diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,21630 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 3084,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00032425421530479895,
+      "grad_norm": 1.1135283708572388,
+      "learning_rate": 1.0000000000000001e-07,
+      "loss": 2.6361,
+      "step": 1
+    },
+    {
+      "epoch": 0.0006485084306095979,
+      "grad_norm": 4.623935222625732,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 5.135,
+      "step": 2
+    },
+    {
+      "epoch": 0.0009727626459143969,
+      "grad_norm": 1.4618090391159058,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 3.1062,
+      "step": 3
+    },
+    {
+      "epoch": 0.0012970168612191958,
+      "grad_norm": 1.6558266878128052,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 3.1041,
+      "step": 4
+    },
+    {
+      "epoch": 0.001621271076523995,
+      "grad_norm": 4.603311538696289,
+      "learning_rate": 5e-07,
+      "loss": 4.9758,
+      "step": 5
+    },
+    {
+      "epoch": 0.0019455252918287938,
+      "grad_norm": 7.0271525382995605,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 6.5161,
+      "step": 6
+    },
+    {
+      "epoch": 0.002269779507133593,
+      "grad_norm": 1.913426160812378,
+      "learning_rate": 7.000000000000001e-07,
+      "loss": 3.1082,
+      "step": 7
+    },
+    {
+      "epoch": 0.0025940337224383916,
+      "grad_norm": 3.367969512939453,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 4.0623,
+      "step": 8
+    },
+    {
+      "epoch": 0.0029182879377431907,
+      "grad_norm": 4.887551784515381,
+      "learning_rate": 9e-07,
+      "loss": 5.1213,
+      "step": 9
+    },
+    {
+      "epoch": 0.00324254215304799,
+      "grad_norm": 3.567915201187134,
+      "learning_rate": 1e-06,
+      "loss": 4.2036,
+      "step": 10
+    },
+    {
+      "epoch": 0.0035667963683527885,
+      "grad_norm": 3.2461798191070557,
+      "learning_rate": 1.1e-06,
+      "loss": 4.1854,
+      "step": 11
+    },
+    {
+      "epoch": 0.0038910505836575876,
+      "grad_norm": 7.888861179351807,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 7.037,
+      "step": 12
+    },
+    {
+      "epoch": 0.004215304798962386,
+      "grad_norm": 1.5571101903915405,
+      "learning_rate": 1.3e-06,
+      "loss": 2.9691,
+      "step": 13
+    },
+    {
+      "epoch": 0.004539559014267186,
+      "grad_norm": 5.124597072601318,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 5.3845,
+      "step": 14
+    },
+    {
+      "epoch": 0.0048638132295719845,
+      "grad_norm": 5.182037353515625,
+      "learning_rate": 1.5e-06,
+      "loss": 5.2594,
+      "step": 15
+    },
+    {
+      "epoch": 0.005188067444876783,
+      "grad_norm": 6.674919605255127,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 6.298,
+      "step": 16
+    },
+    {
+      "epoch": 0.005512321660181583,
+      "grad_norm": 5.65845251083374,
+      "learning_rate": 1.7e-06,
+      "loss": 5.7398,
+      "step": 17
+    },
+    {
+      "epoch": 0.005836575875486381,
+      "grad_norm": 3.8821213245391846,
+      "learning_rate": 1.8e-06,
+      "loss": 4.5671,
+      "step": 18
+    },
+    {
+      "epoch": 0.00616083009079118,
+      "grad_norm": 3.360201120376587,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 4.1446,
+      "step": 19
+    },
+    {
+      "epoch": 0.00648508430609598,
+      "grad_norm": 3.004438638687134,
+      "learning_rate": 2e-06,
+      "loss": 3.8176,
+      "step": 20
+    },
+    {
+      "epoch": 0.006809338521400778,
+      "grad_norm": 2.982933282852173,
+      "learning_rate": 2.1000000000000002e-06,
+      "loss": 3.5871,
+      "step": 21
+    },
+    {
+      "epoch": 0.007133592736705577,
+      "grad_norm": 6.308475971221924,
+      "learning_rate": 2.2e-06,
+      "loss": 5.8924,
+      "step": 22
+    },
+    {
+      "epoch": 0.0074578469520103765,
+      "grad_norm": 2.6446545124053955,
+      "learning_rate": 2.3e-06,
+      "loss": 3.5275,
+      "step": 23
+    },
+    {
+      "epoch": 0.007782101167315175,
+      "grad_norm": 4.7137861251831055,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 4.979,
+      "step": 24
+    },
+    {
+      "epoch": 0.008106355382619975,
+      "grad_norm": 5.097355365753174,
+      "learning_rate": 2.4999999999999998e-06,
+      "loss": 5.2859,
+      "step": 25
+    },
+    {
+      "epoch": 0.008430609597924773,
+      "grad_norm": 2.9971680641174316,
+      "learning_rate": 2.6e-06,
+      "loss": 3.7364,
+      "step": 26
+    },
+    {
+      "epoch": 0.008754863813229572,
+      "grad_norm": 1.8347759246826172,
+      "learning_rate": 2.7e-06,
+      "loss": 3.1322,
+      "step": 27
+    },
+    {
+      "epoch": 0.009079118028534372,
+      "grad_norm": 2.780092477798462,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 4.0401,
+      "step": 28
+    },
+    {
+      "epoch": 0.00940337224383917,
+      "grad_norm": 1.5163801908493042,
+      "learning_rate": 2.9e-06,
+      "loss": 3.0874,
+      "step": 29
+    },
+    {
+      "epoch": 0.009727626459143969,
+      "grad_norm": 1.2515937089920044,
+      "learning_rate": 3e-06,
+      "loss": 3.0699,
+      "step": 30
+    },
+    {
+      "epoch": 0.010051880674448769,
+      "grad_norm": 1.8627383708953857,
+      "learning_rate": 3.1e-06,
+      "loss": 3.3789,
+      "step": 31
+    },
+    {
+      "epoch": 0.010376134889753566,
+      "grad_norm": 2.442812442779541,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 4.4279,
+      "step": 32
+    },
+    {
+      "epoch": 0.010700389105058366,
+      "grad_norm": 1.3900775909423828,
+      "learning_rate": 3.3e-06,
+      "loss": 3.2687,
+      "step": 33
+    },
+    {
+      "epoch": 0.011024643320363165,
+      "grad_norm": 2.0634968280792236,
+      "learning_rate": 3.4e-06,
+      "loss": 4.0014,
+      "step": 34
+    },
+    {
+      "epoch": 0.011348897535667963,
+      "grad_norm": 1.3402507305145264,
+      "learning_rate": 3.5e-06,
+      "loss": 3.2764,
+      "step": 35
+    },
+    {
+      "epoch": 0.011673151750972763,
+      "grad_norm": 1.2550427913665771,
+      "learning_rate": 3.6e-06,
+      "loss": 3.1777,
+      "step": 36
+    },
+    {
+      "epoch": 0.011997405966277562,
+      "grad_norm": 1.6775720119476318,
+      "learning_rate": 3.7e-06,
+      "loss": 3.7751,
+      "step": 37
+    },
+    {
+      "epoch": 0.01232166018158236,
+      "grad_norm": 0.8009628653526306,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 2.6708,
+      "step": 38
+    },
+    {
+      "epoch": 0.01264591439688716,
+      "grad_norm": 1.7612379789352417,
+      "learning_rate": 3.9e-06,
+      "loss": 3.8331,
+      "step": 39
+    },
+    {
+      "epoch": 0.01297016861219196,
+      "grad_norm": 0.9109116196632385,
+      "learning_rate": 4e-06,
+      "loss": 2.9569,
+      "step": 40
+    },
+    {
+      "epoch": 0.013294422827496757,
+      "grad_norm": 1.0260356664657593,
+      "learning_rate": 4.1e-06,
+      "loss": 2.997,
+      "step": 41
+    },
+    {
+      "epoch": 0.013618677042801557,
+      "grad_norm": 1.196084976196289,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 3.1,
+      "step": 42
+    },
+    {
+      "epoch": 0.013942931258106356,
+      "grad_norm": 1.7340625524520874,
+      "learning_rate": 4.3e-06,
+      "loss": 4.1351,
+      "step": 43
+    },
+    {
+      "epoch": 0.014267185473411154,
+      "grad_norm": 3.0273687839508057,
+      "learning_rate": 4.4e-06,
+      "loss": 5.9681,
+      "step": 44
+    },
+    {
+      "epoch": 0.014591439688715954,
+      "grad_norm": 1.9812054634094238,
+      "learning_rate": 4.5e-06,
+      "loss": 4.5326,
+      "step": 45
+    },
+    {
+      "epoch": 0.014915693904020753,
+      "grad_norm": 1.2337762117385864,
+      "learning_rate": 4.6e-06,
+      "loss": 3.2809,
+      "step": 46
+    },
+    {
+      "epoch": 0.01523994811932555,
+      "grad_norm": 1.0402265787124634,
+      "learning_rate": 4.700000000000001e-06,
+      "loss": 3.3071,
+      "step": 47
+    },
+    {
+      "epoch": 0.01556420233463035,
+      "grad_norm": 0.5975327491760254,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 2.7973,
+      "step": 48
+    },
+    {
+      "epoch": 0.015888456549935148,
+      "grad_norm": 0.8598670363426208,
+      "learning_rate": 4.9e-06,
+      "loss": 3.1448,
+      "step": 49
+    },
+    {
+      "epoch": 0.01621271076523995,
+      "grad_norm": 0.4908126890659332,
+      "learning_rate": 4.9999999999999996e-06,
+      "loss": 2.3799,
+      "step": 50
+    },
+    {
+      "epoch": 0.016536964980544747,
+      "grad_norm": 0.8438050746917725,
+      "learning_rate": 5.1e-06,
+      "loss": 3.2936,
+      "step": 51
+    },
+    {
+      "epoch": 0.016861219195849545,
+      "grad_norm": 1.605458378791809,
+      "learning_rate": 5.2e-06,
+      "loss": 4.4773,
+      "step": 52
+    },
+    {
+      "epoch": 0.017185473411154346,
+      "grad_norm": 0.4867059290409088,
+      "learning_rate": 5.3e-06,
+      "loss": 2.4534,
+      "step": 53
+    },
+    {
+      "epoch": 0.017509727626459144,
+      "grad_norm": 1.0149558782577515,
+      "learning_rate": 5.4e-06,
+      "loss": 3.4858,
+      "step": 54
+    },
+    {
+      "epoch": 0.017833981841763942,
+      "grad_norm": 0.6530719995498657,
+      "learning_rate": 5.5e-06,
+      "loss": 2.7231,
+      "step": 55
+    },
+    {
+      "epoch": 0.018158236057068743,
+      "grad_norm": 0.5719746947288513,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 2.5289,
+      "step": 56
+    },
+    {
+      "epoch": 0.01848249027237354,
+      "grad_norm": 0.9637179374694824,
+      "learning_rate": 5.7000000000000005e-06,
+      "loss": 3.4664,
+      "step": 57
+    },
+    {
+      "epoch": 0.01880674448767834,
+      "grad_norm": 0.7422669529914856,
+      "learning_rate": 5.8e-06,
+      "loss": 3.2202,
+      "step": 58
+    },
+    {
+      "epoch": 0.01913099870298314,
+      "grad_norm": 1.1487177610397339,
+      "learning_rate": 5.899999999999999e-06,
+      "loss": 4.4312,
+      "step": 59
+    },
+    {
+      "epoch": 0.019455252918287938,
+      "grad_norm": 1.2654414176940918,
+      "learning_rate": 6e-06,
+      "loss": 4.0673,
+      "step": 60
+    },
+    {
+      "epoch": 0.019779507133592736,
+      "grad_norm": 0.8139782547950745,
+      "learning_rate": 6.1e-06,
+      "loss": 3.5977,
+      "step": 61
+    },
+    {
+      "epoch": 0.020103761348897537,
+      "grad_norm": 0.5601393580436707,
+      "learning_rate": 6.2e-06,
+      "loss": 3.0429,
+      "step": 62
+    },
+    {
+      "epoch": 0.020428015564202335,
+      "grad_norm": 0.4398249387741089,
+      "learning_rate": 6.3e-06,
+      "loss": 2.4796,
+      "step": 63
+    },
+    {
+      "epoch": 0.020752269779507133,
+      "grad_norm": 0.6434876322746277,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 3.1017,
+      "step": 64
+    },
+    {
+      "epoch": 0.021076523994811934,
+      "grad_norm": 1.6901971101760864,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 4.2524,
+      "step": 65
+    },
+    {
+      "epoch": 0.021400778210116732,
+      "grad_norm": 0.6332772970199585,
+      "learning_rate": 6.6e-06,
+      "loss": 3.1918,
+      "step": 66
+    },
+    {
+      "epoch": 0.02172503242542153,
+      "grad_norm": 0.5238966941833496,
+      "learning_rate": 6.7e-06,
+      "loss": 2.9542,
+      "step": 67
+    },
+    {
+      "epoch": 0.02204928664072633,
+      "grad_norm": 0.9129756689071655,
+      "learning_rate": 6.8e-06,
+      "loss": 3.6116,
+      "step": 68
+    },
+    {
+      "epoch": 0.02237354085603113,
+      "grad_norm": 0.6836019158363342,
+      "learning_rate": 6.900000000000001e-06,
+      "loss": 2.9913,
+      "step": 69
+    },
+    {
+      "epoch": 0.022697795071335927,
+      "grad_norm": 0.5503202676773071,
+      "learning_rate": 7e-06,
+      "loss": 3.081,
+      "step": 70
+    },
+    {
+      "epoch": 0.023022049286640728,
+      "grad_norm": 0.7651568055152893,
+      "learning_rate": 7.1e-06,
+      "loss": 3.6429,
+      "step": 71
+    },
+    {
+      "epoch": 0.023346303501945526,
+      "grad_norm": 0.8701173067092896,
+      "learning_rate": 7.2e-06,
+      "loss": 4.156,
+      "step": 72
+    },
+    {
+      "epoch": 0.023670557717250323,
+      "grad_norm": 0.7399364113807678,
+      "learning_rate": 7.3e-06,
+      "loss": 3.2263,
+      "step": 73
+    },
+    {
+      "epoch": 0.023994811932555125,
+      "grad_norm": 0.6664344668388367,
+      "learning_rate": 7.4e-06,
+      "loss": 3.6347,
+      "step": 74
+    },
+    {
+      "epoch": 0.024319066147859923,
+      "grad_norm": 0.825458288192749,
+      "learning_rate": 7.5e-06,
+      "loss": 3.5788,
+      "step": 75
+    },
+    {
+      "epoch": 0.02464332036316472,
+      "grad_norm": 2.444143056869507,
+      "learning_rate": 7.600000000000001e-06,
+      "loss": 4.122,
+      "step": 76
+    },
+    {
+      "epoch": 0.02496757457846952,
+      "grad_norm": 0.43582776188850403,
+      "learning_rate": 7.699999999999999e-06,
+      "loss": 2.5745,
+      "step": 77
+    },
+    {
+      "epoch": 0.02529182879377432,
+      "grad_norm": 0.9375268220901489,
+      "learning_rate": 7.8e-06,
+      "loss": 3.7733,
+      "step": 78
+    },
+    {
+      "epoch": 0.025616083009079117,
+      "grad_norm": 0.5359544157981873,
+      "learning_rate": 7.899999999999999e-06,
+      "loss": 2.8973,
+      "step": 79
+    },
+    {
+      "epoch": 0.02594033722438392,
+      "grad_norm": 0.4063528776168823,
+      "learning_rate": 8e-06,
+      "loss": 2.5174,
+      "step": 80
+    },
+    {
+      "epoch": 0.026264591439688716,
+      "grad_norm": 0.5341715216636658,
+      "learning_rate": 8.1e-06,
+      "loss": 2.4619,
+      "step": 81
+    },
+    {
+      "epoch": 0.026588845654993514,
+      "grad_norm": 0.5937154293060303,
+      "learning_rate": 8.2e-06,
+      "loss": 3.6104,
+      "step": 82
+    },
+    {
+      "epoch": 0.026913099870298315,
+      "grad_norm": 0.7084957957267761,
+      "learning_rate": 8.3e-06,
+      "loss": 4.0348,
+      "step": 83
+    },
+    {
+      "epoch": 0.027237354085603113,
+      "grad_norm": 0.5360614061355591,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 3.2098,
+      "step": 84
+    },
+    {
+      "epoch": 0.02756160830090791,
+      "grad_norm": 2.3136301040649414,
+      "learning_rate": 8.5e-06,
+      "loss": 4.0799,
+      "step": 85
+    },
+    {
+      "epoch": 0.027885862516212712,
+      "grad_norm": 0.35532355308532715,
+      "learning_rate": 8.6e-06,
+      "loss": 2.3292,
+      "step": 86
+    },
+    {
+      "epoch": 0.02821011673151751,
+      "grad_norm": 0.5682448148727417,
+      "learning_rate": 8.7e-06,
+      "loss": 3.2899,
+      "step": 87
+    },
+    {
+      "epoch": 0.028534370946822308,
+      "grad_norm": 0.5726275444030762,
+      "learning_rate": 8.8e-06,
+      "loss": 2.9937,
+      "step": 88
+    },
+    {
+      "epoch": 0.02885862516212711,
+      "grad_norm": 0.48312482237815857,
+      "learning_rate": 8.900000000000001e-06,
+      "loss": 2.6601,
+      "step": 89
+    },
+    {
+      "epoch": 0.029182879377431907,
+      "grad_norm": 0.4579432010650635,
+      "learning_rate": 9e-06,
+      "loss": 2.8515,
+      "step": 90
+    },
+    {
+      "epoch": 0.029507133592736705,
+      "grad_norm": 0.9770523905754089,
+      "learning_rate": 9.100000000000001e-06,
+      "loss": 3.3337,
+      "step": 91
+    },
+    {
+      "epoch": 0.029831387808041506,
+      "grad_norm": 0.9641193747520447,
+      "learning_rate": 9.2e-06,
+      "loss": 3.0227,
+      "step": 92
+    },
+    {
+      "epoch": 0.030155642023346304,
+      "grad_norm": 0.3334672152996063,
+      "learning_rate": 9.3e-06,
+      "loss": 2.3238,
+      "step": 93
+    },
+    {
+      "epoch": 0.0304798962386511,
+      "grad_norm": 0.8481307029724121,
+      "learning_rate": 9.400000000000001e-06,
+      "loss": 3.4508,
+      "step": 94
+    },
+    {
+      "epoch": 0.030804150453955903,
+      "grad_norm": 0.8653482794761658,
+      "learning_rate": 9.5e-06,
+      "loss": 3.7898,
+      "step": 95
+    },
+    {
+      "epoch": 0.0311284046692607,
+      "grad_norm": 0.6192666888237,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 3.3661,
+      "step": 96
+    },
+    {
+      "epoch": 0.0314526588845655,
+      "grad_norm": 0.8940184712409973,
+      "learning_rate": 9.699999999999999e-06,
+      "loss": 3.0296,
+      "step": 97
+    },
+    {
+      "epoch": 0.031776913099870296,
+      "grad_norm": 0.4716889560222626,
+      "learning_rate": 9.8e-06,
+      "loss": 2.699,
+      "step": 98
+    },
+    {
+      "epoch": 0.032101167315175094,
+      "grad_norm": 0.6614878177642822,
+      "learning_rate": 9.9e-06,
+      "loss": 2.9141,
+      "step": 99
+    },
+    {
+      "epoch": 0.0324254215304799,
+      "grad_norm": 0.7267977595329285,
+      "learning_rate": 9.999999999999999e-06,
+      "loss": 3.3407,
+      "step": 100
+    },
+    {
+      "epoch": 0.0327496757457847,
+      "grad_norm": 0.991168737411499,
+      "learning_rate": 1.01e-05,
+      "loss": 3.2398,
+      "step": 101
+    },
+    {
+      "epoch": 0.033073929961089495,
+      "grad_norm": 0.5779916644096375,
+      "learning_rate": 1.02e-05,
+      "loss": 3.1671,
+      "step": 102
+    },
+    {
+      "epoch": 0.03339818417639429,
+      "grad_norm": 0.8053655028343201,
+      "learning_rate": 1.03e-05,
+      "loss": 3.3144,
+      "step": 103
+    },
+    {
+      "epoch": 0.03372243839169909,
+      "grad_norm": 0.3501691222190857,
+      "learning_rate": 1.04e-05,
+      "loss": 2.3235,
+      "step": 104
+    },
+    {
+      "epoch": 0.03404669260700389,
+      "grad_norm": 0.70566725730896,
+      "learning_rate": 1.05e-05,
+      "loss": 3.4987,
+      "step": 105
+    },
+    {
+      "epoch": 0.03437094682230869,
+      "grad_norm": 0.7160869836807251,
+      "learning_rate": 1.06e-05,
+      "loss": 3.2092,
+      "step": 106
+    },
+    {
+      "epoch": 0.03469520103761349,
+      "grad_norm": 0.8717048764228821,
+      "learning_rate": 1.0700000000000001e-05,
+      "loss": 4.0119,
+      "step": 107
+    },
+    {
+      "epoch": 0.03501945525291829,
+      "grad_norm": 0.40830761194229126,
+      "learning_rate": 1.08e-05,
+      "loss": 2.65,
+      "step": 108
+    },
+    {
+      "epoch": 0.035343709468223086,
+      "grad_norm": 1.2620882987976074,
+      "learning_rate": 1.09e-05,
+      "loss": 3.7694,
+      "step": 109
+    },
+    {
+      "epoch": 0.035667963683527884,
+      "grad_norm": 1.0694531202316284,
+      "learning_rate": 1.1e-05,
+      "loss": 3.9461,
+      "step": 110
+    },
+    {
+      "epoch": 0.03599221789883268,
+      "grad_norm": 0.8394907712936401,
+      "learning_rate": 1.11e-05,
+      "loss": 3.5299,
+      "step": 111
+    },
+    {
+      "epoch": 0.03631647211413749,
+      "grad_norm": 0.717319667339325,
+      "learning_rate": 1.1200000000000001e-05,
+      "loss": 3.3029,
+      "step": 112
+    },
+    {
+      "epoch": 0.036640726329442284,
+      "grad_norm": 0.6590234637260437,
+      "learning_rate": 1.13e-05,
+      "loss": 3.0881,
+      "step": 113
+    },
+    {
+      "epoch": 0.03696498054474708,
+      "grad_norm": 0.6868249773979187,
+      "learning_rate": 1.1400000000000001e-05,
+      "loss": 2.9888,
+      "step": 114
+    },
+    {
+      "epoch": 0.03728923476005188,
+      "grad_norm": 1.3315736055374146,
+      "learning_rate": 1.1500000000000002e-05,
+      "loss": 3.686,
+      "step": 115
+    },
+    {
+      "epoch": 0.03761348897535668,
+      "grad_norm": 0.41429704427719116,
+      "learning_rate": 1.16e-05,
+      "loss": 2.7828,
+      "step": 116
+    },
+    {
+      "epoch": 0.037937743190661476,
+      "grad_norm": 0.6155628561973572,
+      "learning_rate": 1.1700000000000001e-05,
+      "loss": 3.1063,
+      "step": 117
+    },
+    {
+      "epoch": 0.03826199740596628,
+      "grad_norm": 0.6803083419799805,
+      "learning_rate": 1.1799999999999999e-05,
+      "loss": 2.9712,
+      "step": 118
+    },
+    {
+      "epoch": 0.03858625162127108,
+      "grad_norm": 0.8247591853141785,
+      "learning_rate": 1.19e-05,
+      "loss": 3.5498,
+      "step": 119
+    },
+    {
+      "epoch": 0.038910505836575876,
+      "grad_norm": 0.6183271408081055,
+      "learning_rate": 1.2e-05,
+      "loss": 2.9966,
+      "step": 120
+    },
+    {
+      "epoch": 0.039234760051880674,
+      "grad_norm": 0.5294755697250366,
+      "learning_rate": 1.21e-05,
+      "loss": 2.7481,
+      "step": 121
+    },
+    {
+      "epoch": 0.03955901426718547,
+      "grad_norm": 0.6849741339683533,
+      "learning_rate": 1.22e-05,
+      "loss": 3.1042,
+      "step": 122
+    },
+    {
+      "epoch": 0.03988326848249027,
+      "grad_norm": 0.43405941128730774,
+      "learning_rate": 1.2299999999999999e-05,
+      "loss": 2.8666,
+      "step": 123
+    },
+    {
+      "epoch": 0.040207522697795074,
+      "grad_norm": 0.43856894969940186,
+      "learning_rate": 1.24e-05,
+      "loss": 2.6219,
+      "step": 124
+    },
+    {
+      "epoch": 0.04053177691309987,
+      "grad_norm": 0.5562130808830261,
+      "learning_rate": 1.25e-05,
+      "loss": 2.6611,
+      "step": 125
+    },
+    {
+      "epoch": 0.04085603112840467,
+      "grad_norm": 0.9194834232330322,
+      "learning_rate": 1.26e-05,
+      "loss": 3.9853,
+      "step": 126
+    },
+    {
+      "epoch": 0.04118028534370947,
+      "grad_norm": 0.3443896770477295,
+      "learning_rate": 1.27e-05,
+      "loss": 2.224,
+      "step": 127
+    },
+    {
+      "epoch": 0.041504539559014265,
+      "grad_norm": 0.4816150963306427,
+      "learning_rate": 1.2800000000000001e-05,
+      "loss": 2.6323,
+      "step": 128
+    },
+    {
+      "epoch": 0.04182879377431906,
+      "grad_norm": 0.9299262166023254,
+      "learning_rate": 1.29e-05,
+      "loss": 3.8284,
+      "step": 129
+    },
+    {
+      "epoch": 0.04215304798962387,
+      "grad_norm": 0.45185068249702454,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 2.6099,
+      "step": 130
+    },
+    {
+      "epoch": 0.042477302204928666,
+      "grad_norm": 0.5144737362861633,
+      "learning_rate": 1.31e-05,
+      "loss": 2.684,
+      "step": 131
+    },
+    {
+      "epoch": 0.042801556420233464,
+      "grad_norm": 0.5446034669876099,
+      "learning_rate": 1.32e-05,
+      "loss": 2.592,
+      "step": 132
+    },
+    {
+      "epoch": 0.04312581063553826,
+      "grad_norm": 0.8398793935775757,
+      "learning_rate": 1.3300000000000001e-05,
+      "loss": 3.02,
+      "step": 133
+    },
+    {
+      "epoch": 0.04345006485084306,
+      "grad_norm": 0.8170154094696045,
+      "learning_rate": 1.34e-05,
+      "loss": 3.7581,
+      "step": 134
+    },
+    {
+      "epoch": 0.04377431906614786,
+      "grad_norm": 0.6092444062232971,
+      "learning_rate": 1.3500000000000001e-05,
+      "loss": 3.374,
+      "step": 135
+    },
+    {
+      "epoch": 0.04409857328145266,
+      "grad_norm": 0.552638053894043,
+      "learning_rate": 1.36e-05,
+      "loss": 2.6497,
+      "step": 136
+    },
+    {
+      "epoch": 0.04442282749675746,
+      "grad_norm": 0.49857938289642334,
+      "learning_rate": 1.3700000000000001e-05,
+      "loss": 2.9354,
+      "step": 137
+    },
+    {
+      "epoch": 0.04474708171206226,
+      "grad_norm": 0.6805559396743774,
+      "learning_rate": 1.3800000000000002e-05,
+      "loss": 2.7692,
+      "step": 138
+    },
+    {
+      "epoch": 0.045071335927367055,
+      "grad_norm": 0.586958110332489,
+      "learning_rate": 1.39e-05,
+      "loss": 2.8949,
+      "step": 139
+    },
+    {
+      "epoch": 0.04539559014267185,
+      "grad_norm": 1.1846500635147095,
+      "learning_rate": 1.4e-05,
+      "loss": 3.5016,
+      "step": 140
+    },
+    {
+      "epoch": 0.04571984435797665,
+      "grad_norm": 0.4898151457309723,
+      "learning_rate": 1.4099999999999999e-05,
+      "loss": 2.6688,
+      "step": 141
+    },
+    {
+      "epoch": 0.046044098573281456,
+      "grad_norm": 0.7297990918159485,
+      "learning_rate": 1.42e-05,
+      "loss": 2.6528,
+      "step": 142
+    },
+    {
+      "epoch": 0.04636835278858625,
+      "grad_norm": 0.6000905632972717,
+      "learning_rate": 1.43e-05,
+      "loss": 2.5211,
+      "step": 143
+    },
+    {
+      "epoch": 0.04669260700389105,
+      "grad_norm": 0.6208663582801819,
+      "learning_rate": 1.44e-05,
+      "loss": 2.8747,
+      "step": 144
+    },
+    {
+      "epoch": 0.04701686121919585,
+      "grad_norm": 0.5884286165237427,
+      "learning_rate": 1.45e-05,
+      "loss": 2.6542,
+      "step": 145
+    },
+    {
+      "epoch": 0.04734111543450065,
+      "grad_norm": 0.4307498335838318,
+      "learning_rate": 1.46e-05,
+      "loss": 2.6078,
+      "step": 146
+    },
+    {
+      "epoch": 0.047665369649805445,
+      "grad_norm": 0.4269621968269348,
+      "learning_rate": 1.47e-05,
+      "loss": 2.421,
+      "step": 147
+    },
+    {
+      "epoch": 0.04798962386511025,
+      "grad_norm": 0.655303955078125,
+      "learning_rate": 1.48e-05,
+      "loss": 2.8831,
+      "step": 148
+    },
+    {
+      "epoch": 0.04831387808041505,
+      "grad_norm": 1.0429331064224243,
+      "learning_rate": 1.49e-05,
+      "loss": 3.4136,
+      "step": 149
+    },
+    {
+      "epoch": 0.048638132295719845,
+      "grad_norm": 0.6508225202560425,
+      "learning_rate": 1.5e-05,
+      "loss": 2.9192,
+      "step": 150
+    },
+    {
+      "epoch": 0.04896238651102464,
+      "grad_norm": 0.663862407207489,
+      "learning_rate": 1.51e-05,
+      "loss": 3.1745,
+      "step": 151
+    },
+    {
+      "epoch": 0.04928664072632944,
+      "grad_norm": 0.581112802028656,
+      "learning_rate": 1.5200000000000002e-05,
+      "loss": 2.9268,
+      "step": 152
+    },
+    {
+      "epoch": 0.04961089494163424,
+      "grad_norm": 0.7333909869194031,
+      "learning_rate": 1.53e-05,
+      "loss": 2.8906,
+      "step": 153
+    },
+    {
+      "epoch": 0.04993514915693904,
+      "grad_norm": 0.9611436724662781,
+      "learning_rate": 1.5399999999999998e-05,
+      "loss": 3.9098,
+      "step": 154
+    },
+    {
+      "epoch": 0.05025940337224384,
+      "grad_norm": 0.5327355265617371,
+      "learning_rate": 1.55e-05,
+      "loss": 2.6769,
+      "step": 155
+    },
+    {
+      "epoch": 0.05058365758754864,
+      "grad_norm": 0.6427381038665771,
+      "learning_rate": 1.56e-05,
+      "loss": 2.8091,
+      "step": 156
+    },
+    {
+      "epoch": 0.05090791180285344,
+      "grad_norm": 1.098859190940857,
+      "learning_rate": 1.57e-05,
+      "loss": 3.1322,
+      "step": 157
+    },
+    {
+      "epoch": 0.051232166018158234,
+      "grad_norm": 0.48087596893310547,
+      "learning_rate": 1.5799999999999998e-05,
+      "loss": 2.3415,
+      "step": 158
+    },
+    {
+      "epoch": 0.05155642023346303,
+      "grad_norm": 1.2534234523773193,
+      "learning_rate": 1.59e-05,
+      "loss": 3.4838,
+      "step": 159
+    },
+    {
+      "epoch": 0.05188067444876784,
+      "grad_norm": 0.5815152525901794,
+      "learning_rate": 1.6e-05,
+      "loss": 2.5035,
+      "step": 160
+    },
+    {
+      "epoch": 0.052204928664072635,
+      "grad_norm": 0.604430615901947,
+      "learning_rate": 1.61e-05,
+      "loss": 2.5306,
+      "step": 161
+    },
+    {
+      "epoch": 0.05252918287937743,
+      "grad_norm": 0.4218912720680237,
+      "learning_rate": 1.62e-05,
+      "loss": 2.1495,
+      "step": 162
+    },
+    {
+      "epoch": 0.05285343709468223,
+      "grad_norm": 1.389872431755066,
+      "learning_rate": 1.63e-05,
+      "loss": 3.1309,
+      "step": 163
+    },
+    {
+      "epoch": 0.05317769130998703,
+      "grad_norm": 1.1290407180786133,
+      "learning_rate": 1.64e-05,
+      "loss": 2.8799,
+      "step": 164
+    },
+    {
+      "epoch": 0.053501945525291826,
+      "grad_norm": 1.4107460975646973,
+      "learning_rate": 1.65e-05,
+      "loss": 3.186,
+      "step": 165
+    },
+    {
+      "epoch": 0.05382619974059663,
+      "grad_norm": 0.875864565372467,
+      "learning_rate": 1.66e-05,
+      "loss": 3.2862,
+      "step": 166
+    },
+    {
+      "epoch": 0.05415045395590143,
+      "grad_norm": 0.5816508531570435,
+      "learning_rate": 1.67e-05,
+      "loss": 2.919,
+      "step": 167
+    },
+    {
+      "epoch": 0.054474708171206226,
+      "grad_norm": 0.5963166356086731,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 3.0483,
+      "step": 168
+    },
+    {
+      "epoch": 0.054798962386511024,
+      "grad_norm": 0.5069915056228638,
+      "learning_rate": 1.69e-05,
+      "loss": 2.5688,
+      "step": 169
+    },
+    {
+      "epoch": 0.05512321660181582,
+      "grad_norm": 0.6019675135612488,
+      "learning_rate": 1.7e-05,
+      "loss": 2.596,
+      "step": 170
+    },
+    {
+      "epoch": 0.05544747081712062,
+      "grad_norm": 1.2352375984191895,
+      "learning_rate": 1.71e-05,
+      "loss": 3.5148,
+      "step": 171
+    },
+    {
+      "epoch": 0.055771725032425425,
+      "grad_norm": 0.7280648946762085,
+      "learning_rate": 1.72e-05,
+      "loss": 3.279,
+      "step": 172
+    },
+    {
+      "epoch": 0.05609597924773022,
+      "grad_norm": 0.67685866355896,
+      "learning_rate": 1.73e-05,
+      "loss": 2.8957,
+      "step": 173
+    },
+    {
+      "epoch": 0.05642023346303502,
+      "grad_norm": 1.390796184539795,
+      "learning_rate": 1.74e-05,
+      "loss": 3.7937,
+      "step": 174
+    },
+    {
+      "epoch": 0.05674448767833982,
+      "grad_norm": 0.7078388929367065,
+      "learning_rate": 1.7500000000000002e-05,
+      "loss": 2.4725,
+      "step": 175
+    },
+    {
+      "epoch": 0.057068741893644616,
+      "grad_norm": 0.6902285814285278,
+      "learning_rate": 1.76e-05,
+      "loss": 3.1511,
+      "step": 176
+    },
+    {
+      "epoch": 0.057392996108949414,
+      "grad_norm": 0.9627773761749268,
+      "learning_rate": 1.77e-05,
+      "loss": 3.9891,
+      "step": 177
+    },
+    {
+      "epoch": 0.05771725032425422,
+      "grad_norm": 0.7457598447799683,
+      "learning_rate": 1.7800000000000002e-05,
+      "loss": 2.9725,
+      "step": 178
+    },
+    {
+      "epoch": 0.058041504539559016,
+      "grad_norm": 0.701324999332428,
+      "learning_rate": 1.79e-05,
+      "loss": 3.2402,
+      "step": 179
+    },
+    {
+      "epoch": 0.058365758754863814,
+      "grad_norm": 0.5984466671943665,
+      "learning_rate": 1.8e-05,
+      "loss": 2.7501,
+      "step": 180
+    },
+    {
+      "epoch": 0.05869001297016861,
+      "grad_norm": 0.6049157977104187,
+      "learning_rate": 1.8100000000000003e-05,
+      "loss": 3.6324,
+      "step": 181
+    },
+    {
+      "epoch": 0.05901426718547341,
+      "grad_norm": 0.6157858371734619,
+      "learning_rate": 1.8200000000000002e-05,
+      "loss": 2.8777,
+      "step": 182
+    },
+    {
+      "epoch": 0.05933852140077821,
+      "grad_norm": 0.5198885202407837,
+      "learning_rate": 1.83e-05,
+      "loss": 2.5689,
+      "step": 183
+    },
+    {
+      "epoch": 0.05966277561608301,
+      "grad_norm": 0.7698058485984802,
+      "learning_rate": 1.84e-05,
+      "loss": 2.6931,
+      "step": 184
+    },
+    {
+      "epoch": 0.05998702983138781,
+      "grad_norm": 0.5717703104019165,
+      "learning_rate": 1.8500000000000002e-05,
+      "loss": 2.8298,
+      "step": 185
+    },
+    {
+      "epoch": 0.06031128404669261,
+      "grad_norm": 0.47829926013946533,
+      "learning_rate": 1.86e-05,
+      "loss": 2.57,
+      "step": 186
+    },
+    {
+      "epoch": 0.060635538261997406,
+      "grad_norm": 0.9569224119186401,
+      "learning_rate": 1.87e-05,
+      "loss": 3.2592,
+      "step": 187
+    },
+    {
+      "epoch": 0.0609597924773022,
+      "grad_norm": 0.5631130933761597,
+      "learning_rate": 1.8800000000000003e-05,
+      "loss": 2.7858,
+      "step": 188
+    },
+    {
+      "epoch": 0.061284046692607,
+      "grad_norm": 0.5489674210548401,
+      "learning_rate": 1.8900000000000002e-05,
+      "loss": 2.5301,
+      "step": 189
+    },
+    {
+      "epoch": 0.061608300907911806,
+      "grad_norm": 1.085650086402893,
+      "learning_rate": 1.9e-05,
+      "loss": 3.9078,
+      "step": 190
+    },
+    {
+      "epoch": 0.061932555123216604,
+      "grad_norm": 0.45189717411994934,
+      "learning_rate": 1.9100000000000003e-05,
+      "loss": 2.4013,
+      "step": 191
+    },
+    {
+      "epoch": 0.0622568093385214,
+      "grad_norm": 0.8459087610244751,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 3.6616,
+      "step": 192
+    },
+    {
+      "epoch": 0.0625810635538262,
+      "grad_norm": 0.799480676651001,
+      "learning_rate": 1.9299999999999998e-05,
+      "loss": 3.0142,
+      "step": 193
+    },
+    {
+      "epoch": 0.062905317769131,
+      "grad_norm": 0.5215678811073303,
+      "learning_rate": 1.9399999999999997e-05,
+      "loss": 2.2615,
+      "step": 194
+    },
+    {
+      "epoch": 0.0632295719844358,
+      "grad_norm": 0.589775025844574,
+      "learning_rate": 1.95e-05,
+      "loss": 2.3304,
+      "step": 195
+    },
+    {
+      "epoch": 0.06355382619974059,
+      "grad_norm": 0.7861542105674744,
+      "learning_rate": 1.96e-05,
+      "loss": 2.8585,
+      "step": 196
+    },
+    {
+      "epoch": 0.06387808041504539,
+      "grad_norm": 0.7212756276130676,
+      "learning_rate": 1.9699999999999998e-05,
+      "loss": 2.7442,
+      "step": 197
+    },
+    {
+      "epoch": 0.06420233463035019,
+      "grad_norm": 0.6401896476745605,
+      "learning_rate": 1.98e-05,
+      "loss": 2.9558,
+      "step": 198
+    },
+    {
+      "epoch": 0.064526588845655,
+      "grad_norm": 0.680672824382782,
+      "learning_rate": 1.99e-05,
+      "loss": 3.0006,
+      "step": 199
+    },
+    {
+      "epoch": 0.0648508430609598,
+      "grad_norm": 0.6738709807395935,
+      "learning_rate": 1.9999999999999998e-05,
+      "loss": 2.9388,
+      "step": 200
+    },
+    {
+      "epoch": 0.0651750972762646,
+      "grad_norm": 0.7580153346061707,
+      "learning_rate": 2.01e-05,
+      "loss": 2.8922,
+      "step": 201
+    },
+    {
+      "epoch": 0.0654993514915694,
+      "grad_norm": 0.542972981929779,
+      "learning_rate": 2.02e-05,
+      "loss": 2.857,
+      "step": 202
+    },
+    {
+      "epoch": 0.06582360570687419,
+      "grad_norm": 0.98100745677948,
+      "learning_rate": 2.03e-05,
+      "loss": 2.8491,
+      "step": 203
+    },
+    {
+      "epoch": 0.06614785992217899,
+      "grad_norm": 1.143560767173767,
+      "learning_rate": 2.04e-05,
+      "loss": 3.6273,
+      "step": 204
+    },
+    {
+      "epoch": 0.06647211413748379,
+      "grad_norm": 0.942961573600769,
+      "learning_rate": 2.05e-05,
+      "loss": 3.253,
+      "step": 205
+    },
+    {
+      "epoch": 0.06679636835278858,
+      "grad_norm": 1.0073987245559692,
+      "learning_rate": 2.06e-05,
+      "loss": 3.2947,
+      "step": 206
+    },
+    {
+      "epoch": 0.06712062256809338,
+      "grad_norm": 0.4021376669406891,
+      "learning_rate": 2.07e-05,
+      "loss": 2.2037,
+      "step": 207
+    },
+    {
+      "epoch": 0.06744487678339818,
+      "grad_norm": 1.0246565341949463,
+      "learning_rate": 2.08e-05,
+      "loss": 3.2796,
+      "step": 208
+    },
+    {
+      "epoch": 0.06776913099870298,
+      "grad_norm": 0.4366953372955322,
+      "learning_rate": 2.09e-05,
+      "loss": 2.2862,
+      "step": 209
+    },
+    {
+      "epoch": 0.06809338521400778,
+      "grad_norm": 0.6385810375213623,
+      "learning_rate": 2.1e-05,
+      "loss": 2.9452,
+      "step": 210
+    },
+    {
+      "epoch": 0.06841763942931259,
+      "grad_norm": 0.6450639367103577,
+      "learning_rate": 2.11e-05,
+      "loss": 2.6522,
+      "step": 211
+    },
+    {
+      "epoch": 0.06874189364461739,
+      "grad_norm": 0.8018207550048828,
+      "learning_rate": 2.12e-05,
+      "loss": 3.1802,
+      "step": 212
+    },
+    {
+      "epoch": 0.06906614785992218,
+      "grad_norm": 0.996006429195404,
+      "learning_rate": 2.13e-05,
+      "loss": 3.115,
+      "step": 213
+    },
+    {
+      "epoch": 0.06939040207522698,
+      "grad_norm": 0.6192570328712463,
+      "learning_rate": 2.1400000000000002e-05,
+      "loss": 2.4743,
+      "step": 214
+    },
+    {
+      "epoch": 0.06971465629053178,
+      "grad_norm": 0.6919751167297363,
+      "learning_rate": 2.15e-05,
+      "loss": 2.8125,
+      "step": 215
+    },
+    {
+      "epoch": 0.07003891050583658,
+      "grad_norm": 0.48847585916519165,
+      "learning_rate": 2.16e-05,
+      "loss": 2.3919,
+      "step": 216
+    },
+    {
+      "epoch": 0.07036316472114137,
+      "grad_norm": 0.9190756678581238,
+      "learning_rate": 2.1700000000000002e-05,
+      "loss": 3.2161,
+      "step": 217
+    },
+    {
+      "epoch": 0.07068741893644617,
+      "grad_norm": 1.0849392414093018,
+      "learning_rate": 2.18e-05,
+      "loss": 3.1678,
+      "step": 218
+    },
+    {
+      "epoch": 0.07101167315175097,
+      "grad_norm": 0.7466356158256531,
+      "learning_rate": 2.19e-05,
+      "loss": 2.4852,
+      "step": 219
+    },
+    {
+      "epoch": 0.07133592736705577,
+      "grad_norm": 1.0212808847427368,
+      "learning_rate": 2.2e-05,
+      "loss": 3.1304,
+      "step": 220
+    },
+    {
+      "epoch": 0.07166018158236057,
+      "grad_norm": 0.5849360227584839,
+      "learning_rate": 2.2100000000000002e-05,
+      "loss": 2.3323,
+      "step": 221
+    },
+    {
+      "epoch": 0.07198443579766536,
+      "grad_norm": 0.8980996012687683,
+      "learning_rate": 2.22e-05,
+      "loss": 2.5766,
+      "step": 222
+    },
+    {
+      "epoch": 0.07230869001297018,
+      "grad_norm": 0.8341001272201538,
+      "learning_rate": 2.23e-05,
+      "loss": 2.9073,
+      "step": 223
+    },
+    {
+      "epoch": 0.07263294422827497,
+      "grad_norm": 0.5301621556282043,
+      "learning_rate": 2.2400000000000002e-05,
+      "loss": 2.8529,
+      "step": 224
+    },
+    {
+      "epoch": 0.07295719844357977,
+      "grad_norm": 0.4614284634590149,
+      "learning_rate": 2.25e-05,
+      "loss": 1.999,
+      "step": 225
+    },
+    {
+      "epoch": 0.07328145265888457,
+      "grad_norm": 0.9701064229011536,
+      "learning_rate": 2.26e-05,
+      "loss": 2.8694,
+      "step": 226
+    },
+    {
+      "epoch": 0.07360570687418937,
+      "grad_norm": 0.6426873207092285,
+      "learning_rate": 2.2700000000000003e-05,
+      "loss": 2.5348,
+      "step": 227
+    },
+    {
+      "epoch": 0.07392996108949416,
+      "grad_norm": 0.5774264931678772,
+      "learning_rate": 2.2800000000000002e-05,
+      "loss": 2.9761,
+      "step": 228
+    },
+    {
+      "epoch": 0.07425421530479896,
+      "grad_norm": 0.4790245294570923,
+      "learning_rate": 2.29e-05,
+      "loss": 2.3544,
+      "step": 229
+    },
+    {
+      "epoch": 0.07457846952010376,
+      "grad_norm": 0.9597713947296143,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 3.5232,
+      "step": 230
+    },
+    {
+      "epoch": 0.07490272373540856,
+      "grad_norm": 0.6499766111373901,
+      "learning_rate": 2.3100000000000002e-05,
+      "loss": 2.8996,
+      "step": 231
+    },
+    {
+      "epoch": 0.07522697795071336,
+      "grad_norm": 0.5623300075531006,
+      "learning_rate": 2.32e-05,
+      "loss": 2.9053,
+      "step": 232
+    },
+    {
+      "epoch": 0.07555123216601815,
+      "grad_norm": 0.5688246488571167,
+      "learning_rate": 2.33e-05,
+      "loss": 2.8407,
+      "step": 233
+    },
+    {
+      "epoch": 0.07587548638132295,
+      "grad_norm": 0.5999271869659424,
+      "learning_rate": 2.3400000000000003e-05,
+      "loss": 2.6887,
+      "step": 234
+    },
+    {
+      "epoch": 0.07619974059662776,
+      "grad_norm": 0.6708094477653503,
+      "learning_rate": 2.3500000000000002e-05,
+      "loss": 3.279,
+      "step": 235
+    },
+    {
+      "epoch": 0.07652399481193256,
+      "grad_norm": 1.0278925895690918,
+      "learning_rate": 2.3599999999999998e-05,
+      "loss": 2.5724,
+      "step": 236
+    },
+    {
+      "epoch": 0.07684824902723736,
+      "grad_norm": 0.638282060623169,
+      "learning_rate": 2.37e-05,
+      "loss": 2.8558,
+      "step": 237
+    },
+    {
+      "epoch": 0.07717250324254216,
+      "grad_norm": 0.7668041586875916,
+      "learning_rate": 2.38e-05,
+      "loss": 3.3197,
+      "step": 238
+    },
+    {
+      "epoch": 0.07749675745784695,
+      "grad_norm": 0.7489424347877502,
+      "learning_rate": 2.3899999999999998e-05,
+      "loss": 3.1294,
+      "step": 239
+    },
+    {
+      "epoch": 0.07782101167315175,
+      "grad_norm": 0.395091712474823,
+      "learning_rate": 2.4e-05,
+      "loss": 2.1787,
+      "step": 240
+    },
+    {
+      "epoch": 0.07814526588845655,
+      "grad_norm": 1.3349931240081787,
+      "learning_rate": 2.41e-05,
+      "loss": 3.7253,
+      "step": 241
+    },
+    {
+      "epoch": 0.07846952010376135,
+      "grad_norm": 0.5610832571983337,
+      "learning_rate": 2.42e-05,
+      "loss": 2.6879,
+      "step": 242
+    },
+    {
+      "epoch": 0.07879377431906615,
+      "grad_norm": 0.9450032114982605,
+      "learning_rate": 2.43e-05,
+      "loss": 3.1517,
+      "step": 243
+    },
+    {
+      "epoch": 0.07911802853437094,
+      "grad_norm": 0.5665799975395203,
+      "learning_rate": 2.44e-05,
+      "loss": 2.3192,
+      "step": 244
+    },
+    {
+      "epoch": 0.07944228274967574,
+      "grad_norm": 0.7385202646255493,
+      "learning_rate": 2.45e-05,
+      "loss": 2.9033,
+      "step": 245
+    },
+    {
+      "epoch": 0.07976653696498054,
+      "grad_norm": 0.6420380473136902,
+      "learning_rate": 2.4599999999999998e-05,
+      "loss": 2.2908,
+      "step": 246
+    },
+    {
+      "epoch": 0.08009079118028535,
+      "grad_norm": 0.6577011346817017,
+      "learning_rate": 2.47e-05,
+      "loss": 2.9001,
+      "step": 247
+    },
+    {
+      "epoch": 0.08041504539559015,
+      "grad_norm": 0.8335537314414978,
+      "learning_rate": 2.48e-05,
+      "loss": 3.1423,
+      "step": 248
+    },
+    {
+      "epoch": 0.08073929961089495,
+      "grad_norm": 0.6767750382423401,
+      "learning_rate": 2.49e-05,
+      "loss": 2.9286,
+      "step": 249
+    },
+    {
+      "epoch": 0.08106355382619974,
+      "grad_norm": 0.4596904218196869,
+      "learning_rate": 2.5e-05,
+      "loss": 2.4935,
+      "step": 250
+    },
+    {
+      "epoch": 0.08138780804150454,
+      "grad_norm": 0.39641085267066956,
+      "learning_rate": 2.51e-05,
+      "loss": 2.4737,
+      "step": 251
+    },
+    {
+      "epoch": 0.08171206225680934,
+      "grad_norm": 0.6800199747085571,
+      "learning_rate": 2.52e-05,
+      "loss": 2.6699,
+      "step": 252
+    },
+    {
+      "epoch": 0.08203631647211414,
+      "grad_norm": 0.8949025869369507,
+      "learning_rate": 2.5300000000000002e-05,
+      "loss": 3.2496,
+      "step": 253
+    },
+    {
+      "epoch": 0.08236057068741894,
+      "grad_norm": 0.893715500831604,
+      "learning_rate": 2.54e-05,
+      "loss": 3.5251,
+      "step": 254
+    },
+    {
+      "epoch": 0.08268482490272373,
+      "grad_norm": 0.6548937559127808,
+      "learning_rate": 2.55e-05,
+      "loss": 3.2173,
+      "step": 255
+    },
+    {
+      "epoch": 0.08300907911802853,
+      "grad_norm": 0.7595838308334351,
+      "learning_rate": 2.5600000000000002e-05,
+      "loss": 3.249,
+      "step": 256
+    },
+    {
+      "epoch": 0.08333333333333333,
+      "grad_norm": 1.198014497756958,
+      "learning_rate": 2.57e-05,
+      "loss": 3.0785,
+      "step": 257
+    },
+    {
+      "epoch": 0.08365758754863813,
+      "grad_norm": 0.6603608131408691,
+      "learning_rate": 2.58e-05,
+      "loss": 2.8998,
+      "step": 258
+    },
+    {
+      "epoch": 0.08398184176394292,
+      "grad_norm": 0.6408155560493469,
+      "learning_rate": 2.59e-05,
+      "loss": 2.9006,
+      "step": 259
+    },
+    {
+      "epoch": 0.08430609597924774,
+      "grad_norm": 0.47020673751831055,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 2.1767,
+      "step": 260
+    },
+    {
+      "epoch": 0.08463035019455253,
+      "grad_norm": 0.9444701075553894,
+      "learning_rate": 2.61e-05,
+      "loss": 3.2923,
+      "step": 261
+    },
+    {
+      "epoch": 0.08495460440985733,
+      "grad_norm": 0.7146949768066406,
+      "learning_rate": 2.62e-05,
+      "loss": 2.8419,
+      "step": 262
+    },
+    {
+      "epoch": 0.08527885862516213,
+      "grad_norm": 0.65700763463974,
+      "learning_rate": 2.6300000000000002e-05,
+      "loss": 2.6358,
+      "step": 263
+    },
+    {
+      "epoch": 0.08560311284046693,
+      "grad_norm": 0.7112804651260376,
+      "learning_rate": 2.64e-05,
+      "loss": 3.3816,
+      "step": 264
+    },
+    {
+      "epoch": 0.08592736705577173,
+      "grad_norm": 0.6024109125137329,
+      "learning_rate": 2.65e-05,
+      "loss": 2.6182,
+      "step": 265
+    },
+    {
+      "epoch": 0.08625162127107652,
+      "grad_norm": 0.6243666410446167,
+      "learning_rate": 2.6600000000000003e-05,
+      "loss": 2.5113,
+      "step": 266
+    },
+    {
+      "epoch": 0.08657587548638132,
+      "grad_norm": 0.6519319415092468,
+      "learning_rate": 2.6700000000000002e-05,
+      "loss": 2.6401,
+      "step": 267
+    },
+    {
+      "epoch": 0.08690012970168612,
+      "grad_norm": 0.5395923852920532,
+      "learning_rate": 2.68e-05,
+      "loss": 2.4758,
+      "step": 268
+    },
+    {
+      "epoch": 0.08722438391699092,
+      "grad_norm": 0.6438803672790527,
+      "learning_rate": 2.69e-05,
+      "loss": 3.0721,
+      "step": 269
+    },
+    {
+      "epoch": 0.08754863813229571,
+      "grad_norm": 0.562529981136322,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 2.9458,
+      "step": 270
+    },
+    {
+      "epoch": 0.08787289234760051,
+      "grad_norm": 0.9241312146186829,
+      "learning_rate": 2.71e-05,
+      "loss": 2.9398,
+      "step": 271
+    },
+    {
+      "epoch": 0.08819714656290532,
+      "grad_norm": 0.735077977180481,
+      "learning_rate": 2.72e-05,
+      "loss": 3.0026,
+      "step": 272
+    },
+    {
+      "epoch": 0.08852140077821012,
+      "grad_norm": 0.6285353899002075,
+      "learning_rate": 2.7300000000000003e-05,
+      "loss": 2.9494,
+      "step": 273
+    },
+    {
+      "epoch": 0.08884565499351492,
+      "grad_norm": 0.7142264246940613,
+      "learning_rate": 2.7400000000000002e-05,
+      "loss": 2.5307,
+      "step": 274
+    },
+    {
+      "epoch": 0.08916990920881972,
+      "grad_norm": 0.46221014857292175,
+      "learning_rate": 2.75e-05,
+      "loss": 2.2512,
+      "step": 275
+    },
+    {
+      "epoch": 0.08949416342412451,
+      "grad_norm": 0.8612362742424011,
+      "learning_rate": 2.7600000000000003e-05,
+      "loss": 3.0653,
+      "step": 276
+    },
+    {
+      "epoch": 0.08981841763942931,
+      "grad_norm": 0.8778195381164551,
+      "learning_rate": 2.7700000000000002e-05,
+      "loss": 2.4644,
+      "step": 277
+    },
+    {
+      "epoch": 0.09014267185473411,
+      "grad_norm": 0.7805315256118774,
+      "learning_rate": 2.78e-05,
+      "loss": 2.8931,
+      "step": 278
+    },
+    {
+      "epoch": 0.09046692607003891,
+      "grad_norm": 1.306007742881775,
+      "learning_rate": 2.79e-05,
+      "loss": 3.7001,
+      "step": 279
+    },
+    {
+      "epoch": 0.0907911802853437,
+      "grad_norm": 0.5904978513717651,
+      "learning_rate": 2.8e-05,
+      "loss": 2.8606,
+      "step": 280
+    },
+    {
+      "epoch": 0.0911154345006485,
+      "grad_norm": 0.5373710989952087,
+      "learning_rate": 2.81e-05,
+      "loss": 2.434,
+      "step": 281
+    },
+    {
+      "epoch": 0.0914396887159533,
+      "grad_norm": 0.5846467614173889,
+      "learning_rate": 2.8199999999999998e-05,
+      "loss": 2.6104,
+      "step": 282
+    },
+    {
+      "epoch": 0.0917639429312581,
+      "grad_norm": 0.6840003132820129,
+      "learning_rate": 2.83e-05,
+      "loss": 2.9327,
+      "step": 283
+    },
+    {
+      "epoch": 0.09208819714656291,
+      "grad_norm": 0.6815200448036194,
+      "learning_rate": 2.84e-05,
+      "loss": 3.1721,
+      "step": 284
+    },
+    {
+      "epoch": 0.09241245136186771,
+      "grad_norm": 0.6062876582145691,
+      "learning_rate": 2.8499999999999998e-05,
+      "loss": 2.6518,
+      "step": 285
+    },
+    {
+      "epoch": 0.0927367055771725,
+      "grad_norm": 0.6957656741142273,
+      "learning_rate": 2.86e-05,
+      "loss": 2.8813,
+      "step": 286
+    },
+    {
+      "epoch": 0.0930609597924773,
+      "grad_norm": 0.5108417868614197,
+      "learning_rate": 2.87e-05,
+      "loss": 2.3557,
+      "step": 287
+    },
+    {
+      "epoch": 0.0933852140077821,
+      "grad_norm": 0.6118603944778442,
+      "learning_rate": 2.88e-05,
+      "loss": 2.5742,
+      "step": 288
+    },
+    {
+      "epoch": 0.0937094682230869,
+      "grad_norm": 0.5081058740615845,
+      "learning_rate": 2.89e-05,
+      "loss": 2.629,
+      "step": 289
+    },
+    {
+      "epoch": 0.0940337224383917,
+      "grad_norm": 1.0502606630325317,
+      "learning_rate": 2.9e-05,
+      "loss": 2.357,
+      "step": 290
+    },
+    {
+      "epoch": 0.0943579766536965,
+      "grad_norm": 0.4737757444381714,
+      "learning_rate": 2.91e-05,
+      "loss": 2.5533,
+      "step": 291
+    },
+    {
+      "epoch": 0.0946822308690013,
+      "grad_norm": 0.7734299302101135,
+      "learning_rate": 2.92e-05,
+      "loss": 3.1602,
+      "step": 292
+    },
+    {
+      "epoch": 0.09500648508430609,
+      "grad_norm": 0.7031993865966797,
+      "learning_rate": 2.93e-05,
+      "loss": 2.4417,
+      "step": 293
+    },
+    {
+      "epoch": 0.09533073929961089,
+      "grad_norm": 0.9395150542259216,
+      "learning_rate": 2.94e-05,
+      "loss": 3.4199,
+      "step": 294
+    },
+    {
+      "epoch": 0.09565499351491569,
+      "grad_norm": 0.40819284319877625,
+      "learning_rate": 2.95e-05,
+      "loss": 2.3519,
+      "step": 295
+    },
+    {
+      "epoch": 0.0959792477302205,
+      "grad_norm": 0.8926293253898621,
+      "learning_rate": 2.96e-05,
+      "loss": 3.2821,
+      "step": 296
+    },
+    {
+      "epoch": 0.0963035019455253,
+      "grad_norm": 3.4780051708221436,
+      "learning_rate": 2.97e-05,
+      "loss": 3.2425,
+      "step": 297
+    },
+    {
+      "epoch": 0.0966277561608301,
+      "grad_norm": 0.8986146450042725,
+      "learning_rate": 2.98e-05,
+      "loss": 3.0981,
+      "step": 298
+    },
+    {
+      "epoch": 0.09695201037613489,
+      "grad_norm": 0.5687377452850342,
+      "learning_rate": 2.9900000000000002e-05,
+      "loss": 2.4905,
+      "step": 299
+    },
+    {
+      "epoch": 0.09727626459143969,
+      "grad_norm": 0.5050517916679382,
+      "learning_rate": 3e-05,
+      "loss": 2.2259,
+      "step": 300
+    },
+    {
+      "epoch": 0.09760051880674449,
+      "grad_norm": 1.3015865087509155,
+      "learning_rate": 3e-05,
+      "loss": 3.6987,
+      "step": 301
+    },
+    {
+      "epoch": 0.09792477302204929,
+      "grad_norm": 0.995035707950592,
+      "learning_rate": 3e-05,
+      "loss": 3.6952,
+      "step": 302
+    },
+    {
+      "epoch": 0.09824902723735408,
+      "grad_norm": 0.6069695949554443,
+      "learning_rate": 3e-05,
+      "loss": 2.3805,
+      "step": 303
+    },
+    {
+      "epoch": 0.09857328145265888,
+      "grad_norm": 0.9736778736114502,
+      "learning_rate": 3e-05,
+      "loss": 2.6599,
+      "step": 304
+    },
+    {
+      "epoch": 0.09889753566796368,
+      "grad_norm": 0.7570983171463013,
+      "learning_rate": 3e-05,
+      "loss": 2.7222,
+      "step": 305
+    },
+    {
+      "epoch": 0.09922178988326848,
+      "grad_norm": 0.7066447138786316,
+      "learning_rate": 3e-05,
+      "loss": 2.593,
+      "step": 306
+    },
+    {
+      "epoch": 0.09954604409857327,
+      "grad_norm": 0.6455000638961792,
+      "learning_rate": 3e-05,
+      "loss": 2.5064,
+      "step": 307
+    },
+    {
+      "epoch": 0.09987029831387809,
+      "grad_norm": 0.9336251020431519,
+      "learning_rate": 3e-05,
+      "loss": 2.9401,
+      "step": 308
+    },
+    {
+      "epoch": 0.10019455252918288,
+      "grad_norm": 1.562284231185913,
+      "learning_rate": 3e-05,
+      "loss": 3.6856,
+      "step": 309
+    },
+    {
+      "epoch": 0.10051880674448768,
+      "grad_norm": 0.7031996250152588,
+      "learning_rate": 3e-05,
+      "loss": 3.1951,
+      "step": 310
+    },
+    {
+      "epoch": 0.10084306095979248,
+      "grad_norm": 0.5966290235519409,
+      "learning_rate": 3e-05,
+      "loss": 2.4069,
+      "step": 311
+    },
+    {
+      "epoch": 0.10116731517509728,
+      "grad_norm": 1.041093111038208,
+      "learning_rate": 3e-05,
+      "loss": 2.8238,
+      "step": 312
+    },
+    {
+      "epoch": 0.10149156939040208,
+      "grad_norm": 0.9774916172027588,
+      "learning_rate": 3e-05,
+      "loss": 3.3699,
+      "step": 313
+    },
+    {
+      "epoch": 0.10181582360570687,
+      "grad_norm": 0.7307340502738953,
+      "learning_rate": 3e-05,
+      "loss": 2.75,
+      "step": 314
+    },
+    {
+      "epoch": 0.10214007782101167,
+      "grad_norm": 1.1388113498687744,
+      "learning_rate": 3e-05,
+      "loss": 3.2411,
+      "step": 315
+    },
+    {
+      "epoch": 0.10246433203631647,
+      "grad_norm": 0.5837814211845398,
+      "learning_rate": 3e-05,
+      "loss": 2.7209,
+      "step": 316
+    },
+    {
+      "epoch": 0.10278858625162127,
+      "grad_norm": 0.5337172150611877,
+      "learning_rate": 3e-05,
+      "loss": 2.6061,
+      "step": 317
+    },
+    {
+      "epoch": 0.10311284046692606,
+      "grad_norm": 0.4444713890552521,
+      "learning_rate": 3e-05,
+      "loss": 2.4059,
+      "step": 318
+    },
+    {
+      "epoch": 0.10343709468223086,
+      "grad_norm": 0.8076606392860413,
+      "learning_rate": 3e-05,
+      "loss": 3.2599,
+      "step": 319
+    },
+    {
+      "epoch": 0.10376134889753567,
+      "grad_norm": 0.8229354619979858,
+      "learning_rate": 3e-05,
+      "loss": 2.9509,
+      "step": 320
+    },
+    {
+      "epoch": 0.10408560311284047,
+      "grad_norm": 0.40909186005592346,
+      "learning_rate": 3e-05,
+      "loss": 2.4976,
+      "step": 321
+    },
+    {
+      "epoch": 0.10440985732814527,
+      "grad_norm": 0.7027782201766968,
+      "learning_rate": 3e-05,
+      "loss": 3.2118,
+      "step": 322
+    },
+    {
+      "epoch": 0.10473411154345007,
+      "grad_norm": 1.305457353591919,
+      "learning_rate": 3e-05,
+      "loss": 3.1277,
+      "step": 323
+    },
+    {
+      "epoch": 0.10505836575875487,
+      "grad_norm": 0.41144242882728577,
+      "learning_rate": 3e-05,
+      "loss": 2.1826,
+      "step": 324
+    },
+    {
+      "epoch": 0.10538261997405966,
+      "grad_norm": 0.9614049792289734,
+      "learning_rate": 3e-05,
+      "loss": 3.1195,
+      "step": 325
+    },
+    {
+      "epoch": 0.10570687418936446,
+      "grad_norm": 0.7106190919876099,
+      "learning_rate": 3e-05,
+      "loss": 3.0072,
+      "step": 326
+    },
+    {
+      "epoch": 0.10603112840466926,
+      "grad_norm": 0.9470122456550598,
+      "learning_rate": 3e-05,
+      "loss": 2.4988,
+      "step": 327
+    },
+    {
+      "epoch": 0.10635538261997406,
+      "grad_norm": 0.7801775336265564,
+      "learning_rate": 3e-05,
+      "loss": 3.6561,
+      "step": 328
+    },
+    {
+      "epoch": 0.10667963683527885,
+      "grad_norm": 0.5154274702072144,
+      "learning_rate": 3e-05,
+      "loss": 2.6729,
+      "step": 329
+    },
+    {
+      "epoch": 0.10700389105058365,
+      "grad_norm": 0.8147680759429932,
+      "learning_rate": 3e-05,
+      "loss": 3.1715,
+      "step": 330
+    },
+    {
+      "epoch": 0.10732814526588845,
+      "grad_norm": 0.9540537595748901,
+      "learning_rate": 3e-05,
+      "loss": 3.3563,
+      "step": 331
+    },
+    {
+      "epoch": 0.10765239948119326,
+      "grad_norm": 0.8451594710350037,
+      "learning_rate": 3e-05,
+      "loss": 3.0807,
+      "step": 332
+    },
+    {
+      "epoch": 0.10797665369649806,
+      "grad_norm": 0.46150726079940796,
+      "learning_rate": 3e-05,
+      "loss": 2.2942,
+      "step": 333
+    },
+    {
+      "epoch": 0.10830090791180286,
+      "grad_norm": 0.7258126735687256,
+      "learning_rate": 3e-05,
+      "loss": 3.426,
+      "step": 334
+    },
+    {
+      "epoch": 0.10862516212710766,
+      "grad_norm": 0.7811896204948425,
+      "learning_rate": 3e-05,
+      "loss": 2.9796,
+      "step": 335
+    },
+    {
+      "epoch": 0.10894941634241245,
+      "grad_norm": 0.5422194004058838,
+      "learning_rate": 3e-05,
+      "loss": 2.6152,
+      "step": 336
+    },
+    {
+      "epoch": 0.10927367055771725,
+      "grad_norm": 0.7674731612205505,
+      "learning_rate": 3e-05,
+      "loss": 3.3383,
+      "step": 337
+    },
+    {
+      "epoch": 0.10959792477302205,
+      "grad_norm": 1.2686983346939087,
+      "learning_rate": 3e-05,
+      "loss": 3.4422,
+      "step": 338
+    },
+    {
+      "epoch": 0.10992217898832685,
+      "grad_norm": 1.2715778350830078,
+      "learning_rate": 3e-05,
+      "loss": 3.5521,
+      "step": 339
+    },
+    {
+      "epoch": 0.11024643320363164,
+      "grad_norm": 0.6377490162849426,
+      "learning_rate": 3e-05,
+      "loss": 2.3667,
+      "step": 340
+    },
+    {
+      "epoch": 0.11057068741893644,
+      "grad_norm": 1.242336630821228,
+      "learning_rate": 3e-05,
+      "loss": 2.764,
+      "step": 341
+    },
+    {
+      "epoch": 0.11089494163424124,
+      "grad_norm": 0.9681991934776306,
+      "learning_rate": 3e-05,
+      "loss": 2.8236,
+      "step": 342
+    },
+    {
+      "epoch": 0.11121919584954604,
+      "grad_norm": 0.6977137327194214,
+      "learning_rate": 3e-05,
+      "loss": 2.8363,
+      "step": 343
+    },
+    {
+      "epoch": 0.11154345006485085,
+      "grad_norm": 0.7238425016403198,
+      "learning_rate": 3e-05,
+      "loss": 2.6898,
+      "step": 344
+    },
+    {
+      "epoch": 0.11186770428015565,
+      "grad_norm": 0.663567841053009,
+      "learning_rate": 3e-05,
+      "loss": 2.4257,
+      "step": 345
+    },
+    {
+      "epoch": 0.11219195849546044,
+      "grad_norm": 1.434211254119873,
+      "learning_rate": 3e-05,
+      "loss": 3.6043,
+      "step": 346
+    },
+    {
+      "epoch": 0.11251621271076524,
+      "grad_norm": 0.36684730648994446,
+      "learning_rate": 3e-05,
+      "loss": 2.1675,
+      "step": 347
+    },
+    {
+      "epoch": 0.11284046692607004,
+      "grad_norm": 0.4982234835624695,
+      "learning_rate": 3e-05,
+      "loss": 2.2719,
+      "step": 348
+    },
+    {
+      "epoch": 0.11316472114137484,
+      "grad_norm": 0.8137962818145752,
+      "learning_rate": 3e-05,
+      "loss": 2.9747,
+      "step": 349
+    },
+    {
+      "epoch": 0.11348897535667964,
+      "grad_norm": 0.5390188097953796,
+      "learning_rate": 3e-05,
+      "loss": 2.2923,
+      "step": 350
+    },
+    {
+      "epoch": 0.11381322957198443,
+      "grad_norm": 0.4610343277454376,
+      "learning_rate": 3e-05,
+      "loss": 2.2999,
+      "step": 351
+    },
+    {
+      "epoch": 0.11413748378728923,
+      "grad_norm": 0.8101584911346436,
+      "learning_rate": 3e-05,
+      "loss": 3.1291,
+      "step": 352
+    },
+    {
+      "epoch": 0.11446173800259403,
+      "grad_norm": 0.42324066162109375,
+      "learning_rate": 3e-05,
+      "loss": 2.4585,
+      "step": 353
+    },
+    {
+      "epoch": 0.11478599221789883,
+      "grad_norm": 0.507593035697937,
+      "learning_rate": 3e-05,
+      "loss": 2.5137,
+      "step": 354
+    },
+    {
+      "epoch": 0.11511024643320363,
+      "grad_norm": 0.6725983023643494,
+      "learning_rate": 3e-05,
+      "loss": 3.1738,
+      "step": 355
+    },
+    {
+      "epoch": 0.11543450064850844,
+      "grad_norm": 0.72255939245224,
+      "learning_rate": 3e-05,
+      "loss": 3.099,
+      "step": 356
+    },
+    {
+      "epoch": 0.11575875486381323,
+      "grad_norm": 0.5508390069007874,
+      "learning_rate": 3e-05,
+      "loss": 2.7165,
+      "step": 357
+    },
+    {
+      "epoch": 0.11608300907911803,
+      "grad_norm": 0.8376918435096741,
+      "learning_rate": 3e-05,
+      "loss": 3.1042,
+      "step": 358
+    },
+    {
+      "epoch": 0.11640726329442283,
+      "grad_norm": 0.627676248550415,
+      "learning_rate": 3e-05,
+      "loss": 2.7748,
+      "step": 359
+    },
+    {
+      "epoch": 0.11673151750972763,
+      "grad_norm": 0.9748745560646057,
+      "learning_rate": 3e-05,
+      "loss": 3.4765,
+      "step": 360
+    },
+    {
+      "epoch": 0.11705577172503243,
+      "grad_norm": 0.6281887888908386,
+      "learning_rate": 3e-05,
+      "loss": 3.0335,
+      "step": 361
+    },
+    {
+      "epoch": 0.11738002594033722,
+      "grad_norm": 0.844294011592865,
+      "learning_rate": 3e-05,
+      "loss": 3.6499,
+      "step": 362
+    },
+    {
+      "epoch": 0.11770428015564202,
+      "grad_norm": 0.7436018586158752,
+      "learning_rate": 3e-05,
+      "loss": 2.5716,
+      "step": 363
+    },
+    {
+      "epoch": 0.11802853437094682,
+      "grad_norm": 1.182637095451355,
+      "learning_rate": 3e-05,
+      "loss": 2.7027,
+      "step": 364
+    },
+    {
+      "epoch": 0.11835278858625162,
+      "grad_norm": 0.8204910159111023,
+      "learning_rate": 3e-05,
+      "loss": 3.1918,
+      "step": 365
+    },
+    {
+      "epoch": 0.11867704280155641,
+      "grad_norm": 0.6965277791023254,
+      "learning_rate": 3e-05,
+      "loss": 2.8156,
+      "step": 366
+    },
+    {
+      "epoch": 0.11900129701686121,
+      "grad_norm": 0.7637869715690613,
+      "learning_rate": 3e-05,
+      "loss": 2.9692,
+      "step": 367
+    },
+    {
+      "epoch": 0.11932555123216602,
+      "grad_norm": 0.5555540919303894,
+      "learning_rate": 3e-05,
+      "loss": 2.2812,
+      "step": 368
+    },
+    {
+      "epoch": 0.11964980544747082,
+      "grad_norm": 1.362791657447815,
+      "learning_rate": 3e-05,
+      "loss": 3.5497,
+      "step": 369
+    },
+    {
+      "epoch": 0.11997405966277562,
+      "grad_norm": 0.39166656136512756,
+      "learning_rate": 3e-05,
+      "loss": 2.2746,
+      "step": 370
+    },
+    {
+      "epoch": 0.12029831387808042,
+      "grad_norm": 0.8686426281929016,
+      "learning_rate": 3e-05,
+      "loss": 3.0663,
+      "step": 371
+    },
+    {
+      "epoch": 0.12062256809338522,
+      "grad_norm": 0.9395362734794617,
+      "learning_rate": 3e-05,
+      "loss": 3.905,
+      "step": 372
+    },
+    {
+      "epoch": 0.12094682230869001,
+      "grad_norm": 1.1586980819702148,
+      "learning_rate": 3e-05,
+      "loss": 3.7696,
+      "step": 373
+    },
+    {
+      "epoch": 0.12127107652399481,
+      "grad_norm": 0.6585375070571899,
+      "learning_rate": 3e-05,
+      "loss": 3.0685,
+      "step": 374
+    },
+    {
+      "epoch": 0.12159533073929961,
+      "grad_norm": 0.5126864910125732,
+      "learning_rate": 3e-05,
+      "loss": 2.718,
+      "step": 375
+    },
+    {
+      "epoch": 0.1219195849546044,
+      "grad_norm": 0.7193174362182617,
+      "learning_rate": 3e-05,
+      "loss": 2.7747,
+      "step": 376
+    },
+    {
+      "epoch": 0.1222438391699092,
+      "grad_norm": 1.0809192657470703,
+      "learning_rate": 3e-05,
+      "loss": 3.4805,
+      "step": 377
+    },
+    {
+      "epoch": 0.122568093385214,
+      "grad_norm": 0.4331657290458679,
+      "learning_rate": 3e-05,
+      "loss": 2.7252,
+      "step": 378
+    },
+    {
+      "epoch": 0.1228923476005188,
+      "grad_norm": 0.7881174683570862,
+      "learning_rate": 3e-05,
+      "loss": 2.8244,
+      "step": 379
+    },
+    {
+      "epoch": 0.12321660181582361,
+      "grad_norm": 0.7106895446777344,
+      "learning_rate": 3e-05,
+      "loss": 3.2227,
+      "step": 380
+    },
+    {
+      "epoch": 0.12354085603112841,
+      "grad_norm": 0.7195374965667725,
+      "learning_rate": 3e-05,
+      "loss": 3.005,
+      "step": 381
+    },
+    {
+      "epoch": 0.12386511024643321,
+      "grad_norm": 0.5068550705909729,
+      "learning_rate": 3e-05,
+      "loss": 2.3562,
+      "step": 382
+    },
+    {
+      "epoch": 0.124189364461738,
+      "grad_norm": 0.4919268786907196,
+      "learning_rate": 3e-05,
+      "loss": 2.3283,
+      "step": 383
+    },
+    {
+      "epoch": 0.1245136186770428,
+      "grad_norm": 0.8128411769866943,
+      "learning_rate": 3e-05,
+      "loss": 2.9122,
+      "step": 384
+    },
+    {
+      "epoch": 0.1248378728923476,
+      "grad_norm": 0.8163199424743652,
+      "learning_rate": 3e-05,
+      "loss": 3.0264,
+      "step": 385
+    },
+    {
+      "epoch": 0.1251621271076524,
+      "grad_norm": 1.0967957973480225,
+      "learning_rate": 3e-05,
+      "loss": 2.5559,
+      "step": 386
+    },
+    {
+      "epoch": 0.1254863813229572,
+      "grad_norm": 0.7630811333656311,
+      "learning_rate": 3e-05,
+      "loss": 2.6751,
+      "step": 387
+    },
+    {
+      "epoch": 0.125810635538262,
+      "grad_norm": 0.6830684542655945,
+      "learning_rate": 3e-05,
+      "loss": 2.9549,
+      "step": 388
+    },
+    {
+      "epoch": 0.1261348897535668,
+      "grad_norm": 0.5710088014602661,
+      "learning_rate": 3e-05,
+      "loss": 2.8603,
+      "step": 389
+    },
+    {
+      "epoch": 0.1264591439688716,
+      "grad_norm": 0.631334125995636,
+      "learning_rate": 3e-05,
+      "loss": 2.7823,
+      "step": 390
+    },
+    {
+      "epoch": 0.1267833981841764,
+      "grad_norm": 0.5275784730911255,
+      "learning_rate": 3e-05,
+      "loss": 2.5567,
+      "step": 391
+    },
+    {
+      "epoch": 0.12710765239948119,
+      "grad_norm": 0.7490580677986145,
+      "learning_rate": 3e-05,
+      "loss": 3.0601,
+      "step": 392
+    },
+    {
+      "epoch": 0.12743190661478598,
+      "grad_norm": 0.6533567905426025,
+      "learning_rate": 3e-05,
+      "loss": 2.8383,
+      "step": 393
+    },
+    {
+      "epoch": 0.12775616083009078,
+      "grad_norm": 0.390969455242157,
+      "learning_rate": 3e-05,
+      "loss": 2.2685,
+      "step": 394
+    },
+    {
+      "epoch": 0.12808041504539558,
+      "grad_norm": 0.6746393442153931,
+      "learning_rate": 3e-05,
+      "loss": 2.6476,
+      "step": 395
+    },
+    {
+      "epoch": 0.12840466926070038,
+      "grad_norm": 0.6837244629859924,
+      "learning_rate": 3e-05,
+      "loss": 3.2313,
+      "step": 396
+    },
+    {
+      "epoch": 0.12872892347600517,
+      "grad_norm": 0.3665929436683655,
+      "learning_rate": 3e-05,
+      "loss": 2.1611,
+      "step": 397
+    },
+    {
+      "epoch": 0.12905317769131,
+      "grad_norm": 0.673858642578125,
+      "learning_rate": 3e-05,
+      "loss": 3.1608,
+      "step": 398
+    },
+    {
+      "epoch": 0.1293774319066148,
+      "grad_norm": 0.7747320532798767,
+      "learning_rate": 3e-05,
+      "loss": 2.8139,
+      "step": 399
+    },
+    {
+      "epoch": 0.1297016861219196,
+      "grad_norm": 0.46531644463539124,
+      "learning_rate": 3e-05,
+      "loss": 2.5512,
+      "step": 400
+    },
+    {
+      "epoch": 0.1300259403372244,
+      "grad_norm": 0.5171082019805908,
+      "learning_rate": 3e-05,
+      "loss": 3.1556,
+      "step": 401
+    },
+    {
+      "epoch": 0.1303501945525292,
+      "grad_norm": 0.6828058362007141,
+      "learning_rate": 3e-05,
+      "loss": 3.1344,
+      "step": 402
+    },
+    {
+      "epoch": 0.130674448767834,
+      "grad_norm": 0.8618952631950378,
+      "learning_rate": 3e-05,
+      "loss": 3.3386,
+      "step": 403
+    },
+    {
+      "epoch": 0.1309987029831388,
+      "grad_norm": 0.6293363571166992,
+      "learning_rate": 3e-05,
+      "loss": 2.6808,
+      "step": 404
+    },
+    {
+      "epoch": 0.13132295719844359,
+      "grad_norm": 0.5605633854866028,
+      "learning_rate": 3e-05,
+      "loss": 2.98,
+      "step": 405
+    },
+    {
+      "epoch": 0.13164721141374838,
+      "grad_norm": 0.8432521224021912,
+      "learning_rate": 3e-05,
+      "loss": 2.5335,
+      "step": 406
+    },
+    {
+      "epoch": 0.13197146562905318,
+      "grad_norm": 0.8483084440231323,
+      "learning_rate": 3e-05,
+      "loss": 3.3009,
+      "step": 407
+    },
+    {
+      "epoch": 0.13229571984435798,
+      "grad_norm": 0.45350831747055054,
+      "learning_rate": 3e-05,
+      "loss": 2.3413,
+      "step": 408
+    },
+    {
+      "epoch": 0.13261997405966278,
+      "grad_norm": 0.917023241519928,
+      "learning_rate": 3e-05,
+      "loss": 3.5875,
+      "step": 409
+    },
+    {
+      "epoch": 0.13294422827496757,
+      "grad_norm": 0.6137032508850098,
+      "learning_rate": 3e-05,
+      "loss": 2.495,
+      "step": 410
+    },
+    {
+      "epoch": 0.13326848249027237,
+      "grad_norm": 0.821355938911438,
+      "learning_rate": 3e-05,
+      "loss": 2.8902,
+      "step": 411
+    },
+    {
+      "epoch": 0.13359273670557717,
+      "grad_norm": 0.4815053343772888,
+      "learning_rate": 3e-05,
+      "loss": 2.289,
+      "step": 412
+    },
+    {
+      "epoch": 0.13391699092088197,
+      "grad_norm": 0.5894195437431335,
+      "learning_rate": 3e-05,
+      "loss": 2.468,
+      "step": 413
+    },
+    {
+      "epoch": 0.13424124513618677,
+      "grad_norm": 0.5890939235687256,
+      "learning_rate": 3e-05,
+      "loss": 2.5555,
+      "step": 414
+    },
+    {
+      "epoch": 0.13456549935149156,
+      "grad_norm": 0.5914024114608765,
+      "learning_rate": 3e-05,
+      "loss": 2.7318,
+      "step": 415
+    },
+    {
+      "epoch": 0.13488975356679636,
+      "grad_norm": 0.376447468996048,
+      "learning_rate": 3e-05,
+      "loss": 2.2713,
+      "step": 416
+    },
+    {
+      "epoch": 0.13521400778210116,
+      "grad_norm": 0.6079317927360535,
+      "learning_rate": 3e-05,
+      "loss": 2.9882,
+      "step": 417
+    },
+    {
+      "epoch": 0.13553826199740596,
+      "grad_norm": 0.3256724774837494,
+      "learning_rate": 3e-05,
+      "loss": 2.0586,
+      "step": 418
+    },
+    {
+      "epoch": 0.13586251621271075,
+      "grad_norm": 0.4198353588581085,
+      "learning_rate": 3e-05,
+      "loss": 2.4441,
+      "step": 419
+    },
+    {
+      "epoch": 0.13618677042801555,
+      "grad_norm": 0.8752949237823486,
+      "learning_rate": 3e-05,
+      "loss": 3.4775,
+      "step": 420
+    },
+    {
+      "epoch": 0.13651102464332035,
+      "grad_norm": 0.6420135498046875,
+      "learning_rate": 3e-05,
+      "loss": 3.0084,
+      "step": 421
+    },
+    {
+      "epoch": 0.13683527885862518,
+      "grad_norm": 0.3771823048591614,
+      "learning_rate": 3e-05,
+      "loss": 1.9751,
+      "step": 422
+    },
+    {
+      "epoch": 0.13715953307392997,
+      "grad_norm": 0.9496719837188721,
+      "learning_rate": 3e-05,
+      "loss": 2.4832,
+      "step": 423
+    },
+    {
+      "epoch": 0.13748378728923477,
+      "grad_norm": 1.209071397781372,
+      "learning_rate": 3e-05,
+      "loss": 2.7216,
+      "step": 424
+    },
+    {
+      "epoch": 0.13780804150453957,
+      "grad_norm": 0.6273159384727478,
+      "learning_rate": 3e-05,
+      "loss": 3.1355,
+      "step": 425
+    },
+    {
+      "epoch": 0.13813229571984437,
+      "grad_norm": 0.9874957203865051,
+      "learning_rate": 3e-05,
+      "loss": 3.849,
+      "step": 426
+    },
+    {
+      "epoch": 0.13845654993514916,
+      "grad_norm": 0.6410277485847473,
+      "learning_rate": 3e-05,
+      "loss": 2.8417,
+      "step": 427
+    },
+    {
+      "epoch": 0.13878080415045396,
+      "grad_norm": 0.5870115160942078,
+      "learning_rate": 3e-05,
+      "loss": 2.905,
+      "step": 428
+    },
+    {
+      "epoch": 0.13910505836575876,
+      "grad_norm": 0.556688666343689,
+      "learning_rate": 3e-05,
+      "loss": 2.656,
+      "step": 429
+    },
+    {
+      "epoch": 0.13942931258106356,
+      "grad_norm": 0.4426051378250122,
+      "learning_rate": 3e-05,
+      "loss": 2.5187,
+      "step": 430
+    },
+    {
+      "epoch": 0.13975356679636836,
+      "grad_norm": 0.6465773582458496,
+      "learning_rate": 3e-05,
+      "loss": 3.0021,
+      "step": 431
+    },
+    {
+      "epoch": 0.14007782101167315,
+      "grad_norm": 0.6202263236045837,
+      "learning_rate": 3e-05,
+      "loss": 2.5958,
+      "step": 432
+    },
+    {
+      "epoch": 0.14040207522697795,
+      "grad_norm": 0.489071786403656,
+      "learning_rate": 3e-05,
+      "loss": 2.837,
+      "step": 433
+    },
+    {
+      "epoch": 0.14072632944228275,
+      "grad_norm": 0.42626988887786865,
+      "learning_rate": 3e-05,
+      "loss": 2.4393,
+      "step": 434
+    },
+    {
+      "epoch": 0.14105058365758755,
+      "grad_norm": 0.8826153874397278,
+      "learning_rate": 3e-05,
+      "loss": 3.7325,
+      "step": 435
+    },
+    {
+      "epoch": 0.14137483787289234,
+      "grad_norm": 0.5003890991210938,
+      "learning_rate": 3e-05,
+      "loss": 2.4484,
+      "step": 436
+    },
+    {
+      "epoch": 0.14169909208819714,
+      "grad_norm": 0.43842563033103943,
+      "learning_rate": 3e-05,
+      "loss": 2.5548,
+      "step": 437
+    },
+    {
+      "epoch": 0.14202334630350194,
+      "grad_norm": 0.7168633937835693,
+      "learning_rate": 3e-05,
+      "loss": 2.6935,
+      "step": 438
+    },
+    {
+      "epoch": 0.14234760051880674,
+      "grad_norm": 0.5617209672927856,
+      "learning_rate": 3e-05,
+      "loss": 3.2058,
+      "step": 439
+    },
+    {
+      "epoch": 0.14267185473411154,
+      "grad_norm": 0.6762505173683167,
+      "learning_rate": 3e-05,
+      "loss": 3.291,
+      "step": 440
+    },
+    {
+      "epoch": 0.14299610894941633,
+      "grad_norm": 0.37547358870506287,
+      "learning_rate": 3e-05,
+      "loss": 2.4053,
+      "step": 441
+    },
+    {
+      "epoch": 0.14332036316472113,
+      "grad_norm": 0.4991322159767151,
+      "learning_rate": 3e-05,
+      "loss": 2.9337,
+      "step": 442
+    },
+    {
+      "epoch": 0.14364461738002593,
+      "grad_norm": 0.5902826189994812,
+      "learning_rate": 3e-05,
+      "loss": 2.652,
+      "step": 443
+    },
+    {
+      "epoch": 0.14396887159533073,
+      "grad_norm": 0.5595560669898987,
+      "learning_rate": 3e-05,
+      "loss": 2.9651,
+      "step": 444
+    },
+    {
+      "epoch": 0.14429312581063553,
+      "grad_norm": 0.44947323203086853,
+      "learning_rate": 3e-05,
+      "loss": 2.5268,
+      "step": 445
+    },
+    {
+      "epoch": 0.14461738002594035,
+      "grad_norm": 1.01261568069458,
+      "learning_rate": 3e-05,
+      "loss": 3.5606,
+      "step": 446
+    },
+    {
+      "epoch": 0.14494163424124515,
+      "grad_norm": 1.060111165046692,
+      "learning_rate": 3e-05,
+      "loss": 3.958,
+      "step": 447
+    },
+    {
+      "epoch": 0.14526588845654995,
+      "grad_norm": 0.6862601637840271,
+      "learning_rate": 3e-05,
+      "loss": 2.3801,
+      "step": 448
+    },
+    {
+      "epoch": 0.14559014267185474,
+      "grad_norm": 0.5765560865402222,
+      "learning_rate": 3e-05,
+      "loss": 2.8744,
+      "step": 449
+    },
+    {
+      "epoch": 0.14591439688715954,
+      "grad_norm": 0.6669028997421265,
+      "learning_rate": 3e-05,
+      "loss": 2.5161,
+      "step": 450
+    },
+    {
+      "epoch": 0.14623865110246434,
+      "grad_norm": 0.40496715903282166,
+      "learning_rate": 3e-05,
+      "loss": 2.3413,
+      "step": 451
+    },
+    {
+      "epoch": 0.14656290531776914,
+      "grad_norm": 0.6844581961631775,
+      "learning_rate": 3e-05,
+      "loss": 3.0118,
+      "step": 452
+    },
+    {
+      "epoch": 0.14688715953307394,
+      "grad_norm": 0.6622430086135864,
+      "learning_rate": 3e-05,
+      "loss": 2.9144,
+      "step": 453
+    },
+    {
+      "epoch": 0.14721141374837873,
+      "grad_norm": 0.7101048827171326,
+      "learning_rate": 3e-05,
+      "loss": 2.9111,
+      "step": 454
+    },
+    {
+      "epoch": 0.14753566796368353,
+      "grad_norm": 0.6050754189491272,
+      "learning_rate": 3e-05,
+      "loss": 2.4829,
+      "step": 455
+    },
+    {
+      "epoch": 0.14785992217898833,
+      "grad_norm": 0.4387664794921875,
+      "learning_rate": 3e-05,
+      "loss": 2.5053,
+      "step": 456
+    },
+    {
+      "epoch": 0.14818417639429313,
+      "grad_norm": 0.8280428647994995,
+      "learning_rate": 3e-05,
+      "loss": 3.0798,
+      "step": 457
+    },
+    {
+      "epoch": 0.14850843060959792,
+      "grad_norm": 0.5333096385002136,
+      "learning_rate": 3e-05,
+      "loss": 2.5485,
+      "step": 458
+    },
+    {
+      "epoch": 0.14883268482490272,
+      "grad_norm": 0.39863231778144836,
+      "learning_rate": 3e-05,
+      "loss": 2.2686,
+      "step": 459
+    },
+    {
+      "epoch": 0.14915693904020752,
+      "grad_norm": 0.6192648410797119,
+      "learning_rate": 3e-05,
+      "loss": 2.6228,
+      "step": 460
+    },
+    {
+      "epoch": 0.14948119325551232,
+      "grad_norm": 0.448277086019516,
+      "learning_rate": 3e-05,
+      "loss": 2.3449,
+      "step": 461
+    },
+    {
+      "epoch": 0.14980544747081712,
+      "grad_norm": 0.6989661455154419,
+      "learning_rate": 3e-05,
+      "loss": 3.2166,
+      "step": 462
+    },
+    {
+      "epoch": 0.1501297016861219,
+      "grad_norm": 0.5062859058380127,
+      "learning_rate": 3e-05,
+      "loss": 2.5943,
+      "step": 463
+    },
+    {
+      "epoch": 0.1504539559014267,
+      "grad_norm": 0.6428647041320801,
+      "learning_rate": 3e-05,
+      "loss": 3.1886,
+      "step": 464
+    },
+    {
+      "epoch": 0.1507782101167315,
+      "grad_norm": 0.41773104667663574,
+      "learning_rate": 3e-05,
+      "loss": 2.6191,
+      "step": 465
+    },
+    {
+      "epoch": 0.1511024643320363,
+      "grad_norm": 0.5408780574798584,
+      "learning_rate": 3e-05,
+      "loss": 2.7527,
+      "step": 466
+    },
+    {
+      "epoch": 0.1514267185473411,
+      "grad_norm": 0.6191630363464355,
+      "learning_rate": 3e-05,
+      "loss": 3.1536,
+      "step": 467
+    },
+    {
+      "epoch": 0.1517509727626459,
+      "grad_norm": 0.47470274567604065,
+      "learning_rate": 3e-05,
+      "loss": 2.5838,
+      "step": 468
+    },
+    {
+      "epoch": 0.1520752269779507,
+      "grad_norm": 0.5481436252593994,
+      "learning_rate": 3e-05,
+      "loss": 2.4394,
+      "step": 469
+    },
+    {
+      "epoch": 0.15239948119325553,
+      "grad_norm": 0.47394832968711853,
+      "learning_rate": 3e-05,
+      "loss": 2.7565,
+      "step": 470
+    },
+    {
+      "epoch": 0.15272373540856032,
+      "grad_norm": 0.45594075322151184,
+      "learning_rate": 3e-05,
+      "loss": 2.9315,
+      "step": 471
+    },
+    {
+      "epoch": 0.15304798962386512,
+      "grad_norm": 0.37394964694976807,
+      "learning_rate": 3e-05,
+      "loss": 2.4531,
+      "step": 472
+    },
+    {
+      "epoch": 0.15337224383916992,
+      "grad_norm": 0.8126447796821594,
+      "learning_rate": 3e-05,
+      "loss": 2.6631,
+      "step": 473
+    },
+    {
+      "epoch": 0.15369649805447472,
+      "grad_norm": 0.49204790592193604,
+      "learning_rate": 3e-05,
+      "loss": 2.7256,
+      "step": 474
+    },
+    {
+      "epoch": 0.15402075226977952,
+      "grad_norm": 0.3882718086242676,
+      "learning_rate": 3e-05,
+      "loss": 2.2979,
+      "step": 475
+    },
+    {
+      "epoch": 0.1543450064850843,
+      "grad_norm": 0.38599106669425964,
+      "learning_rate": 3e-05,
+      "loss": 2.3577,
+      "step": 476
+    },
+    {
+      "epoch": 0.1546692607003891,
+      "grad_norm": 0.6724908351898193,
+      "learning_rate": 3e-05,
+      "loss": 2.856,
+      "step": 477
+    },
+    {
+      "epoch": 0.1549935149156939,
+      "grad_norm": 0.4690714478492737,
+      "learning_rate": 3e-05,
+      "loss": 2.4168,
+      "step": 478
+    },
+    {
+      "epoch": 0.1553177691309987,
+      "grad_norm": 0.6049445271492004,
+      "learning_rate": 3e-05,
+      "loss": 2.8894,
+      "step": 479
+    },
+    {
+      "epoch": 0.1556420233463035,
+      "grad_norm": 0.4164368510246277,
+      "learning_rate": 3e-05,
+      "loss": 2.258,
+      "step": 480
+    },
+    {
+      "epoch": 0.1559662775616083,
+      "grad_norm": 1.2326778173446655,
+      "learning_rate": 3e-05,
+      "loss": 4.142,
+      "step": 481
+    },
+    {
+      "epoch": 0.1562905317769131,
+      "grad_norm": 1.1433262825012207,
+      "learning_rate": 3e-05,
+      "loss": 3.349,
+      "step": 482
+    },
+    {
+      "epoch": 0.1566147859922179,
+      "grad_norm": 0.7982839941978455,
+      "learning_rate": 3e-05,
+      "loss": 2.8084,
+      "step": 483
+    },
+    {
+      "epoch": 0.1569390402075227,
+      "grad_norm": 0.44566285610198975,
+      "learning_rate": 3e-05,
+      "loss": 2.7924,
+      "step": 484
+    },
+    {
+      "epoch": 0.1572632944228275,
+      "grad_norm": 0.5944378972053528,
+      "learning_rate": 3e-05,
+      "loss": 2.8917,
+      "step": 485
+    },
+    {
+      "epoch": 0.1575875486381323,
+      "grad_norm": 0.39894750714302063,
+      "learning_rate": 3e-05,
+      "loss": 2.0111,
+      "step": 486
+    },
+    {
+      "epoch": 0.1579118028534371,
+      "grad_norm": 0.6556453108787537,
+      "learning_rate": 3e-05,
+      "loss": 2.8785,
+      "step": 487
+    },
+    {
+      "epoch": 0.1582360570687419,
+      "grad_norm": 0.5721779465675354,
+      "learning_rate": 3e-05,
+      "loss": 2.4929,
+      "step": 488
+    },
+    {
+      "epoch": 0.15856031128404668,
+      "grad_norm": 0.4211887717247009,
+      "learning_rate": 3e-05,
+      "loss": 2.4014,
+      "step": 489
+    },
+    {
+      "epoch": 0.15888456549935148,
+      "grad_norm": 0.686249315738678,
+      "learning_rate": 3e-05,
+      "loss": 2.9079,
+      "step": 490
+    },
+    {
+      "epoch": 0.15920881971465628,
+      "grad_norm": 0.40154188871383667,
+      "learning_rate": 3e-05,
+      "loss": 2.5882,
+      "step": 491
+    },
+    {
+      "epoch": 0.15953307392996108,
+      "grad_norm": 0.6357690691947937,
+      "learning_rate": 3e-05,
+      "loss": 2.7811,
+      "step": 492
+    },
+    {
+      "epoch": 0.15985732814526588,
+      "grad_norm": 0.9853246808052063,
+      "learning_rate": 3e-05,
+      "loss": 3.3597,
+      "step": 493
+    },
+    {
+      "epoch": 0.1601815823605707,
+      "grad_norm": 0.6755144000053406,
+      "learning_rate": 3e-05,
+      "loss": 2.8055,
+      "step": 494
+    },
+    {
+      "epoch": 0.1605058365758755,
+      "grad_norm": 0.8584931492805481,
+      "learning_rate": 3e-05,
+      "loss": 2.7198,
+      "step": 495
+    },
+    {
+      "epoch": 0.1608300907911803,
+      "grad_norm": 0.504004955291748,
+      "learning_rate": 3e-05,
+      "loss": 2.6595,
+      "step": 496
+    },
+    {
+      "epoch": 0.1611543450064851,
+      "grad_norm": 0.5366716980934143,
+      "learning_rate": 3e-05,
+      "loss": 2.6483,
+      "step": 497
+    },
+    {
+      "epoch": 0.1614785992217899,
+      "grad_norm": 0.574421226978302,
+      "learning_rate": 3e-05,
+      "loss": 2.5282,
+      "step": 498
+    },
+    {
+      "epoch": 0.1618028534370947,
+      "grad_norm": 0.5026317238807678,
+      "learning_rate": 3e-05,
+      "loss": 2.4465,
+      "step": 499
+    },
+    {
+      "epoch": 0.1621271076523995,
+      "grad_norm": 0.38147929310798645,
+      "learning_rate": 3e-05,
+      "loss": 2.3433,
+      "step": 500
+    },
+    {
+      "epoch": 0.16245136186770429,
+      "grad_norm": 0.9142938256263733,
+      "learning_rate": 3e-05,
+      "loss": 3.3797,
+      "step": 501
+    },
+    {
+      "epoch": 0.16277561608300908,
+      "grad_norm": 0.5788193941116333,
+      "learning_rate": 3e-05,
+      "loss": 2.8753,
+      "step": 502
+    },
+    {
+      "epoch": 0.16309987029831388,
+      "grad_norm": 0.5824872255325317,
+      "learning_rate": 3e-05,
+      "loss": 2.6571,
+      "step": 503
+    },
+    {
+      "epoch": 0.16342412451361868,
+      "grad_norm": 0.6864482164382935,
+      "learning_rate": 3e-05,
+      "loss": 2.7588,
+      "step": 504
+    },
+    {
+      "epoch": 0.16374837872892348,
+      "grad_norm": 0.6305482387542725,
+      "learning_rate": 3e-05,
+      "loss": 2.5787,
+      "step": 505
+    },
+    {
+      "epoch": 0.16407263294422827,
+      "grad_norm": 0.6366297602653503,
+      "learning_rate": 3e-05,
+      "loss": 3.4284,
+      "step": 506
+    },
+    {
+      "epoch": 0.16439688715953307,
+      "grad_norm": 0.8215645551681519,
+      "learning_rate": 3e-05,
+      "loss": 3.557,
+      "step": 507
+    },
+    {
+      "epoch": 0.16472114137483787,
+      "grad_norm": 0.38058093190193176,
+      "learning_rate": 3e-05,
+      "loss": 2.2598,
+      "step": 508
+    },
+    {
+      "epoch": 0.16504539559014267,
+      "grad_norm": 0.45505577325820923,
+      "learning_rate": 3e-05,
+      "loss": 2.3108,
+      "step": 509
+    },
+    {
+      "epoch": 0.16536964980544747,
+      "grad_norm": 0.8403217792510986,
+      "learning_rate": 3e-05,
+      "loss": 3.4055,
+      "step": 510
+    },
+    {
+      "epoch": 0.16569390402075226,
+      "grad_norm": 0.6367074847221375,
+      "learning_rate": 3e-05,
+      "loss": 2.711,
+      "step": 511
+    },
+    {
+      "epoch": 0.16601815823605706,
+      "grad_norm": 0.7691447138786316,
+      "learning_rate": 3e-05,
+      "loss": 3.0367,
+      "step": 512
+    },
+    {
+      "epoch": 0.16634241245136186,
+      "grad_norm": 0.508536696434021,
+      "learning_rate": 3e-05,
+      "loss": 2.666,
+      "step": 513
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 0.6860784292221069,
+      "learning_rate": 3e-05,
+      "loss": 3.1029,
+      "step": 514
+    },
+    {
+      "epoch": 0.16699092088197146,
+      "grad_norm": 0.5672210454940796,
+      "learning_rate": 3e-05,
+      "loss": 2.7271,
+      "step": 515
+    },
+    {
+      "epoch": 0.16731517509727625,
+      "grad_norm": 0.44537854194641113,
+      "learning_rate": 3e-05,
+      "loss": 2.7513,
+      "step": 516
+    },
+    {
+      "epoch": 0.16763942931258105,
+      "grad_norm": 0.46655553579330444,
+      "learning_rate": 3e-05,
+      "loss": 2.7629,
+      "step": 517
+    },
+    {
+      "epoch": 0.16796368352788585,
+      "grad_norm": 0.6430413126945496,
+      "learning_rate": 3e-05,
+      "loss": 2.8935,
+      "step": 518
+    },
+    {
+      "epoch": 0.16828793774319067,
+      "grad_norm": 0.6596621870994568,
+      "learning_rate": 3e-05,
+      "loss": 3.0873,
+      "step": 519
+    },
+    {
+      "epoch": 0.16861219195849547,
+      "grad_norm": 0.4391716718673706,
+      "learning_rate": 3e-05,
+      "loss": 2.3547,
+      "step": 520
+    },
+    {
+      "epoch": 0.16893644617380027,
+      "grad_norm": 0.587926983833313,
+      "learning_rate": 3e-05,
+      "loss": 2.959,
+      "step": 521
+    },
+    {
+      "epoch": 0.16926070038910507,
+      "grad_norm": 0.5624407529830933,
+      "learning_rate": 3e-05,
+      "loss": 3.0203,
+      "step": 522
+    },
+    {
+      "epoch": 0.16958495460440987,
+      "grad_norm": 0.5354810953140259,
+      "learning_rate": 3e-05,
+      "loss": 2.9057,
+      "step": 523
+    },
+    {
+      "epoch": 0.16990920881971466,
+      "grad_norm": 0.5903101563453674,
+      "learning_rate": 3e-05,
+      "loss": 3.0467,
+      "step": 524
+    },
+    {
+      "epoch": 0.17023346303501946,
+      "grad_norm": 0.4741097390651703,
+      "learning_rate": 3e-05,
+      "loss": 2.3771,
+      "step": 525
+    },
+    {
+      "epoch": 0.17055771725032426,
+      "grad_norm": 0.4670526385307312,
+      "learning_rate": 3e-05,
+      "loss": 2.5792,
+      "step": 526
+    },
+    {
+      "epoch": 0.17088197146562906,
+      "grad_norm": 0.5363697409629822,
+      "learning_rate": 3e-05,
+      "loss": 2.7219,
+      "step": 527
+    },
+    {
+      "epoch": 0.17120622568093385,
+      "grad_norm": 0.6387732028961182,
+      "learning_rate": 3e-05,
+      "loss": 2.9619,
+      "step": 528
+    },
+    {
+      "epoch": 0.17153047989623865,
+      "grad_norm": 0.5168539881706238,
+      "learning_rate": 3e-05,
+      "loss": 2.7336,
+      "step": 529
+    },
+    {
+      "epoch": 0.17185473411154345,
+      "grad_norm": 0.5983394384384155,
+      "learning_rate": 3e-05,
+      "loss": 2.767,
+      "step": 530
+    },
+    {
+      "epoch": 0.17217898832684825,
+      "grad_norm": 0.4355388283729553,
+      "learning_rate": 3e-05,
+      "loss": 2.5355,
+      "step": 531
+    },
+    {
+      "epoch": 0.17250324254215305,
+      "grad_norm": 0.4762032628059387,
+      "learning_rate": 3e-05,
+      "loss": 2.8563,
+      "step": 532
+    },
+    {
+      "epoch": 0.17282749675745784,
+      "grad_norm": 0.6238677501678467,
+      "learning_rate": 3e-05,
+      "loss": 2.9724,
+      "step": 533
+    },
+    {
+      "epoch": 0.17315175097276264,
+      "grad_norm": 0.5993486046791077,
+      "learning_rate": 3e-05,
+      "loss": 3.181,
+      "step": 534
+    },
+    {
+      "epoch": 0.17347600518806744,
+      "grad_norm": 0.48513278365135193,
+      "learning_rate": 3e-05,
+      "loss": 2.3545,
+      "step": 535
+    },
+    {
+      "epoch": 0.17380025940337224,
+      "grad_norm": 0.38926875591278076,
+      "learning_rate": 3e-05,
+      "loss": 2.346,
+      "step": 536
+    },
+    {
+      "epoch": 0.17412451361867703,
+      "grad_norm": 0.5772921442985535,
+      "learning_rate": 3e-05,
+      "loss": 3.2224,
+      "step": 537
+    },
+    {
+      "epoch": 0.17444876783398183,
+      "grad_norm": 0.4297749400138855,
+      "learning_rate": 3e-05,
+      "loss": 2.4488,
+      "step": 538
+    },
+    {
+      "epoch": 0.17477302204928663,
+      "grad_norm": 0.6404263973236084,
+      "learning_rate": 3e-05,
+      "loss": 2.9496,
+      "step": 539
+    },
+    {
+      "epoch": 0.17509727626459143,
+      "grad_norm": 0.4602503478527069,
+      "learning_rate": 3e-05,
+      "loss": 2.5866,
+      "step": 540
+    },
+    {
+      "epoch": 0.17542153047989623,
+      "grad_norm": 0.6406030058860779,
+      "learning_rate": 3e-05,
+      "loss": 2.785,
+      "step": 541
+    },
+    {
+      "epoch": 0.17574578469520102,
+      "grad_norm": 0.7343503832817078,
+      "learning_rate": 3e-05,
+      "loss": 2.862,
+      "step": 542
+    },
+    {
+      "epoch": 0.17607003891050585,
+      "grad_norm": 1.2962440252304077,
+      "learning_rate": 3e-05,
+      "loss": 3.4819,
+      "step": 543
+    },
+    {
+      "epoch": 0.17639429312581065,
+      "grad_norm": 0.9998407959938049,
+      "learning_rate": 3e-05,
+      "loss": 3.6895,
+      "step": 544
+    },
+    {
+      "epoch": 0.17671854734111545,
+      "grad_norm": 0.5724481344223022,
+      "learning_rate": 3e-05,
+      "loss": 3.0255,
+      "step": 545
+    },
+    {
+      "epoch": 0.17704280155642024,
+      "grad_norm": 0.7746886014938354,
+      "learning_rate": 3e-05,
+      "loss": 3.1269,
+      "step": 546
+    },
+    {
+      "epoch": 0.17736705577172504,
+      "grad_norm": 0.7206681966781616,
+      "learning_rate": 3e-05,
+      "loss": 3.3506,
+      "step": 547
+    },
+    {
+      "epoch": 0.17769130998702984,
+      "grad_norm": 0.4826478064060211,
+      "learning_rate": 3e-05,
+      "loss": 2.438,
+      "step": 548
+    },
+    {
+      "epoch": 0.17801556420233464,
+      "grad_norm": 1.012331247329712,
+      "learning_rate": 3e-05,
+      "loss": 3.2215,
+      "step": 549
+    },
+    {
+      "epoch": 0.17833981841763943,
+      "grad_norm": 0.8271002173423767,
+      "learning_rate": 3e-05,
+      "loss": 3.2743,
+      "step": 550
+    },
+    {
+      "epoch": 0.17866407263294423,
+      "grad_norm": 0.6614644527435303,
+      "learning_rate": 3e-05,
+      "loss": 2.7949,
+      "step": 551
+    },
+    {
+      "epoch": 0.17898832684824903,
+      "grad_norm": 0.7599791288375854,
+      "learning_rate": 3e-05,
+      "loss": 2.9938,
+      "step": 552
+    },
+    {
+      "epoch": 0.17931258106355383,
+      "grad_norm": 0.624595046043396,
+      "learning_rate": 3e-05,
+      "loss": 2.6115,
+      "step": 553
+    },
+    {
+      "epoch": 0.17963683527885863,
+      "grad_norm": 0.4255031645298004,
+      "learning_rate": 3e-05,
+      "loss": 2.4856,
+      "step": 554
+    },
+    {
+      "epoch": 0.17996108949416342,
+      "grad_norm": 0.737805187702179,
+      "learning_rate": 3e-05,
+      "loss": 3.2104,
+      "step": 555
+    },
+    {
+      "epoch": 0.18028534370946822,
+      "grad_norm": 0.4640255272388458,
+      "learning_rate": 3e-05,
+      "loss": 2.5888,
+      "step": 556
+    },
+    {
+      "epoch": 0.18060959792477302,
+      "grad_norm": 0.6386939883232117,
+      "learning_rate": 3e-05,
+      "loss": 3.0775,
+      "step": 557
+    },
+    {
+      "epoch": 0.18093385214007782,
+      "grad_norm": 0.6156755089759827,
+      "learning_rate": 3e-05,
+      "loss": 2.33,
+      "step": 558
+    },
+    {
+      "epoch": 0.18125810635538261,
+      "grad_norm": 0.9192589521408081,
+      "learning_rate": 3e-05,
+      "loss": 2.9808,
+      "step": 559
+    },
+    {
+      "epoch": 0.1815823605706874,
+      "grad_norm": 0.5457234382629395,
+      "learning_rate": 3e-05,
+      "loss": 2.5262,
+      "step": 560
+    },
+    {
+      "epoch": 0.1819066147859922,
+      "grad_norm": 0.5139357447624207,
+      "learning_rate": 3e-05,
+      "loss": 2.7627,
+      "step": 561
+    },
+    {
+      "epoch": 0.182230869001297,
+      "grad_norm": 0.8737220764160156,
+      "learning_rate": 3e-05,
+      "loss": 2.8111,
+      "step": 562
+    },
+    {
+      "epoch": 0.1825551232166018,
+      "grad_norm": 0.7253252267837524,
+      "learning_rate": 3e-05,
+      "loss": 2.5158,
+      "step": 563
+    },
+    {
+      "epoch": 0.1828793774319066,
+      "grad_norm": 0.37919893860816956,
+      "learning_rate": 3e-05,
+      "loss": 2.321,
+      "step": 564
+    },
+    {
+      "epoch": 0.1832036316472114,
+      "grad_norm": 0.46546998620033264,
+      "learning_rate": 3e-05,
+      "loss": 2.6096,
+      "step": 565
+    },
+    {
+      "epoch": 0.1835278858625162,
+      "grad_norm": 0.5480428338050842,
+      "learning_rate": 3e-05,
+      "loss": 2.7706,
+      "step": 566
+    },
+    {
+      "epoch": 0.18385214007782102,
+      "grad_norm": 1.2214484214782715,
+      "learning_rate": 3e-05,
+      "loss": 3.3446,
+      "step": 567
+    },
+    {
+      "epoch": 0.18417639429312582,
+      "grad_norm": 0.5335822105407715,
+      "learning_rate": 3e-05,
+      "loss": 2.863,
+      "step": 568
+    },
+    {
+      "epoch": 0.18450064850843062,
+      "grad_norm": 0.4341515302658081,
+      "learning_rate": 3e-05,
+      "loss": 2.5814,
+      "step": 569
+    },
+    {
+      "epoch": 0.18482490272373542,
+      "grad_norm": 0.3196038007736206,
+      "learning_rate": 3e-05,
+      "loss": 1.8526,
+      "step": 570
+    },
+    {
+      "epoch": 0.18514915693904022,
+      "grad_norm": 0.472495973110199,
+      "learning_rate": 3e-05,
+      "loss": 2.1448,
+      "step": 571
+    },
+    {
+      "epoch": 0.185473411154345,
+      "grad_norm": 0.8463444113731384,
+      "learning_rate": 3e-05,
+      "loss": 3.3341,
+      "step": 572
+    },
+    {
+      "epoch": 0.1857976653696498,
+      "grad_norm": 0.6128259897232056,
+      "learning_rate": 3e-05,
+      "loss": 3.2903,
+      "step": 573
+    },
+    {
+      "epoch": 0.1861219195849546,
+      "grad_norm": 0.8299535512924194,
+      "learning_rate": 3e-05,
+      "loss": 3.9265,
+      "step": 574
+    },
+    {
+      "epoch": 0.1864461738002594,
+      "grad_norm": 0.3511716425418854,
+      "learning_rate": 3e-05,
+      "loss": 2.0337,
+      "step": 575
+    },
+    {
+      "epoch": 0.1867704280155642,
+      "grad_norm": 0.4506371319293976,
+      "learning_rate": 3e-05,
+      "loss": 2.7435,
+      "step": 576
+    },
+    {
+      "epoch": 0.187094682230869,
+      "grad_norm": 0.7277122735977173,
+      "learning_rate": 3e-05,
+      "loss": 3.1784,
+      "step": 577
+    },
+    {
+      "epoch": 0.1874189364461738,
+      "grad_norm": 0.7074690461158752,
+      "learning_rate": 3e-05,
+      "loss": 2.7359,
+      "step": 578
+    },
+    {
+      "epoch": 0.1877431906614786,
+      "grad_norm": 0.5255841612815857,
+      "learning_rate": 3e-05,
+      "loss": 2.8919,
+      "step": 579
+    },
+    {
+      "epoch": 0.1880674448767834,
+      "grad_norm": 0.6812916994094849,
+      "learning_rate": 3e-05,
+      "loss": 3.5616,
+      "step": 580
+    },
+    {
+      "epoch": 0.1883916990920882,
+      "grad_norm": 0.8738076090812683,
+      "learning_rate": 3e-05,
+      "loss": 3.1305,
+      "step": 581
+    },
+    {
+      "epoch": 0.188715953307393,
+      "grad_norm": 0.3921210765838623,
+      "learning_rate": 3e-05,
+      "loss": 2.601,
+      "step": 582
+    },
+    {
+      "epoch": 0.1890402075226978,
+      "grad_norm": 0.4673861563205719,
+      "learning_rate": 3e-05,
+      "loss": 2.0567,
+      "step": 583
+    },
+    {
+      "epoch": 0.1893644617380026,
+      "grad_norm": 0.5435258746147156,
+      "learning_rate": 3e-05,
+      "loss": 2.7957,
+      "step": 584
+    },
+    {
+      "epoch": 0.18968871595330739,
+      "grad_norm": 0.47920048236846924,
+      "learning_rate": 3e-05,
+      "loss": 2.6303,
+      "step": 585
+    },
+    {
+      "epoch": 0.19001297016861218,
+      "grad_norm": 0.6047441959381104,
+      "learning_rate": 3e-05,
+      "loss": 2.8921,
+      "step": 586
+    },
+    {
+      "epoch": 0.19033722438391698,
+      "grad_norm": 1.3351110219955444,
+      "learning_rate": 3e-05,
+      "loss": 4.2067,
+      "step": 587
+    },
+    {
+      "epoch": 0.19066147859922178,
+      "grad_norm": 0.5865879654884338,
+      "learning_rate": 3e-05,
+      "loss": 2.8102,
+      "step": 588
+    },
+    {
+      "epoch": 0.19098573281452658,
+      "grad_norm": 0.3187553584575653,
+      "learning_rate": 3e-05,
+      "loss": 1.882,
+      "step": 589
+    },
+    {
+      "epoch": 0.19130998702983137,
+      "grad_norm": 0.8730062246322632,
+      "learning_rate": 3e-05,
+      "loss": 3.2953,
+      "step": 590
+    },
+    {
+      "epoch": 0.1916342412451362,
+      "grad_norm": 0.6745346784591675,
+      "learning_rate": 3e-05,
+      "loss": 3.2988,
+      "step": 591
+    },
+    {
+      "epoch": 0.191958495460441,
+      "grad_norm": 0.5809904932975769,
+      "learning_rate": 3e-05,
+      "loss": 2.4448,
+      "step": 592
+    },
+    {
+      "epoch": 0.1922827496757458,
+      "grad_norm": 0.5214886665344238,
+      "learning_rate": 3e-05,
+      "loss": 2.9776,
+      "step": 593
+    },
+    {
+      "epoch": 0.1926070038910506,
+      "grad_norm": 0.9443441033363342,
+      "learning_rate": 3e-05,
+      "loss": 3.3658,
+      "step": 594
+    },
+    {
+      "epoch": 0.1929312581063554,
+      "grad_norm": 0.6118474006652832,
+      "learning_rate": 3e-05,
+      "loss": 2.5735,
+      "step": 595
+    },
+    {
+      "epoch": 0.1932555123216602,
+      "grad_norm": 1.3823214769363403,
+      "learning_rate": 3e-05,
+      "loss": 3.1607,
+      "step": 596
+    },
+    {
+      "epoch": 0.193579766536965,
+      "grad_norm": 0.5286158919334412,
+      "learning_rate": 3e-05,
+      "loss": 2.6437,
+      "step": 597
+    },
+    {
+      "epoch": 0.19390402075226978,
+      "grad_norm": 0.7138181924819946,
+      "learning_rate": 3e-05,
+      "loss": 3.4811,
+      "step": 598
+    },
+    {
+      "epoch": 0.19422827496757458,
+      "grad_norm": 0.6960625052452087,
+      "learning_rate": 3e-05,
+      "loss": 2.9869,
+      "step": 599
+    },
+    {
+      "epoch": 0.19455252918287938,
+      "grad_norm": 1.146842360496521,
+      "learning_rate": 3e-05,
+      "loss": 2.9548,
+      "step": 600
+    },
+    {
+      "epoch": 0.19487678339818418,
+      "grad_norm": 1.038044810295105,
+      "learning_rate": 3e-05,
+      "loss": 3.1584,
+      "step": 601
+    },
+    {
+      "epoch": 0.19520103761348898,
+      "grad_norm": 0.6122036576271057,
+      "learning_rate": 3e-05,
+      "loss": 3.3223,
+      "step": 602
+    },
+    {
+      "epoch": 0.19552529182879377,
+      "grad_norm": 0.9626026749610901,
+      "learning_rate": 3e-05,
+      "loss": 3.1786,
+      "step": 603
+    },
+    {
+      "epoch": 0.19584954604409857,
+      "grad_norm": 0.6192612051963806,
+      "learning_rate": 3e-05,
+      "loss": 2.6205,
+      "step": 604
+    },
+    {
+      "epoch": 0.19617380025940337,
+      "grad_norm": 0.6861389875411987,
+      "learning_rate": 3e-05,
+      "loss": 2.7874,
+      "step": 605
+    },
+    {
+      "epoch": 0.19649805447470817,
+      "grad_norm": 0.7164813876152039,
+      "learning_rate": 3e-05,
+      "loss": 2.593,
+      "step": 606
+    },
+    {
+      "epoch": 0.19682230869001296,
+      "grad_norm": 0.45475277304649353,
+      "learning_rate": 3e-05,
+      "loss": 2.5288,
+      "step": 607
+    },
+    {
+      "epoch": 0.19714656290531776,
+      "grad_norm": 0.8449373245239258,
+      "learning_rate": 3e-05,
+      "loss": 2.8585,
+      "step": 608
+    },
+    {
+      "epoch": 0.19747081712062256,
+      "grad_norm": 0.4997235834598541,
+      "learning_rate": 3e-05,
+      "loss": 2.868,
+      "step": 609
+    },
+    {
+      "epoch": 0.19779507133592736,
+      "grad_norm": 0.5529733300209045,
+      "learning_rate": 3e-05,
+      "loss": 2.6726,
+      "step": 610
+    },
+    {
+      "epoch": 0.19811932555123216,
+      "grad_norm": 0.34082671999931335,
+      "learning_rate": 3e-05,
+      "loss": 2.3458,
+      "step": 611
+    },
+    {
+      "epoch": 0.19844357976653695,
+      "grad_norm": 0.4057732820510864,
+      "learning_rate": 3e-05,
+      "loss": 2.5625,
+      "step": 612
+    },
+    {
+      "epoch": 0.19876783398184175,
+      "grad_norm": 0.7269879579544067,
+      "learning_rate": 3e-05,
+      "loss": 2.8705,
+      "step": 613
+    },
+    {
+      "epoch": 0.19909208819714655,
+      "grad_norm": 0.4163285493850708,
+      "learning_rate": 3e-05,
+      "loss": 2.4406,
+      "step": 614
+    },
+    {
+      "epoch": 0.19941634241245138,
+      "grad_norm": 0.8123195767402649,
+      "learning_rate": 3e-05,
+      "loss": 2.7734,
+      "step": 615
+    },
+    {
+      "epoch": 0.19974059662775617,
+      "grad_norm": 0.670996367931366,
+      "learning_rate": 3e-05,
+      "loss": 3.0694,
+      "step": 616
+    },
+    {
+      "epoch": 0.20006485084306097,
+      "grad_norm": 0.6573699712753296,
+      "learning_rate": 3e-05,
+      "loss": 3.1377,
+      "step": 617
+    },
+    {
+      "epoch": 0.20038910505836577,
+      "grad_norm": 0.4109354019165039,
+      "learning_rate": 3e-05,
+      "loss": 2.8498,
+      "step": 618
+    },
+    {
+      "epoch": 0.20071335927367057,
+      "grad_norm": 0.4426591396331787,
+      "learning_rate": 3e-05,
+      "loss": 2.5303,
+      "step": 619
+    },
+    {
+      "epoch": 0.20103761348897536,
+      "grad_norm": 0.3850404620170593,
+      "learning_rate": 3e-05,
+      "loss": 2.4879,
+      "step": 620
+    },
+    {
+      "epoch": 0.20136186770428016,
+      "grad_norm": 0.3282124102115631,
+      "learning_rate": 3e-05,
+      "loss": 2.3304,
+      "step": 621
+    },
+    {
+      "epoch": 0.20168612191958496,
+      "grad_norm": 0.8417193293571472,
+      "learning_rate": 3e-05,
+      "loss": 3.7596,
+      "step": 622
+    },
+    {
+      "epoch": 0.20201037613488976,
+      "grad_norm": 0.3370361626148224,
+      "learning_rate": 3e-05,
+      "loss": 2.3659,
+      "step": 623
+    },
+    {
+      "epoch": 0.20233463035019456,
+      "grad_norm": 0.47991448640823364,
+      "learning_rate": 3e-05,
+      "loss": 2.6795,
+      "step": 624
+    },
+    {
+      "epoch": 0.20265888456549935,
+      "grad_norm": 0.5947468280792236,
+      "learning_rate": 3e-05,
+      "loss": 3.067,
+      "step": 625
+    },
+    {
+      "epoch": 0.20298313878080415,
+      "grad_norm": 0.5366941094398499,
+      "learning_rate": 3e-05,
+      "loss": 2.7266,
+      "step": 626
+    },
+    {
+      "epoch": 0.20330739299610895,
+      "grad_norm": 0.5762605667114258,
+      "learning_rate": 3e-05,
+      "loss": 2.5823,
+      "step": 627
+    },
+    {
+      "epoch": 0.20363164721141375,
+      "grad_norm": 0.3306184709072113,
+      "learning_rate": 3e-05,
+      "loss": 2.216,
+      "step": 628
+    },
+    {
+      "epoch": 0.20395590142671854,
+      "grad_norm": 0.5724536776542664,
+      "learning_rate": 3e-05,
+      "loss": 2.6854,
+      "step": 629
+    },
+    {
+      "epoch": 0.20428015564202334,
+      "grad_norm": 0.6035904288291931,
+      "learning_rate": 3e-05,
+      "loss": 2.666,
+      "step": 630
+    },
+    {
+      "epoch": 0.20460440985732814,
+      "grad_norm": 0.4908602237701416,
+      "learning_rate": 3e-05,
+      "loss": 2.2219,
+      "step": 631
+    },
+    {
+      "epoch": 0.20492866407263294,
+      "grad_norm": 0.40935036540031433,
+      "learning_rate": 3e-05,
+      "loss": 2.2492,
+      "step": 632
+    },
+    {
+      "epoch": 0.20525291828793774,
+      "grad_norm": 0.8115411400794983,
+      "learning_rate": 3e-05,
+      "loss": 2.8117,
+      "step": 633
+    },
+    {
+      "epoch": 0.20557717250324253,
+      "grad_norm": 0.36841490864753723,
+      "learning_rate": 3e-05,
+      "loss": 2.3384,
+      "step": 634
+    },
+    {
+      "epoch": 0.20590142671854733,
+      "grad_norm": 0.6712201833724976,
+      "learning_rate": 3e-05,
+      "loss": 3.0101,
+      "step": 635
+    },
+    {
+      "epoch": 0.20622568093385213,
+      "grad_norm": 0.8300108909606934,
+      "learning_rate": 3e-05,
+      "loss": 3.1934,
+      "step": 636
+    },
+    {
+      "epoch": 0.20654993514915693,
+      "grad_norm": 0.6002539992332458,
+      "learning_rate": 3e-05,
+      "loss": 2.6231,
+      "step": 637
+    },
+    {
+      "epoch": 0.20687418936446172,
+      "grad_norm": 0.5007717609405518,
+      "learning_rate": 3e-05,
+      "loss": 3.0547,
+      "step": 638
+    },
+    {
+      "epoch": 0.20719844357976655,
+      "grad_norm": 0.37764233350753784,
+      "learning_rate": 3e-05,
+      "loss": 2.44,
+      "step": 639
+    },
+    {
+      "epoch": 0.20752269779507135,
+      "grad_norm": 0.6653828024864197,
+      "learning_rate": 3e-05,
+      "loss": 3.0345,
+      "step": 640
+    },
+    {
+      "epoch": 0.20784695201037615,
+      "grad_norm": 0.6956347227096558,
+      "learning_rate": 3e-05,
+      "loss": 3.0612,
+      "step": 641
+    },
+    {
+      "epoch": 0.20817120622568094,
+      "grad_norm": 0.637048602104187,
+      "learning_rate": 3e-05,
+      "loss": 2.6811,
+      "step": 642
+    },
+    {
+      "epoch": 0.20849546044098574,
+      "grad_norm": 0.5668857097625732,
+      "learning_rate": 3e-05,
+      "loss": 3.187,
+      "step": 643
+    },
+    {
+      "epoch": 0.20881971465629054,
+      "grad_norm": 0.6416916251182556,
+      "learning_rate": 3e-05,
+      "loss": 3.3468,
+      "step": 644
+    },
+    {
+      "epoch": 0.20914396887159534,
+      "grad_norm": 1.0018491744995117,
+      "learning_rate": 3e-05,
+      "loss": 2.9971,
+      "step": 645
+    },
+    {
+      "epoch": 0.20946822308690013,
+      "grad_norm": 0.5023985505104065,
+      "learning_rate": 3e-05,
+      "loss": 2.4579,
+      "step": 646
+    },
+    {
+      "epoch": 0.20979247730220493,
+      "grad_norm": 0.5833659172058105,
+      "learning_rate": 3e-05,
+      "loss": 2.8735,
+      "step": 647
+    },
+    {
+      "epoch": 0.21011673151750973,
+      "grad_norm": 0.9475619196891785,
+      "learning_rate": 3e-05,
+      "loss": 3.6852,
+      "step": 648
+    },
+    {
+      "epoch": 0.21044098573281453,
+      "grad_norm": 0.5293980836868286,
+      "learning_rate": 3e-05,
+      "loss": 2.9791,
+      "step": 649
+    },
+    {
+      "epoch": 0.21076523994811933,
+      "grad_norm": 0.49721434712409973,
+      "learning_rate": 3e-05,
+      "loss": 2.6844,
+      "step": 650
+    },
+    {
+      "epoch": 0.21108949416342412,
+      "grad_norm": 0.5258075594902039,
+      "learning_rate": 3e-05,
+      "loss": 3.3374,
+      "step": 651
+    },
+    {
+      "epoch": 0.21141374837872892,
+      "grad_norm": 1.0863511562347412,
+      "learning_rate": 3e-05,
+      "loss": 3.5361,
+      "step": 652
+    },
+    {
+      "epoch": 0.21173800259403372,
+      "grad_norm": 0.7522532343864441,
+      "learning_rate": 3e-05,
+      "loss": 3.1369,
+      "step": 653
+    },
+    {
+      "epoch": 0.21206225680933852,
+      "grad_norm": 0.48903346061706543,
+      "learning_rate": 3e-05,
+      "loss": 2.8043,
+      "step": 654
+    },
+    {
+      "epoch": 0.21238651102464332,
+      "grad_norm": 0.7583860754966736,
+      "learning_rate": 3e-05,
+      "loss": 3.6502,
+      "step": 655
+    },
+    {
+      "epoch": 0.2127107652399481,
+      "grad_norm": 0.3995281457901001,
+      "learning_rate": 3e-05,
+      "loss": 2.4326,
+      "step": 656
+    },
+    {
+      "epoch": 0.2130350194552529,
+      "grad_norm": 0.6274381875991821,
+      "learning_rate": 3e-05,
+      "loss": 2.8351,
+      "step": 657
+    },
+    {
+      "epoch": 0.2133592736705577,
+      "grad_norm": 1.104507565498352,
+      "learning_rate": 3e-05,
+      "loss": 3.6037,
+      "step": 658
+    },
+    {
+      "epoch": 0.2136835278858625,
+      "grad_norm": 0.4919705092906952,
+      "learning_rate": 3e-05,
+      "loss": 2.6095,
+      "step": 659
+    },
+    {
+      "epoch": 0.2140077821011673,
+      "grad_norm": 0.8049091696739197,
+      "learning_rate": 3e-05,
+      "loss": 3.2209,
+      "step": 660
+    },
+    {
+      "epoch": 0.2143320363164721,
+      "grad_norm": 0.5571126341819763,
+      "learning_rate": 3e-05,
+      "loss": 3.0004,
+      "step": 661
+    },
+    {
+      "epoch": 0.2146562905317769,
+      "grad_norm": 0.5219164490699768,
+      "learning_rate": 3e-05,
+      "loss": 2.8952,
+      "step": 662
+    },
+    {
+      "epoch": 0.21498054474708173,
+      "grad_norm": 0.5346251130104065,
+      "learning_rate": 3e-05,
+      "loss": 3.2931,
+      "step": 663
+    },
+    {
+      "epoch": 0.21530479896238652,
+      "grad_norm": 0.591167688369751,
+      "learning_rate": 3e-05,
+      "loss": 3.0325,
+      "step": 664
+    },
+    {
+      "epoch": 0.21562905317769132,
+      "grad_norm": 0.5756276845932007,
+      "learning_rate": 3e-05,
+      "loss": 3.2101,
+      "step": 665
+    },
+    {
+      "epoch": 0.21595330739299612,
+      "grad_norm": 1.084157943725586,
+      "learning_rate": 3e-05,
+      "loss": 3.6147,
+      "step": 666
+    },
+    {
+      "epoch": 0.21627756160830092,
+      "grad_norm": 0.5247547030448914,
+      "learning_rate": 3e-05,
+      "loss": 3.1252,
+      "step": 667
+    },
+    {
+      "epoch": 0.21660181582360571,
+      "grad_norm": 0.7057146430015564,
+      "learning_rate": 3e-05,
+      "loss": 3.3921,
+      "step": 668
+    },
+    {
+      "epoch": 0.2169260700389105,
+      "grad_norm": 0.6902710795402527,
+      "learning_rate": 3e-05,
+      "loss": 2.6694,
+      "step": 669
+    },
+    {
+      "epoch": 0.2172503242542153,
+      "grad_norm": 0.5417840480804443,
+      "learning_rate": 3e-05,
+      "loss": 3.1882,
+      "step": 670
+    },
+    {
+      "epoch": 0.2175745784695201,
+      "grad_norm": 0.3860093653202057,
+      "learning_rate": 3e-05,
+      "loss": 2.473,
+      "step": 671
+    },
+    {
+      "epoch": 0.2178988326848249,
+      "grad_norm": 0.3936418890953064,
+      "learning_rate": 3e-05,
+      "loss": 2.343,
+      "step": 672
+    },
+    {
+      "epoch": 0.2182230869001297,
+      "grad_norm": 0.4849075973033905,
+      "learning_rate": 3e-05,
+      "loss": 2.7487,
+      "step": 673
+    },
+    {
+      "epoch": 0.2185473411154345,
+      "grad_norm": 0.5063043236732483,
+      "learning_rate": 3e-05,
+      "loss": 2.4949,
+      "step": 674
+    },
+    {
+      "epoch": 0.2188715953307393,
+      "grad_norm": 0.5540835857391357,
+      "learning_rate": 3e-05,
+      "loss": 2.8841,
+      "step": 675
+    },
+    {
+      "epoch": 0.2191958495460441,
+      "grad_norm": 0.43374526500701904,
+      "learning_rate": 3e-05,
+      "loss": 2.8368,
+      "step": 676
+    },
+    {
+      "epoch": 0.2195201037613489,
+      "grad_norm": 0.5101787447929382,
+      "learning_rate": 3e-05,
+      "loss": 2.7815,
+      "step": 677
+    },
+    {
+      "epoch": 0.2198443579766537,
+      "grad_norm": 0.4803546071052551,
+      "learning_rate": 3e-05,
+      "loss": 2.6116,
+      "step": 678
+    },
+    {
+      "epoch": 0.2201686121919585,
+      "grad_norm": 0.45843422412872314,
+      "learning_rate": 3e-05,
+      "loss": 2.8503,
+      "step": 679
+    },
+    {
+      "epoch": 0.2204928664072633,
+      "grad_norm": 0.654393196105957,
+      "learning_rate": 3e-05,
+      "loss": 3.2359,
+      "step": 680
+    },
+    {
+      "epoch": 0.22081712062256809,
+      "grad_norm": 0.7138974666595459,
+      "learning_rate": 3e-05,
+      "loss": 3.1234,
+      "step": 681
+    },
+    {
+      "epoch": 0.22114137483787288,
+      "grad_norm": 0.35799074172973633,
+      "learning_rate": 3e-05,
+      "loss": 2.154,
+      "step": 682
+    },
+    {
+      "epoch": 0.22146562905317768,
+      "grad_norm": 0.49411740899086,
+      "learning_rate": 3e-05,
+      "loss": 2.3735,
+      "step": 683
+    },
+    {
+      "epoch": 0.22178988326848248,
+      "grad_norm": 0.5778548121452332,
+      "learning_rate": 3e-05,
+      "loss": 2.6286,
+      "step": 684
+    },
+    {
+      "epoch": 0.22211413748378728,
+      "grad_norm": 0.5416801571846008,
+      "learning_rate": 3e-05,
+      "loss": 2.5704,
+      "step": 685
+    },
+    {
+      "epoch": 0.22243839169909208,
+      "grad_norm": 0.8196271061897278,
+      "learning_rate": 3e-05,
+      "loss": 3.2575,
+      "step": 686
+    },
+    {
+      "epoch": 0.2227626459143969,
+      "grad_norm": 0.6107181906700134,
+      "learning_rate": 3e-05,
+      "loss": 3.2336,
+      "step": 687
+    },
+    {
+      "epoch": 0.2230869001297017,
+      "grad_norm": 1.0187643766403198,
+      "learning_rate": 3e-05,
+      "loss": 4.0318,
+      "step": 688
+    },
+    {
+      "epoch": 0.2234111543450065,
+      "grad_norm": 0.47535061836242676,
+      "learning_rate": 3e-05,
+      "loss": 2.6467,
+      "step": 689
+    },
+    {
+      "epoch": 0.2237354085603113,
+      "grad_norm": 0.5755884051322937,
+      "learning_rate": 3e-05,
+      "loss": 3.1385,
+      "step": 690
+    },
+    {
+      "epoch": 0.2240596627756161,
+      "grad_norm": 0.5970344543457031,
+      "learning_rate": 3e-05,
+      "loss": 2.8521,
+      "step": 691
+    },
+    {
+      "epoch": 0.2243839169909209,
+      "grad_norm": 0.6138725876808167,
+      "learning_rate": 3e-05,
+      "loss": 3.6003,
+      "step": 692
+    },
+    {
+      "epoch": 0.2247081712062257,
+      "grad_norm": 0.649166464805603,
+      "learning_rate": 3e-05,
+      "loss": 3.5793,
+      "step": 693
+    },
+    {
+      "epoch": 0.22503242542153049,
+      "grad_norm": 0.5929542183876038,
+      "learning_rate": 3e-05,
+      "loss": 2.6602,
+      "step": 694
+    },
+    {
+      "epoch": 0.22535667963683528,
+      "grad_norm": 0.48995569348335266,
+      "learning_rate": 3e-05,
+      "loss": 2.648,
+      "step": 695
+    },
+    {
+      "epoch": 0.22568093385214008,
+      "grad_norm": 0.6453479528427124,
+      "learning_rate": 3e-05,
+      "loss": 2.6031,
+      "step": 696
+    },
+    {
+      "epoch": 0.22600518806744488,
+      "grad_norm": 0.46323737502098083,
+      "learning_rate": 3e-05,
+      "loss": 2.3561,
+      "step": 697
+    },
+    {
+      "epoch": 0.22632944228274968,
+      "grad_norm": 0.4369601309299469,
+      "learning_rate": 3e-05,
+      "loss": 2.5121,
+      "step": 698
+    },
+    {
+      "epoch": 0.22665369649805447,
+      "grad_norm": 0.8469499349594116,
+      "learning_rate": 3e-05,
+      "loss": 3.0536,
+      "step": 699
+    },
+    {
+      "epoch": 0.22697795071335927,
+      "grad_norm": 0.5674737691879272,
+      "learning_rate": 3e-05,
+      "loss": 3.4239,
+      "step": 700
+    },
+    {
+      "epoch": 0.22730220492866407,
+      "grad_norm": 0.5338563919067383,
+      "learning_rate": 3e-05,
+      "loss": 2.8795,
+      "step": 701
+    },
+    {
+      "epoch": 0.22762645914396887,
+      "grad_norm": 0.7521704435348511,
+      "learning_rate": 3e-05,
+      "loss": 2.692,
+      "step": 702
+    },
+    {
+      "epoch": 0.22795071335927367,
+      "grad_norm": 0.7360763549804688,
+      "learning_rate": 3e-05,
+      "loss": 2.8246,
+      "step": 703
+    },
+    {
+      "epoch": 0.22827496757457846,
+      "grad_norm": 0.3520442247390747,
+      "learning_rate": 3e-05,
+      "loss": 2.252,
+      "step": 704
+    },
+    {
+      "epoch": 0.22859922178988326,
+      "grad_norm": 0.5350028872489929,
+      "learning_rate": 3e-05,
+      "loss": 2.7432,
+      "step": 705
+    },
+    {
+      "epoch": 0.22892347600518806,
+      "grad_norm": 0.42630794644355774,
+      "learning_rate": 3e-05,
+      "loss": 2.3031,
+      "step": 706
+    },
+    {
+      "epoch": 0.22924773022049286,
+      "grad_norm": 0.40515443682670593,
+      "learning_rate": 3e-05,
+      "loss": 2.5623,
+      "step": 707
+    },
+    {
+      "epoch": 0.22957198443579765,
+      "grad_norm": 0.5237495303153992,
+      "learning_rate": 3e-05,
+      "loss": 2.5982,
+      "step": 708
+    },
+    {
+      "epoch": 0.22989623865110245,
+      "grad_norm": 0.7773265242576599,
+      "learning_rate": 3e-05,
+      "loss": 2.8697,
+      "step": 709
+    },
+    {
+      "epoch": 0.23022049286640725,
+      "grad_norm": 0.9476958513259888,
+      "learning_rate": 3e-05,
+      "loss": 3.1217,
+      "step": 710
+    },
+    {
+      "epoch": 0.23054474708171208,
+      "grad_norm": 0.7086666226387024,
+      "learning_rate": 3e-05,
+      "loss": 3.3149,
+      "step": 711
+    },
+    {
+      "epoch": 0.23086900129701687,
+      "grad_norm": 0.7462046146392822,
+      "learning_rate": 3e-05,
+      "loss": 3.3795,
+      "step": 712
+    },
+    {
+      "epoch": 0.23119325551232167,
+      "grad_norm": 0.724816620349884,
+      "learning_rate": 3e-05,
+      "loss": 2.8584,
+      "step": 713
+    },
+    {
+      "epoch": 0.23151750972762647,
+      "grad_norm": 0.39265692234039307,
+      "learning_rate": 3e-05,
+      "loss": 2.2117,
+      "step": 714
+    },
+    {
+      "epoch": 0.23184176394293127,
+      "grad_norm": 0.6535349488258362,
+      "learning_rate": 3e-05,
+      "loss": 2.9606,
+      "step": 715
+    },
+    {
+      "epoch": 0.23216601815823606,
+      "grad_norm": 0.6737574934959412,
+      "learning_rate": 3e-05,
+      "loss": 3.0101,
+      "step": 716
+    },
+    {
+      "epoch": 0.23249027237354086,
+      "grad_norm": 0.6305509805679321,
+      "learning_rate": 3e-05,
+      "loss": 3.1431,
+      "step": 717
+    },
+    {
+      "epoch": 0.23281452658884566,
+      "grad_norm": 0.47166550159454346,
+      "learning_rate": 3e-05,
+      "loss": 2.6362,
+      "step": 718
+    },
+    {
+      "epoch": 0.23313878080415046,
+      "grad_norm": 0.5452982187271118,
+      "learning_rate": 3e-05,
+      "loss": 2.9785,
+      "step": 719
+    },
+    {
+      "epoch": 0.23346303501945526,
+      "grad_norm": 0.5587636232376099,
+      "learning_rate": 3e-05,
+      "loss": 3.0811,
+      "step": 720
+    },
+    {
+      "epoch": 0.23378728923476005,
+      "grad_norm": 0.6038868427276611,
+      "learning_rate": 3e-05,
+      "loss": 3.1882,
+      "step": 721
+    },
+    {
+      "epoch": 0.23411154345006485,
+      "grad_norm": 0.675064206123352,
+      "learning_rate": 3e-05,
+      "loss": 3.2042,
+      "step": 722
+    },
+    {
+      "epoch": 0.23443579766536965,
+      "grad_norm": 0.6770980954170227,
+      "learning_rate": 3e-05,
+      "loss": 3.2027,
+      "step": 723
+    },
+    {
+      "epoch": 0.23476005188067445,
+      "grad_norm": 0.6449234485626221,
+      "learning_rate": 3e-05,
+      "loss": 3.1643,
+      "step": 724
+    },
+    {
+      "epoch": 0.23508430609597925,
+      "grad_norm": 0.9414512515068054,
+      "learning_rate": 3e-05,
+      "loss": 3.1188,
+      "step": 725
+    },
+    {
+      "epoch": 0.23540856031128404,
+      "grad_norm": 0.46925151348114014,
+      "learning_rate": 3e-05,
+      "loss": 2.6131,
+      "step": 726
+    },
+    {
+      "epoch": 0.23573281452658884,
+      "grad_norm": 0.5054044127464294,
+      "learning_rate": 3e-05,
+      "loss": 2.88,
+      "step": 727
+    },
+    {
+      "epoch": 0.23605706874189364,
+      "grad_norm": 0.43413135409355164,
+      "learning_rate": 3e-05,
+      "loss": 2.597,
+      "step": 728
+    },
+    {
+      "epoch": 0.23638132295719844,
+      "grad_norm": 0.39313945174217224,
+      "learning_rate": 3e-05,
+      "loss": 2.1194,
+      "step": 729
+    },
+    {
+      "epoch": 0.23670557717250323,
+      "grad_norm": 0.5625256299972534,
+      "learning_rate": 3e-05,
+      "loss": 2.9969,
+      "step": 730
+    },
+    {
+      "epoch": 0.23702983138780803,
+      "grad_norm": 0.5629372000694275,
+      "learning_rate": 3e-05,
+      "loss": 3.1004,
+      "step": 731
+    },
+    {
+      "epoch": 0.23735408560311283,
+      "grad_norm": 0.4292953312397003,
+      "learning_rate": 3e-05,
+      "loss": 2.3363,
+      "step": 732
+    },
+    {
+      "epoch": 0.23767833981841763,
+      "grad_norm": 0.7561744451522827,
+      "learning_rate": 3e-05,
+      "loss": 3.0642,
+      "step": 733
+    },
+    {
+      "epoch": 0.23800259403372243,
+      "grad_norm": 0.6839069128036499,
+      "learning_rate": 3e-05,
+      "loss": 2.8802,
+      "step": 734
+    },
+    {
+      "epoch": 0.23832684824902725,
+      "grad_norm": 0.4319092333316803,
+      "learning_rate": 3e-05,
+      "loss": 2.4812,
+      "step": 735
+    },
+    {
+      "epoch": 0.23865110246433205,
+      "grad_norm": 0.3969837427139282,
+      "learning_rate": 3e-05,
+      "loss": 2.4295,
+      "step": 736
+    },
+    {
+      "epoch": 0.23897535667963685,
+      "grad_norm": 0.4676837921142578,
+      "learning_rate": 3e-05,
+      "loss": 2.4484,
+      "step": 737
+    },
+    {
+      "epoch": 0.23929961089494164,
+      "grad_norm": 0.5048397183418274,
+      "learning_rate": 3e-05,
+      "loss": 2.7932,
+      "step": 738
+    },
+    {
+      "epoch": 0.23962386511024644,
+      "grad_norm": 0.6628028750419617,
+      "learning_rate": 3e-05,
+      "loss": 2.9782,
+      "step": 739
+    },
+    {
+      "epoch": 0.23994811932555124,
+      "grad_norm": 0.3311907947063446,
+      "learning_rate": 3e-05,
+      "loss": 2.4179,
+      "step": 740
+    },
+    {
+      "epoch": 0.24027237354085604,
+      "grad_norm": 0.41412487626075745,
+      "learning_rate": 3e-05,
+      "loss": 2.7539,
+      "step": 741
+    },
+    {
+      "epoch": 0.24059662775616084,
+      "grad_norm": 0.36422622203826904,
+      "learning_rate": 3e-05,
+      "loss": 2.5021,
+      "step": 742
+    },
+    {
+      "epoch": 0.24092088197146563,
+      "grad_norm": 0.5025641322135925,
+      "learning_rate": 3e-05,
+      "loss": 2.2962,
+      "step": 743
+    },
+    {
+      "epoch": 0.24124513618677043,
+      "grad_norm": 0.7100929021835327,
+      "learning_rate": 3e-05,
+      "loss": 3.3462,
+      "step": 744
+    },
+    {
+      "epoch": 0.24156939040207523,
+      "grad_norm": 0.8952285051345825,
+      "learning_rate": 3e-05,
+      "loss": 3.393,
+      "step": 745
+    },
+    {
+      "epoch": 0.24189364461738003,
+      "grad_norm": 0.43522584438323975,
+      "learning_rate": 3e-05,
+      "loss": 2.7596,
+      "step": 746
+    },
+    {
+      "epoch": 0.24221789883268482,
+      "grad_norm": 0.7664096355438232,
+      "learning_rate": 3e-05,
+      "loss": 2.7851,
+      "step": 747
+    },
+    {
+      "epoch": 0.24254215304798962,
+      "grad_norm": 0.47800686955451965,
+      "learning_rate": 3e-05,
+      "loss": 2.4666,
+      "step": 748
+    },
+    {
+      "epoch": 0.24286640726329442,
+      "grad_norm": 1.1177982091903687,
+      "learning_rate": 3e-05,
+      "loss": 3.6211,
+      "step": 749
+    },
+    {
+      "epoch": 0.24319066147859922,
+      "grad_norm": 0.7494988441467285,
+      "learning_rate": 3e-05,
+      "loss": 2.9515,
+      "step": 750
+    },
+    {
+      "epoch": 0.24351491569390402,
+      "grad_norm": 0.8488116264343262,
+      "learning_rate": 3e-05,
+      "loss": 3.2553,
+      "step": 751
+    },
+    {
+      "epoch": 0.2438391699092088,
+      "grad_norm": 0.6400964856147766,
+      "learning_rate": 3e-05,
+      "loss": 2.6258,
+      "step": 752
+    },
+    {
+      "epoch": 0.2441634241245136,
+      "grad_norm": 0.8031836152076721,
+      "learning_rate": 3e-05,
+      "loss": 3.3766,
+      "step": 753
+    },
+    {
+      "epoch": 0.2444876783398184,
+      "grad_norm": 0.6949998736381531,
+      "learning_rate": 3e-05,
+      "loss": 2.8059,
+      "step": 754
+    },
+    {
+      "epoch": 0.2448119325551232,
+      "grad_norm": 0.355684369802475,
+      "learning_rate": 3e-05,
+      "loss": 2.086,
+      "step": 755
+    },
+    {
+      "epoch": 0.245136186770428,
+      "grad_norm": 0.48180344700813293,
+      "learning_rate": 3e-05,
+      "loss": 2.9503,
+      "step": 756
+    },
+    {
+      "epoch": 0.2454604409857328,
+      "grad_norm": 0.4545673131942749,
+      "learning_rate": 3e-05,
+      "loss": 2.3665,
+      "step": 757
+    },
+    {
+      "epoch": 0.2457846952010376,
+      "grad_norm": 0.49011802673339844,
+      "learning_rate": 3e-05,
+      "loss": 2.4522,
+      "step": 758
+    },
+    {
+      "epoch": 0.24610894941634243,
+      "grad_norm": 0.5392249822616577,
+      "learning_rate": 3e-05,
+      "loss": 2.7421,
+      "step": 759
+    },
+    {
+      "epoch": 0.24643320363164722,
+      "grad_norm": 0.472016841173172,
+      "learning_rate": 3e-05,
+      "loss": 2.4666,
+      "step": 760
+    },
+    {
+      "epoch": 0.24675745784695202,
+      "grad_norm": 0.5145785212516785,
+      "learning_rate": 3e-05,
+      "loss": 2.8343,
+      "step": 761
+    },
+    {
+      "epoch": 0.24708171206225682,
+      "grad_norm": 0.8675739169120789,
+      "learning_rate": 3e-05,
+      "loss": 3.4052,
+      "step": 762
+    },
+    {
+      "epoch": 0.24740596627756162,
+      "grad_norm": 0.6294595003128052,
+      "learning_rate": 3e-05,
+      "loss": 2.6779,
+      "step": 763
+    },
+    {
+      "epoch": 0.24773022049286642,
+      "grad_norm": 0.7377346158027649,
+      "learning_rate": 3e-05,
+      "loss": 2.7951,
+      "step": 764
+    },
+    {
+      "epoch": 0.2480544747081712,
+      "grad_norm": 0.3930434584617615,
+      "learning_rate": 3e-05,
+      "loss": 2.051,
+      "step": 765
+    },
+    {
+      "epoch": 0.248378728923476,
+      "grad_norm": 0.5885437726974487,
+      "learning_rate": 3e-05,
+      "loss": 2.5326,
+      "step": 766
+    },
+    {
+      "epoch": 0.2487029831387808,
+      "grad_norm": 0.6963761448860168,
+      "learning_rate": 3e-05,
+      "loss": 3.2813,
+      "step": 767
+    },
+    {
+      "epoch": 0.2490272373540856,
+      "grad_norm": 0.39287203550338745,
+      "learning_rate": 3e-05,
+      "loss": 2.502,
+      "step": 768
+    },
+    {
+      "epoch": 0.2493514915693904,
+      "grad_norm": 0.3966052830219269,
+      "learning_rate": 3e-05,
+      "loss": 2.1918,
+      "step": 769
+    },
+    {
+      "epoch": 0.2496757457846952,
+      "grad_norm": 0.5804434418678284,
+      "learning_rate": 3e-05,
+      "loss": 3.1225,
+      "step": 770
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4645136892795563,
+      "learning_rate": 3e-05,
+      "loss": 2.3203,
+      "step": 771
+    },
+    {
+      "epoch": 0.2503242542153048,
+      "grad_norm": 0.9936898946762085,
+      "learning_rate": 3e-05,
+      "loss": 3.2265,
+      "step": 772
+    },
+    {
+      "epoch": 0.2506485084306096,
+      "grad_norm": 0.46387016773223877,
+      "learning_rate": 3e-05,
+      "loss": 2.6085,
+      "step": 773
+    },
+    {
+      "epoch": 0.2509727626459144,
+      "grad_norm": 0.7485678195953369,
+      "learning_rate": 3e-05,
+      "loss": 3.8427,
+      "step": 774
+    },
+    {
+      "epoch": 0.2512970168612192,
+      "grad_norm": 0.4775388836860657,
+      "learning_rate": 3e-05,
+      "loss": 2.4701,
+      "step": 775
+    },
+    {
+      "epoch": 0.251621271076524,
+      "grad_norm": 0.7365149855613708,
+      "learning_rate": 3e-05,
+      "loss": 2.9643,
+      "step": 776
+    },
+    {
+      "epoch": 0.2519455252918288,
+      "grad_norm": 0.4263695180416107,
+      "learning_rate": 3e-05,
+      "loss": 2.3099,
+      "step": 777
+    },
+    {
+      "epoch": 0.2522697795071336,
+      "grad_norm": 0.5459638237953186,
+      "learning_rate": 3e-05,
+      "loss": 3.7147,
+      "step": 778
+    },
+    {
+      "epoch": 0.2525940337224384,
+      "grad_norm": 0.4120403826236725,
+      "learning_rate": 3e-05,
+      "loss": 2.4033,
+      "step": 779
+    },
+    {
+      "epoch": 0.2529182879377432,
+      "grad_norm": 0.6184048056602478,
+      "learning_rate": 3e-05,
+      "loss": 2.7862,
+      "step": 780
+    },
+    {
+      "epoch": 0.253242542153048,
+      "grad_norm": 0.49757036566734314,
+      "learning_rate": 3e-05,
+      "loss": 2.2505,
+      "step": 781
+    },
+    {
+      "epoch": 0.2535667963683528,
+      "grad_norm": 0.32198652625083923,
+      "learning_rate": 3e-05,
+      "loss": 2.1048,
+      "step": 782
+    },
+    {
+      "epoch": 0.2538910505836576,
+      "grad_norm": 0.7427743673324585,
+      "learning_rate": 3e-05,
+      "loss": 3.5673,
+      "step": 783
+    },
+    {
+      "epoch": 0.25421530479896237,
+      "grad_norm": 0.5337818264961243,
+      "learning_rate": 3e-05,
+      "loss": 2.4778,
+      "step": 784
+    },
+    {
+      "epoch": 0.25453955901426717,
+      "grad_norm": 0.5355744361877441,
+      "learning_rate": 3e-05,
+      "loss": 2.3546,
+      "step": 785
+    },
+    {
+      "epoch": 0.25486381322957197,
+      "grad_norm": 0.6086364388465881,
+      "learning_rate": 3e-05,
+      "loss": 3.3215,
+      "step": 786
+    },
+    {
+      "epoch": 0.25518806744487676,
+      "grad_norm": 0.41851088404655457,
+      "learning_rate": 3e-05,
+      "loss": 2.8199,
+      "step": 787
+    },
+    {
+      "epoch": 0.25551232166018156,
+      "grad_norm": 0.3716970980167389,
+      "learning_rate": 3e-05,
+      "loss": 2.5396,
+      "step": 788
+    },
+    {
+      "epoch": 0.25583657587548636,
+      "grad_norm": 0.43586140871047974,
+      "learning_rate": 3e-05,
+      "loss": 2.59,
+      "step": 789
+    },
+    {
+      "epoch": 0.25616083009079116,
+      "grad_norm": 0.3024660348892212,
+      "learning_rate": 3e-05,
+      "loss": 2.1819,
+      "step": 790
+    },
+    {
+      "epoch": 0.25648508430609596,
+      "grad_norm": 0.6013676524162292,
+      "learning_rate": 3e-05,
+      "loss": 2.6067,
+      "step": 791
+    },
+    {
+      "epoch": 0.25680933852140075,
+      "grad_norm": 0.47237104177474976,
+      "learning_rate": 3e-05,
+      "loss": 2.5544,
+      "step": 792
+    },
+    {
+      "epoch": 0.25713359273670555,
+      "grad_norm": 0.36199885606765747,
+      "learning_rate": 3e-05,
+      "loss": 2.2197,
+      "step": 793
+    },
+    {
+      "epoch": 0.25745784695201035,
+      "grad_norm": 0.5068061351776123,
+      "learning_rate": 3e-05,
+      "loss": 2.8955,
+      "step": 794
+    },
+    {
+      "epoch": 0.25778210116731515,
+      "grad_norm": 0.4486326575279236,
+      "learning_rate": 3e-05,
+      "loss": 2.5571,
+      "step": 795
+    },
+    {
+      "epoch": 0.25810635538262,
+      "grad_norm": 0.4050910174846649,
+      "learning_rate": 3e-05,
+      "loss": 2.3838,
+      "step": 796
+    },
+    {
+      "epoch": 0.2584306095979248,
+      "grad_norm": 0.5301112532615662,
+      "learning_rate": 3e-05,
+      "loss": 2.6702,
+      "step": 797
+    },
+    {
+      "epoch": 0.2587548638132296,
+      "grad_norm": 0.5540606379508972,
+      "learning_rate": 3e-05,
+      "loss": 2.8537,
+      "step": 798
+    },
+    {
+      "epoch": 0.2590791180285344,
+      "grad_norm": 0.341996967792511,
+      "learning_rate": 3e-05,
+      "loss": 2.3042,
+      "step": 799
+    },
+    {
+      "epoch": 0.2594033722438392,
+      "grad_norm": 0.6332293152809143,
+      "learning_rate": 3e-05,
+      "loss": 2.8644,
+      "step": 800
+    },
+    {
+      "epoch": 0.259727626459144,
+      "grad_norm": 0.667199432849884,
+      "learning_rate": 3e-05,
+      "loss": 2.9346,
+      "step": 801
+    },
+    {
+      "epoch": 0.2600518806744488,
+      "grad_norm": 0.5970064997673035,
+      "learning_rate": 3e-05,
+      "loss": 3.3909,
+      "step": 802
+    },
+    {
+      "epoch": 0.2603761348897536,
+      "grad_norm": 0.46726590394973755,
+      "learning_rate": 3e-05,
+      "loss": 2.8297,
+      "step": 803
+    },
+    {
+      "epoch": 0.2607003891050584,
+      "grad_norm": 0.6965469121932983,
+      "learning_rate": 3e-05,
+      "loss": 3.159,
+      "step": 804
+    },
+    {
+      "epoch": 0.2610246433203632,
+      "grad_norm": 0.4107237458229065,
+      "learning_rate": 3e-05,
+      "loss": 2.5608,
+      "step": 805
+    },
+    {
+      "epoch": 0.261348897535668,
+      "grad_norm": 0.4953273832798004,
+      "learning_rate": 3e-05,
+      "loss": 2.9133,
+      "step": 806
+    },
+    {
+      "epoch": 0.2616731517509728,
+      "grad_norm": 0.43170103430747986,
+      "learning_rate": 3e-05,
+      "loss": 2.4828,
+      "step": 807
+    },
+    {
+      "epoch": 0.2619974059662776,
+      "grad_norm": 0.496584951877594,
+      "learning_rate": 3e-05,
+      "loss": 2.7734,
+      "step": 808
+    },
+    {
+      "epoch": 0.2623216601815824,
+      "grad_norm": 0.47588759660720825,
+      "learning_rate": 3e-05,
+      "loss": 2.8555,
+      "step": 809
+    },
+    {
+      "epoch": 0.26264591439688717,
+      "grad_norm": 0.6026840806007385,
+      "learning_rate": 3e-05,
+      "loss": 2.667,
+      "step": 810
+    },
+    {
+      "epoch": 0.26297016861219197,
+      "grad_norm": 0.664490282535553,
+      "learning_rate": 3e-05,
+      "loss": 2.5149,
+      "step": 811
+    },
+    {
+      "epoch": 0.26329442282749677,
+      "grad_norm": 0.5206149220466614,
+      "learning_rate": 3e-05,
+      "loss": 3.1756,
+      "step": 812
+    },
+    {
+      "epoch": 0.26361867704280156,
+      "grad_norm": 0.4277188777923584,
+      "learning_rate": 3e-05,
+      "loss": 2.8447,
+      "step": 813
+    },
+    {
+      "epoch": 0.26394293125810636,
+      "grad_norm": 0.2984434962272644,
+      "learning_rate": 3e-05,
+      "loss": 2.2154,
+      "step": 814
+    },
+    {
+      "epoch": 0.26426718547341116,
+      "grad_norm": 0.7023221254348755,
+      "learning_rate": 3e-05,
+      "loss": 2.9555,
+      "step": 815
+    },
+    {
+      "epoch": 0.26459143968871596,
+      "grad_norm": 0.5019450783729553,
+      "learning_rate": 3e-05,
+      "loss": 2.4785,
+      "step": 816
+    },
+    {
+      "epoch": 0.26491569390402075,
+      "grad_norm": 0.7492998242378235,
+      "learning_rate": 3e-05,
+      "loss": 3.1367,
+      "step": 817
+    },
+    {
+      "epoch": 0.26523994811932555,
+      "grad_norm": 0.4578784108161926,
+      "learning_rate": 3e-05,
+      "loss": 2.5546,
+      "step": 818
+    },
+    {
+      "epoch": 0.26556420233463035,
+      "grad_norm": 0.45463910698890686,
+      "learning_rate": 3e-05,
+      "loss": 2.8147,
+      "step": 819
+    },
+    {
+      "epoch": 0.26588845654993515,
+      "grad_norm": 0.630575954914093,
+      "learning_rate": 3e-05,
+      "loss": 2.8994,
+      "step": 820
+    },
+    {
+      "epoch": 0.26621271076523995,
+      "grad_norm": 0.5443377494812012,
+      "learning_rate": 3e-05,
+      "loss": 2.5136,
+      "step": 821
+    },
+    {
+      "epoch": 0.26653696498054474,
+      "grad_norm": 0.39724206924438477,
+      "learning_rate": 3e-05,
+      "loss": 2.473,
+      "step": 822
+    },
+    {
+      "epoch": 0.26686121919584954,
+      "grad_norm": 0.3576723337173462,
+      "learning_rate": 3e-05,
+      "loss": 2.3109,
+      "step": 823
+    },
+    {
+      "epoch": 0.26718547341115434,
+      "grad_norm": 0.3579493463039398,
+      "learning_rate": 3e-05,
+      "loss": 2.1788,
+      "step": 824
+    },
+    {
+      "epoch": 0.26750972762645914,
+      "grad_norm": 0.6163134574890137,
+      "learning_rate": 3e-05,
+      "loss": 2.9397,
+      "step": 825
+    },
+    {
+      "epoch": 0.26783398184176394,
+      "grad_norm": 0.5778406858444214,
+      "learning_rate": 3e-05,
+      "loss": 2.7516,
+      "step": 826
+    },
+    {
+      "epoch": 0.26815823605706873,
+      "grad_norm": 0.5430405735969543,
+      "learning_rate": 3e-05,
+      "loss": 3.0809,
+      "step": 827
+    },
+    {
+      "epoch": 0.26848249027237353,
+      "grad_norm": 0.36714935302734375,
+      "learning_rate": 3e-05,
+      "loss": 2.5097,
+      "step": 828
+    },
+    {
+      "epoch": 0.26880674448767833,
+      "grad_norm": 0.38578131794929504,
+      "learning_rate": 3e-05,
+      "loss": 2.4949,
+      "step": 829
+    },
+    {
+      "epoch": 0.2691309987029831,
+      "grad_norm": 0.5868868231773376,
+      "learning_rate": 3e-05,
+      "loss": 2.8886,
+      "step": 830
+    },
+    {
+      "epoch": 0.2694552529182879,
+      "grad_norm": 0.4620862901210785,
+      "learning_rate": 3e-05,
+      "loss": 2.4331,
+      "step": 831
+    },
+    {
+      "epoch": 0.2697795071335927,
+      "grad_norm": 0.43147629499435425,
+      "learning_rate": 3e-05,
+      "loss": 2.3762,
+      "step": 832
+    },
+    {
+      "epoch": 0.2701037613488975,
+      "grad_norm": 0.48113328218460083,
+      "learning_rate": 3e-05,
+      "loss": 2.687,
+      "step": 833
+    },
+    {
+      "epoch": 0.2704280155642023,
+      "grad_norm": 0.3894902467727661,
+      "learning_rate": 3e-05,
+      "loss": 2.6019,
+      "step": 834
+    },
+    {
+      "epoch": 0.2707522697795071,
+      "grad_norm": 0.3427148461341858,
+      "learning_rate": 3e-05,
+      "loss": 2.2522,
+      "step": 835
+    },
+    {
+      "epoch": 0.2710765239948119,
+      "grad_norm": 0.7869563698768616,
+      "learning_rate": 3e-05,
+      "loss": 3.1511,
+      "step": 836
+    },
+    {
+      "epoch": 0.2714007782101167,
+      "grad_norm": 0.6199852824211121,
+      "learning_rate": 3e-05,
+      "loss": 3.6137,
+      "step": 837
+    },
+    {
+      "epoch": 0.2717250324254215,
+      "grad_norm": 0.3947635293006897,
+      "learning_rate": 3e-05,
+      "loss": 2.4244,
+      "step": 838
+    },
+    {
+      "epoch": 0.2720492866407263,
+      "grad_norm": 0.5009225010871887,
+      "learning_rate": 3e-05,
+      "loss": 2.8318,
+      "step": 839
+    },
+    {
+      "epoch": 0.2723735408560311,
+      "grad_norm": 0.8884918689727783,
+      "learning_rate": 3e-05,
+      "loss": 3.036,
+      "step": 840
+    },
+    {
+      "epoch": 0.2726977950713359,
+      "grad_norm": 0.432415634393692,
+      "learning_rate": 3e-05,
+      "loss": 2.8178,
+      "step": 841
+    },
+    {
+      "epoch": 0.2730220492866407,
+      "grad_norm": 0.5464527010917664,
+      "learning_rate": 3e-05,
+      "loss": 3.2804,
+      "step": 842
+    },
+    {
+      "epoch": 0.2733463035019455,
+      "grad_norm": 0.5110059380531311,
+      "learning_rate": 3e-05,
+      "loss": 2.8568,
+      "step": 843
+    },
+    {
+      "epoch": 0.27367055771725035,
+      "grad_norm": 0.3482668995857239,
+      "learning_rate": 3e-05,
+      "loss": 2.1653,
+      "step": 844
+    },
+    {
+      "epoch": 0.27399481193255515,
+      "grad_norm": 0.5129956007003784,
+      "learning_rate": 3e-05,
+      "loss": 2.7335,
+      "step": 845
+    },
+    {
+      "epoch": 0.27431906614785995,
+      "grad_norm": 0.8042261004447937,
+      "learning_rate": 3e-05,
+      "loss": 3.1353,
+      "step": 846
+    },
+    {
+      "epoch": 0.27464332036316474,
+      "grad_norm": 0.49207305908203125,
+      "learning_rate": 3e-05,
+      "loss": 3.029,
+      "step": 847
+    },
+    {
+      "epoch": 0.27496757457846954,
+      "grad_norm": 0.6267735958099365,
+      "learning_rate": 3e-05,
+      "loss": 3.1218,
+      "step": 848
+    },
+    {
+      "epoch": 0.27529182879377434,
+      "grad_norm": 0.4071832299232483,
+      "learning_rate": 3e-05,
+      "loss": 2.7786,
+      "step": 849
+    },
+    {
+      "epoch": 0.27561608300907914,
+      "grad_norm": 0.5524840354919434,
+      "learning_rate": 3e-05,
+      "loss": 3.0923,
+      "step": 850
+    },
+    {
+      "epoch": 0.27594033722438394,
+      "grad_norm": 1.148716688156128,
+      "learning_rate": 3e-05,
+      "loss": 2.7018,
+      "step": 851
+    },
+    {
+      "epoch": 0.27626459143968873,
+      "grad_norm": 0.37708598375320435,
+      "learning_rate": 3e-05,
+      "loss": 2.6549,
+      "step": 852
+    },
+    {
+      "epoch": 0.27658884565499353,
+      "grad_norm": 0.4228946566581726,
+      "learning_rate": 3e-05,
+      "loss": 2.4748,
+      "step": 853
+    },
+    {
+      "epoch": 0.27691309987029833,
+      "grad_norm": 0.5255026817321777,
+      "learning_rate": 3e-05,
+      "loss": 2.688,
+      "step": 854
+    },
+    {
+      "epoch": 0.2772373540856031,
+      "grad_norm": 0.8235050439834595,
+      "learning_rate": 3e-05,
+      "loss": 3.0743,
+      "step": 855
+    },
+    {
+      "epoch": 0.2775616083009079,
+      "grad_norm": 0.42073825001716614,
+      "learning_rate": 3e-05,
+      "loss": 2.5207,
+      "step": 856
+    },
+    {
+      "epoch": 0.2778858625162127,
+      "grad_norm": 0.5220509171485901,
+      "learning_rate": 3e-05,
+      "loss": 3.3045,
+      "step": 857
+    },
+    {
+      "epoch": 0.2782101167315175,
+      "grad_norm": 0.8334913849830627,
+      "learning_rate": 3e-05,
+      "loss": 3.7851,
+      "step": 858
+    },
+    {
+      "epoch": 0.2785343709468223,
+      "grad_norm": 0.4133909046649933,
+      "learning_rate": 3e-05,
+      "loss": 2.7929,
+      "step": 859
+    },
+    {
+      "epoch": 0.2788586251621271,
+      "grad_norm": 0.8429019451141357,
+      "learning_rate": 3e-05,
+      "loss": 3.5546,
+      "step": 860
+    },
+    {
+      "epoch": 0.2791828793774319,
+      "grad_norm": 0.5576422810554504,
+      "learning_rate": 3e-05,
+      "loss": 2.9545,
+      "step": 861
+    },
+    {
+      "epoch": 0.2795071335927367,
+      "grad_norm": 0.4912596642971039,
+      "learning_rate": 3e-05,
+      "loss": 2.5834,
+      "step": 862
+    },
+    {
+      "epoch": 0.2798313878080415,
+      "grad_norm": 0.7404374480247498,
+      "learning_rate": 3e-05,
+      "loss": 3.1569,
+      "step": 863
+    },
+    {
+      "epoch": 0.2801556420233463,
+      "grad_norm": 0.49707213044166565,
+      "learning_rate": 3e-05,
+      "loss": 2.7859,
+      "step": 864
+    },
+    {
+      "epoch": 0.2804798962386511,
+      "grad_norm": 0.4106294810771942,
+      "learning_rate": 3e-05,
+      "loss": 2.2702,
+      "step": 865
+    },
+    {
+      "epoch": 0.2808041504539559,
+      "grad_norm": 0.5578663349151611,
+      "learning_rate": 3e-05,
+      "loss": 2.7579,
+      "step": 866
+    },
+    {
+      "epoch": 0.2811284046692607,
+      "grad_norm": 0.569461464881897,
+      "learning_rate": 3e-05,
+      "loss": 2.6383,
+      "step": 867
+    },
+    {
+      "epoch": 0.2814526588845655,
+      "grad_norm": 0.7622687220573425,
+      "learning_rate": 3e-05,
+      "loss": 2.9861,
+      "step": 868
+    },
+    {
+      "epoch": 0.2817769130998703,
+      "grad_norm": 0.35921913385391235,
+      "learning_rate": 3e-05,
+      "loss": 2.187,
+      "step": 869
+    },
+    {
+      "epoch": 0.2821011673151751,
+      "grad_norm": 0.6600784659385681,
+      "learning_rate": 3e-05,
+      "loss": 2.7447,
+      "step": 870
+    },
+    {
+      "epoch": 0.2824254215304799,
+      "grad_norm": 0.7182983756065369,
+      "learning_rate": 3e-05,
+      "loss": 2.6222,
+      "step": 871
+    },
+    {
+      "epoch": 0.2827496757457847,
+      "grad_norm": 0.6544340252876282,
+      "learning_rate": 3e-05,
+      "loss": 2.5752,
+      "step": 872
+    },
+    {
+      "epoch": 0.2830739299610895,
+      "grad_norm": 0.818039059638977,
+      "learning_rate": 3e-05,
+      "loss": 3.52,
+      "step": 873
+    },
+    {
+      "epoch": 0.2833981841763943,
+      "grad_norm": 0.7677620649337769,
+      "learning_rate": 3e-05,
+      "loss": 2.9811,
+      "step": 874
+    },
+    {
+      "epoch": 0.2837224383916991,
+      "grad_norm": 0.5289186835289001,
+      "learning_rate": 3e-05,
+      "loss": 2.7088,
+      "step": 875
+    },
+    {
+      "epoch": 0.2840466926070039,
+      "grad_norm": 0.4958803057670593,
+      "learning_rate": 3e-05,
+      "loss": 2.8632,
+      "step": 876
+    },
+    {
+      "epoch": 0.2843709468223087,
+      "grad_norm": 0.6255384683609009,
+      "learning_rate": 3e-05,
+      "loss": 2.701,
+      "step": 877
+    },
+    {
+      "epoch": 0.2846952010376135,
+      "grad_norm": 0.6244052052497864,
+      "learning_rate": 3e-05,
+      "loss": 2.8518,
+      "step": 878
+    },
+    {
+      "epoch": 0.2850194552529183,
+      "grad_norm": 0.6885880827903748,
+      "learning_rate": 3e-05,
+      "loss": 3.5767,
+      "step": 879
+    },
+    {
+      "epoch": 0.2853437094682231,
+      "grad_norm": 0.47798091173171997,
+      "learning_rate": 3e-05,
+      "loss": 2.5703,
+      "step": 880
+    },
+    {
+      "epoch": 0.28566796368352787,
+      "grad_norm": 0.4266071319580078,
+      "learning_rate": 3e-05,
+      "loss": 2.2979,
+      "step": 881
+    },
+    {
+      "epoch": 0.28599221789883267,
+      "grad_norm": 0.4127390384674072,
+      "learning_rate": 3e-05,
+      "loss": 2.3447,
+      "step": 882
+    },
+    {
+      "epoch": 0.28631647211413747,
+      "grad_norm": 0.601061224937439,
+      "learning_rate": 3e-05,
+      "loss": 3.0573,
+      "step": 883
+    },
+    {
+      "epoch": 0.28664072632944226,
+      "grad_norm": 0.6732063293457031,
+      "learning_rate": 3e-05,
+      "loss": 3.0664,
+      "step": 884
+    },
+    {
+      "epoch": 0.28696498054474706,
+      "grad_norm": 0.6474917531013489,
+      "learning_rate": 3e-05,
+      "loss": 3.2366,
+      "step": 885
+    },
+    {
+      "epoch": 0.28728923476005186,
+      "grad_norm": 0.4453222453594208,
+      "learning_rate": 3e-05,
+      "loss": 2.3936,
+      "step": 886
+    },
+    {
+      "epoch": 0.28761348897535666,
+      "grad_norm": 0.6809524297714233,
+      "learning_rate": 3e-05,
+      "loss": 2.9435,
+      "step": 887
+    },
+    {
+      "epoch": 0.28793774319066145,
+      "grad_norm": 0.36348336935043335,
+      "learning_rate": 3e-05,
+      "loss": 2.074,
+      "step": 888
+    },
+    {
+      "epoch": 0.28826199740596625,
+      "grad_norm": 0.47681447863578796,
+      "learning_rate": 3e-05,
+      "loss": 3.0509,
+      "step": 889
+    },
+    {
+      "epoch": 0.28858625162127105,
+      "grad_norm": 0.5100139379501343,
+      "learning_rate": 3e-05,
+      "loss": 3.0448,
+      "step": 890
+    },
+    {
+      "epoch": 0.28891050583657585,
+      "grad_norm": 0.31945228576660156,
+      "learning_rate": 3e-05,
+      "loss": 2.1639,
+      "step": 891
+    },
+    {
+      "epoch": 0.2892347600518807,
+      "grad_norm": 0.46525561809539795,
+      "learning_rate": 3e-05,
+      "loss": 2.9871,
+      "step": 892
+    },
+    {
+      "epoch": 0.2895590142671855,
+      "grad_norm": 0.5058289170265198,
+      "learning_rate": 3e-05,
+      "loss": 2.6879,
+      "step": 893
+    },
+    {
+      "epoch": 0.2898832684824903,
+      "grad_norm": 0.4673522710800171,
+      "learning_rate": 3e-05,
+      "loss": 3.315,
+      "step": 894
+    },
+    {
+      "epoch": 0.2902075226977951,
+      "grad_norm": 0.5846499800682068,
+      "learning_rate": 3e-05,
+      "loss": 2.9919,
+      "step": 895
+    },
+    {
+      "epoch": 0.2905317769130999,
+      "grad_norm": 0.42340484261512756,
+      "learning_rate": 3e-05,
+      "loss": 2.4663,
+      "step": 896
+    },
+    {
+      "epoch": 0.2908560311284047,
+      "grad_norm": 0.39643198251724243,
+      "learning_rate": 3e-05,
+      "loss": 2.4072,
+      "step": 897
+    },
+    {
+      "epoch": 0.2911802853437095,
+      "grad_norm": 0.443149209022522,
+      "learning_rate": 3e-05,
+      "loss": 3.1246,
+      "step": 898
+    },
+    {
+      "epoch": 0.2915045395590143,
+      "grad_norm": 0.5193252563476562,
+      "learning_rate": 3e-05,
+      "loss": 2.7696,
+      "step": 899
+    },
+    {
+      "epoch": 0.2918287937743191,
+      "grad_norm": 0.6589711308479309,
+      "learning_rate": 3e-05,
+      "loss": 2.361,
+      "step": 900
+    },
+    {
+      "epoch": 0.2921530479896239,
+      "grad_norm": 0.3732461929321289,
+      "learning_rate": 3e-05,
+      "loss": 2.2649,
+      "step": 901
+    },
+    {
+      "epoch": 0.2924773022049287,
+      "grad_norm": 0.6968868970870972,
+      "learning_rate": 3e-05,
+      "loss": 3.2707,
+      "step": 902
+    },
+    {
+      "epoch": 0.2928015564202335,
+      "grad_norm": 0.496919721364975,
+      "learning_rate": 3e-05,
+      "loss": 2.9642,
+      "step": 903
+    },
+    {
+      "epoch": 0.2931258106355383,
+      "grad_norm": 0.5677239894866943,
+      "learning_rate": 3e-05,
+      "loss": 2.9902,
+      "step": 904
+    },
+    {
+      "epoch": 0.2934500648508431,
+      "grad_norm": 0.6409761309623718,
+      "learning_rate": 3e-05,
+      "loss": 3.2508,
+      "step": 905
+    },
+    {
+      "epoch": 0.29377431906614787,
+      "grad_norm": 0.6694306135177612,
+      "learning_rate": 3e-05,
+      "loss": 2.7424,
+      "step": 906
+    },
+    {
+      "epoch": 0.29409857328145267,
+      "grad_norm": 0.6164602041244507,
+      "learning_rate": 3e-05,
+      "loss": 3.3699,
+      "step": 907
+    },
+    {
+      "epoch": 0.29442282749675747,
+      "grad_norm": 0.4653756320476532,
+      "learning_rate": 3e-05,
+      "loss": 3.1649,
+      "step": 908
+    },
+    {
+      "epoch": 0.29474708171206226,
+      "grad_norm": 0.33215054869651794,
+      "learning_rate": 3e-05,
+      "loss": 2.488,
+      "step": 909
+    },
+    {
+      "epoch": 0.29507133592736706,
+      "grad_norm": 0.40489378571510315,
+      "learning_rate": 3e-05,
+      "loss": 2.5423,
+      "step": 910
+    },
+    {
+      "epoch": 0.29539559014267186,
+      "grad_norm": 0.5569796562194824,
+      "learning_rate": 3e-05,
+      "loss": 2.8831,
+      "step": 911
+    },
+    {
+      "epoch": 0.29571984435797666,
+      "grad_norm": 0.6345402598381042,
+      "learning_rate": 3e-05,
+      "loss": 3.2924,
+      "step": 912
+    },
+    {
+      "epoch": 0.29604409857328146,
+      "grad_norm": 0.5806029438972473,
+      "learning_rate": 3e-05,
+      "loss": 2.9051,
+      "step": 913
+    },
+    {
+      "epoch": 0.29636835278858625,
+      "grad_norm": 0.42891520261764526,
+      "learning_rate": 3e-05,
+      "loss": 2.9067,
+      "step": 914
+    },
+    {
+      "epoch": 0.29669260700389105,
+      "grad_norm": 0.42912954092025757,
+      "learning_rate": 3e-05,
+      "loss": 2.8417,
+      "step": 915
+    },
+    {
+      "epoch": 0.29701686121919585,
+      "grad_norm": 0.52364182472229,
+      "learning_rate": 3e-05,
+      "loss": 2.8594,
+      "step": 916
+    },
+    {
+      "epoch": 0.29734111543450065,
+      "grad_norm": 0.4221334159374237,
+      "learning_rate": 3e-05,
+      "loss": 2.2914,
+      "step": 917
+    },
+    {
+      "epoch": 0.29766536964980544,
+      "grad_norm": 0.5189850926399231,
+      "learning_rate": 3e-05,
+      "loss": 2.9155,
+      "step": 918
+    },
+    {
+      "epoch": 0.29798962386511024,
+      "grad_norm": 0.5195404887199402,
+      "learning_rate": 3e-05,
+      "loss": 2.8723,
+      "step": 919
+    },
+    {
+      "epoch": 0.29831387808041504,
+      "grad_norm": 0.43594616651535034,
+      "learning_rate": 3e-05,
+      "loss": 2.4063,
+      "step": 920
+    },
+    {
+      "epoch": 0.29863813229571984,
+      "grad_norm": 0.7093835473060608,
+      "learning_rate": 3e-05,
+      "loss": 2.9251,
+      "step": 921
+    },
+    {
+      "epoch": 0.29896238651102464,
+      "grad_norm": 0.3778212070465088,
+      "learning_rate": 3e-05,
+      "loss": 2.5804,
+      "step": 922
+    },
+    {
+      "epoch": 0.29928664072632943,
+      "grad_norm": 0.7705602049827576,
+      "learning_rate": 3e-05,
+      "loss": 3.5105,
+      "step": 923
+    },
+    {
+      "epoch": 0.29961089494163423,
+      "grad_norm": 0.5839797854423523,
+      "learning_rate": 3e-05,
+      "loss": 2.7166,
+      "step": 924
+    },
+    {
+      "epoch": 0.29993514915693903,
+      "grad_norm": 0.6594537496566772,
+      "learning_rate": 3e-05,
+      "loss": 3.6343,
+      "step": 925
+    },
+    {
+      "epoch": 0.3002594033722438,
+      "grad_norm": 0.6010946035385132,
+      "learning_rate": 3e-05,
+      "loss": 3.373,
+      "step": 926
+    },
+    {
+      "epoch": 0.3005836575875486,
+      "grad_norm": 0.4086730182170868,
+      "learning_rate": 3e-05,
+      "loss": 2.3902,
+      "step": 927
+    },
+    {
+      "epoch": 0.3009079118028534,
+      "grad_norm": 0.6263770461082458,
+      "learning_rate": 3e-05,
+      "loss": 2.8743,
+      "step": 928
+    },
+    {
+      "epoch": 0.3012321660181582,
+      "grad_norm": 0.8926978707313538,
+      "learning_rate": 3e-05,
+      "loss": 3.2324,
+      "step": 929
+    },
+    {
+      "epoch": 0.301556420233463,
+      "grad_norm": 0.7156753540039062,
+      "learning_rate": 3e-05,
+      "loss": 3.3918,
+      "step": 930
+    },
+    {
+      "epoch": 0.3018806744487678,
+      "grad_norm": 0.48723042011260986,
+      "learning_rate": 3e-05,
+      "loss": 2.6189,
+      "step": 931
+    },
+    {
+      "epoch": 0.3022049286640726,
+      "grad_norm": 0.528663158416748,
+      "learning_rate": 3e-05,
+      "loss": 2.8344,
+      "step": 932
+    },
+    {
+      "epoch": 0.3025291828793774,
+      "grad_norm": 0.7297291159629822,
+      "learning_rate": 3e-05,
+      "loss": 2.5947,
+      "step": 933
+    },
+    {
+      "epoch": 0.3028534370946822,
+      "grad_norm": 0.3236597776412964,
+      "learning_rate": 3e-05,
+      "loss": 2.0619,
+      "step": 934
+    },
+    {
+      "epoch": 0.303177691309987,
+      "grad_norm": 0.6719183921813965,
+      "learning_rate": 3e-05,
+      "loss": 2.851,
+      "step": 935
+    },
+    {
+      "epoch": 0.3035019455252918,
+      "grad_norm": 0.49292534589767456,
+      "learning_rate": 3e-05,
+      "loss": 2.5548,
+      "step": 936
+    },
+    {
+      "epoch": 0.3038261997405966,
+      "grad_norm": 0.7527304887771606,
+      "learning_rate": 3e-05,
+      "loss": 2.8254,
+      "step": 937
+    },
+    {
+      "epoch": 0.3041504539559014,
+      "grad_norm": 0.4877021610736847,
+      "learning_rate": 3e-05,
+      "loss": 2.7812,
+      "step": 938
+    },
+    {
+      "epoch": 0.3044747081712062,
+      "grad_norm": 0.4564758837223053,
+      "learning_rate": 3e-05,
+      "loss": 2.8608,
+      "step": 939
+    },
+    {
+      "epoch": 0.30479896238651105,
+      "grad_norm": 0.450410932302475,
+      "learning_rate": 3e-05,
+      "loss": 2.574,
+      "step": 940
+    },
+    {
+      "epoch": 0.30512321660181585,
+      "grad_norm": 0.5457193851470947,
+      "learning_rate": 3e-05,
+      "loss": 2.6935,
+      "step": 941
+    },
+    {
+      "epoch": 0.30544747081712065,
+      "grad_norm": 0.5094232559204102,
+      "learning_rate": 3e-05,
+      "loss": 2.4729,
+      "step": 942
+    },
+    {
+      "epoch": 0.30577172503242545,
+      "grad_norm": 0.400168776512146,
+      "learning_rate": 3e-05,
+      "loss": 2.3652,
+      "step": 943
+    },
+    {
+      "epoch": 0.30609597924773024,
+      "grad_norm": 0.378609836101532,
+      "learning_rate": 3e-05,
+      "loss": 2.3545,
+      "step": 944
+    },
+    {
+      "epoch": 0.30642023346303504,
+      "grad_norm": 0.7327134013175964,
+      "learning_rate": 3e-05,
+      "loss": 3.009,
+      "step": 945
+    },
+    {
+      "epoch": 0.30674448767833984,
+      "grad_norm": 0.28948792815208435,
+      "learning_rate": 3e-05,
+      "loss": 2.0493,
+      "step": 946
+    },
+    {
+      "epoch": 0.30706874189364464,
+      "grad_norm": 0.511046290397644,
+      "learning_rate": 3e-05,
+      "loss": 2.5156,
+      "step": 947
+    },
+    {
+      "epoch": 0.30739299610894943,
+      "grad_norm": 1.279922604560852,
+      "learning_rate": 3e-05,
+      "loss": 3.918,
+      "step": 948
+    },
+    {
+      "epoch": 0.30771725032425423,
+      "grad_norm": 0.2863001525402069,
+      "learning_rate": 3e-05,
+      "loss": 2.0697,
+      "step": 949
+    },
+    {
+      "epoch": 0.30804150453955903,
+      "grad_norm": 0.41506022214889526,
+      "learning_rate": 3e-05,
+      "loss": 2.5597,
+      "step": 950
+    },
+    {
+      "epoch": 0.30836575875486383,
+      "grad_norm": 0.6336697936058044,
+      "learning_rate": 3e-05,
+      "loss": 2.9766,
+      "step": 951
+    },
+    {
+      "epoch": 0.3086900129701686,
+      "grad_norm": 0.8107984662055969,
+      "learning_rate": 3e-05,
+      "loss": 2.9716,
+      "step": 952
+    },
+    {
+      "epoch": 0.3090142671854734,
+      "grad_norm": 0.45455336570739746,
+      "learning_rate": 3e-05,
+      "loss": 2.6655,
+      "step": 953
+    },
+    {
+      "epoch": 0.3093385214007782,
+      "grad_norm": 0.3446507751941681,
+      "learning_rate": 3e-05,
+      "loss": 2.329,
+      "step": 954
+    },
+    {
+      "epoch": 0.309662775616083,
+      "grad_norm": 0.6133052706718445,
+      "learning_rate": 3e-05,
+      "loss": 3.369,
+      "step": 955
+    },
+    {
+      "epoch": 0.3099870298313878,
+      "grad_norm": 0.46877843141555786,
+      "learning_rate": 3e-05,
+      "loss": 2.6596,
+      "step": 956
+    },
+    {
+      "epoch": 0.3103112840466926,
+      "grad_norm": 0.4552307724952698,
+      "learning_rate": 3e-05,
+      "loss": 2.91,
+      "step": 957
+    },
+    {
+      "epoch": 0.3106355382619974,
+      "grad_norm": 0.6491840481758118,
+      "learning_rate": 3e-05,
+      "loss": 3.4627,
+      "step": 958
+    },
+    {
+      "epoch": 0.3109597924773022,
+      "grad_norm": 0.452769011259079,
+      "learning_rate": 3e-05,
+      "loss": 2.4148,
+      "step": 959
+    },
+    {
+      "epoch": 0.311284046692607,
+      "grad_norm": 0.4854589104652405,
+      "learning_rate": 3e-05,
+      "loss": 2.8621,
+      "step": 960
+    },
+    {
+      "epoch": 0.3116083009079118,
+      "grad_norm": 0.4908556342124939,
+      "learning_rate": 3e-05,
+      "loss": 2.6032,
+      "step": 961
+    },
+    {
+      "epoch": 0.3119325551232166,
+      "grad_norm": 0.371963232755661,
+      "learning_rate": 3e-05,
+      "loss": 2.3494,
+      "step": 962
+    },
+    {
+      "epoch": 0.3122568093385214,
+      "grad_norm": 0.37373873591423035,
+      "learning_rate": 3e-05,
+      "loss": 2.3036,
+      "step": 963
+    },
+    {
+      "epoch": 0.3125810635538262,
+      "grad_norm": 0.5100082159042358,
+      "learning_rate": 3e-05,
+      "loss": 3.3213,
+      "step": 964
+    },
+    {
+      "epoch": 0.312905317769131,
+      "grad_norm": 0.543237030506134,
+      "learning_rate": 3e-05,
+      "loss": 2.8775,
+      "step": 965
+    },
+    {
+      "epoch": 0.3132295719844358,
+      "grad_norm": 0.3861331641674042,
+      "learning_rate": 3e-05,
+      "loss": 2.3915,
+      "step": 966
+    },
+    {
+      "epoch": 0.3135538261997406,
+      "grad_norm": 0.37816861271858215,
+      "learning_rate": 3e-05,
+      "loss": 2.3732,
+      "step": 967
+    },
+    {
+      "epoch": 0.3138780804150454,
+      "grad_norm": 0.5352786779403687,
+      "learning_rate": 3e-05,
+      "loss": 2.6437,
+      "step": 968
+    },
+    {
+      "epoch": 0.3142023346303502,
+      "grad_norm": 0.48626938462257385,
+      "learning_rate": 3e-05,
+      "loss": 3.1061,
+      "step": 969
+    },
+    {
+      "epoch": 0.314526588845655,
+      "grad_norm": 0.6336450576782227,
+      "learning_rate": 3e-05,
+      "loss": 3.4607,
+      "step": 970
+    },
+    {
+      "epoch": 0.3148508430609598,
+      "grad_norm": 0.5795568823814392,
+      "learning_rate": 3e-05,
+      "loss": 3.0808,
+      "step": 971
+    },
+    {
+      "epoch": 0.3151750972762646,
+      "grad_norm": 0.501320481300354,
+      "learning_rate": 3e-05,
+      "loss": 3.0145,
+      "step": 972
+    },
+    {
+      "epoch": 0.3154993514915694,
+      "grad_norm": 0.47971445322036743,
+      "learning_rate": 3e-05,
+      "loss": 3.0998,
+      "step": 973
+    },
+    {
+      "epoch": 0.3158236057068742,
+      "grad_norm": 0.35247766971588135,
+      "learning_rate": 3e-05,
+      "loss": 2.2002,
+      "step": 974
+    },
+    {
+      "epoch": 0.316147859922179,
+      "grad_norm": 0.5320780277252197,
+      "learning_rate": 3e-05,
+      "loss": 2.8033,
+      "step": 975
+    },
+    {
+      "epoch": 0.3164721141374838,
+      "grad_norm": 0.45627561211586,
+      "learning_rate": 3e-05,
+      "loss": 2.6405,
+      "step": 976
+    },
+    {
+      "epoch": 0.31679636835278857,
+      "grad_norm": 0.36019641160964966,
+      "learning_rate": 3e-05,
+      "loss": 2.4511,
+      "step": 977
+    },
+    {
+      "epoch": 0.31712062256809337,
+      "grad_norm": 0.4169032573699951,
+      "learning_rate": 3e-05,
+      "loss": 2.7013,
+      "step": 978
+    },
+    {
+      "epoch": 0.31744487678339817,
+      "grad_norm": 0.8212884664535522,
+      "learning_rate": 3e-05,
+      "loss": 3.3463,
+      "step": 979
+    },
+    {
+      "epoch": 0.31776913099870296,
+      "grad_norm": 0.5073938369750977,
+      "learning_rate": 3e-05,
+      "loss": 2.6465,
+      "step": 980
+    },
+    {
+      "epoch": 0.31809338521400776,
+      "grad_norm": 0.5281322598457336,
+      "learning_rate": 3e-05,
+      "loss": 3.745,
+      "step": 981
+    },
+    {
+      "epoch": 0.31841763942931256,
+      "grad_norm": 0.41538822650909424,
+      "learning_rate": 3e-05,
+      "loss": 2.5166,
+      "step": 982
+    },
+    {
+      "epoch": 0.31874189364461736,
+      "grad_norm": 0.5122720003128052,
+      "learning_rate": 3e-05,
+      "loss": 2.3055,
+      "step": 983
+    },
+    {
+      "epoch": 0.31906614785992216,
+      "grad_norm": 0.4331333041191101,
+      "learning_rate": 3e-05,
+      "loss": 2.1991,
+      "step": 984
+    },
+    {
+      "epoch": 0.31939040207522695,
+      "grad_norm": 0.38671496510505676,
+      "learning_rate": 3e-05,
+      "loss": 2.8095,
+      "step": 985
+    },
+    {
+      "epoch": 0.31971465629053175,
+      "grad_norm": 0.39397186040878296,
+      "learning_rate": 3e-05,
+      "loss": 2.7517,
+      "step": 986
+    },
+    {
+      "epoch": 0.32003891050583655,
+      "grad_norm": 0.44841620326042175,
+      "learning_rate": 3e-05,
+      "loss": 2.896,
+      "step": 987
+    },
+    {
+      "epoch": 0.3203631647211414,
+      "grad_norm": 0.5253480076789856,
+      "learning_rate": 3e-05,
+      "loss": 3.1864,
+      "step": 988
+    },
+    {
+      "epoch": 0.3206874189364462,
+      "grad_norm": 0.398269921541214,
+      "learning_rate": 3e-05,
+      "loss": 2.2448,
+      "step": 989
+    },
+    {
+      "epoch": 0.321011673151751,
+      "grad_norm": 0.36403292417526245,
+      "learning_rate": 3e-05,
+      "loss": 2.3267,
+      "step": 990
+    },
+    {
+      "epoch": 0.3213359273670558,
+      "grad_norm": 0.39670422673225403,
+      "learning_rate": 3e-05,
+      "loss": 2.6214,
+      "step": 991
+    },
+    {
+      "epoch": 0.3216601815823606,
+      "grad_norm": 0.497109979391098,
+      "learning_rate": 3e-05,
+      "loss": 2.6469,
+      "step": 992
+    },
+    {
+      "epoch": 0.3219844357976654,
+      "grad_norm": 0.42875874042510986,
+      "learning_rate": 3e-05,
+      "loss": 2.4714,
+      "step": 993
+    },
+    {
+      "epoch": 0.3223086900129702,
+      "grad_norm": 0.6803904175758362,
+      "learning_rate": 3e-05,
+      "loss": 2.6934,
+      "step": 994
+    },
+    {
+      "epoch": 0.322632944228275,
+      "grad_norm": 0.4994969367980957,
+      "learning_rate": 3e-05,
+      "loss": 2.6917,
+      "step": 995
+    },
+    {
+      "epoch": 0.3229571984435798,
+      "grad_norm": 0.294212281703949,
+      "learning_rate": 3e-05,
+      "loss": 2.3076,
+      "step": 996
+    },
+    {
+      "epoch": 0.3232814526588846,
+      "grad_norm": 0.3361668288707733,
+      "learning_rate": 3e-05,
+      "loss": 2.5125,
+      "step": 997
+    },
+    {
+      "epoch": 0.3236057068741894,
+      "grad_norm": 0.5549931526184082,
+      "learning_rate": 3e-05,
+      "loss": 2.947,
+      "step": 998
+    },
+    {
+      "epoch": 0.3239299610894942,
+      "grad_norm": 0.4241522252559662,
+      "learning_rate": 3e-05,
+      "loss": 2.4376,
+      "step": 999
+    },
+    {
+      "epoch": 0.324254215304799,
+      "grad_norm": 0.4088788628578186,
+      "learning_rate": 3e-05,
+      "loss": 2.7194,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3245784695201038,
+      "grad_norm": 0.38929829001426697,
+      "learning_rate": 3e-05,
+      "loss": 2.3808,
+      "step": 1001
+    },
+    {
+      "epoch": 0.32490272373540857,
+      "grad_norm": 0.41293174028396606,
+      "learning_rate": 3e-05,
+      "loss": 2.6707,
+      "step": 1002
+    },
+    {
+      "epoch": 0.32522697795071337,
+      "grad_norm": 0.5914104580879211,
+      "learning_rate": 3e-05,
+      "loss": 2.9137,
+      "step": 1003
+    },
+    {
+      "epoch": 0.32555123216601817,
+      "grad_norm": 0.34951502084732056,
+      "learning_rate": 3e-05,
+      "loss": 2.3238,
+      "step": 1004
+    },
+    {
+      "epoch": 0.32587548638132297,
+      "grad_norm": 0.5906566977500916,
+      "learning_rate": 3e-05,
+      "loss": 3.38,
+      "step": 1005
+    },
+    {
+      "epoch": 0.32619974059662776,
+      "grad_norm": 0.4735203683376312,
+      "learning_rate": 3e-05,
+      "loss": 2.7574,
+      "step": 1006
+    },
+    {
+      "epoch": 0.32652399481193256,
+      "grad_norm": 0.49539265036582947,
+      "learning_rate": 3e-05,
+      "loss": 2.6175,
+      "step": 1007
+    },
+    {
+      "epoch": 0.32684824902723736,
+      "grad_norm": 0.9257025122642517,
+      "learning_rate": 3e-05,
+      "loss": 3.3341,
+      "step": 1008
+    },
+    {
+      "epoch": 0.32717250324254216,
+      "grad_norm": 0.31978079676628113,
+      "learning_rate": 3e-05,
+      "loss": 2.2215,
+      "step": 1009
+    },
+    {
+      "epoch": 0.32749675745784695,
+      "grad_norm": 0.4244193136692047,
+      "learning_rate": 3e-05,
+      "loss": 2.1486,
+      "step": 1010
+    },
+    {
+      "epoch": 0.32782101167315175,
+      "grad_norm": 0.3573285937309265,
+      "learning_rate": 3e-05,
+      "loss": 2.1674,
+      "step": 1011
+    },
+    {
+      "epoch": 0.32814526588845655,
+      "grad_norm": 0.45863983035087585,
+      "learning_rate": 3e-05,
+      "loss": 2.5534,
+      "step": 1012
+    },
+    {
+      "epoch": 0.32846952010376135,
+      "grad_norm": 0.34895676374435425,
+      "learning_rate": 3e-05,
+      "loss": 2.2315,
+      "step": 1013
+    },
+    {
+      "epoch": 0.32879377431906615,
+      "grad_norm": 0.48339521884918213,
+      "learning_rate": 3e-05,
+      "loss": 2.7302,
+      "step": 1014
+    },
+    {
+      "epoch": 0.32911802853437094,
+      "grad_norm": 0.5254601240158081,
+      "learning_rate": 3e-05,
+      "loss": 3.0446,
+      "step": 1015
+    },
+    {
+      "epoch": 0.32944228274967574,
+      "grad_norm": 0.4702230393886566,
+      "learning_rate": 3e-05,
+      "loss": 2.3611,
+      "step": 1016
+    },
+    {
+      "epoch": 0.32976653696498054,
+      "grad_norm": 0.41626113653182983,
+      "learning_rate": 3e-05,
+      "loss": 2.8728,
+      "step": 1017
+    },
+    {
+      "epoch": 0.33009079118028534,
+      "grad_norm": 0.5229095220565796,
+      "learning_rate": 3e-05,
+      "loss": 3.0887,
+      "step": 1018
+    },
+    {
+      "epoch": 0.33041504539559013,
+      "grad_norm": 0.32266777753829956,
+      "learning_rate": 3e-05,
+      "loss": 2.2818,
+      "step": 1019
+    },
+    {
+      "epoch": 0.33073929961089493,
+      "grad_norm": 0.5472150444984436,
+      "learning_rate": 3e-05,
+      "loss": 3.0028,
+      "step": 1020
+    },
+    {
+      "epoch": 0.33106355382619973,
+      "grad_norm": 0.5138072967529297,
+      "learning_rate": 3e-05,
+      "loss": 2.2646,
+      "step": 1021
+    },
+    {
+      "epoch": 0.33138780804150453,
+      "grad_norm": 0.5459330081939697,
+      "learning_rate": 3e-05,
+      "loss": 2.4625,
+      "step": 1022
+    },
+    {
+      "epoch": 0.3317120622568093,
+      "grad_norm": 0.6895681619644165,
+      "learning_rate": 3e-05,
+      "loss": 2.9059,
+      "step": 1023
+    },
+    {
+      "epoch": 0.3320363164721141,
+      "grad_norm": 0.36621585488319397,
+      "learning_rate": 3e-05,
+      "loss": 2.3707,
+      "step": 1024
+    },
+    {
+      "epoch": 0.3323605706874189,
+      "grad_norm": 0.3927241563796997,
+      "learning_rate": 3e-05,
+      "loss": 2.1017,
+      "step": 1025
+    },
+    {
+      "epoch": 0.3326848249027237,
+      "grad_norm": 0.625253438949585,
+      "learning_rate": 3e-05,
+      "loss": 3.1149,
+      "step": 1026
+    },
+    {
+      "epoch": 0.3330090791180285,
+      "grad_norm": 0.5783314108848572,
+      "learning_rate": 3e-05,
+      "loss": 3.1748,
+      "step": 1027
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.5214437246322632,
+      "learning_rate": 3e-05,
+      "loss": 3.1947,
+      "step": 1028
+    },
+    {
+      "epoch": 0.3336575875486381,
+      "grad_norm": 0.661126971244812,
+      "learning_rate": 3e-05,
+      "loss": 3.2207,
+      "step": 1029
+    },
+    {
+      "epoch": 0.3339818417639429,
+      "grad_norm": 0.37396642565727234,
+      "learning_rate": 3e-05,
+      "loss": 2.3853,
+      "step": 1030
+    },
+    {
+      "epoch": 0.3343060959792477,
+      "grad_norm": 0.415732741355896,
+      "learning_rate": 3e-05,
+      "loss": 2.9434,
+      "step": 1031
+    },
+    {
+      "epoch": 0.3346303501945525,
+      "grad_norm": 0.5611098408699036,
+      "learning_rate": 3e-05,
+      "loss": 3.2908,
+      "step": 1032
+    },
+    {
+      "epoch": 0.3349546044098573,
+      "grad_norm": 0.6200509071350098,
+      "learning_rate": 3e-05,
+      "loss": 2.5109,
+      "step": 1033
+    },
+    {
+      "epoch": 0.3352788586251621,
+      "grad_norm": 0.53252112865448,
+      "learning_rate": 3e-05,
+      "loss": 2.6049,
+      "step": 1034
+    },
+    {
+      "epoch": 0.3356031128404669,
+      "grad_norm": 0.33023950457572937,
+      "learning_rate": 3e-05,
+      "loss": 2.1467,
+      "step": 1035
+    },
+    {
+      "epoch": 0.3359273670557717,
+      "grad_norm": 1.1515405178070068,
+      "learning_rate": 3e-05,
+      "loss": 3.3127,
+      "step": 1036
+    },
+    {
+      "epoch": 0.33625162127107655,
+      "grad_norm": 1.0085968971252441,
+      "learning_rate": 3e-05,
+      "loss": 3.0238,
+      "step": 1037
+    },
+    {
+      "epoch": 0.33657587548638135,
+      "grad_norm": 0.3712306618690491,
+      "learning_rate": 3e-05,
+      "loss": 2.3189,
+      "step": 1038
+    },
+    {
+      "epoch": 0.33690012970168615,
+      "grad_norm": 0.6699020266532898,
+      "learning_rate": 3e-05,
+      "loss": 3.3792,
+      "step": 1039
+    },
+    {
+      "epoch": 0.33722438391699094,
+      "grad_norm": 0.6088201999664307,
+      "learning_rate": 3e-05,
+      "loss": 2.6401,
+      "step": 1040
+    },
+    {
+      "epoch": 0.33754863813229574,
+      "grad_norm": 0.8415403962135315,
+      "learning_rate": 3e-05,
+      "loss": 2.7451,
+      "step": 1041
+    },
+    {
+      "epoch": 0.33787289234760054,
+      "grad_norm": 0.7582104802131653,
+      "learning_rate": 3e-05,
+      "loss": 3.2412,
+      "step": 1042
+    },
+    {
+      "epoch": 0.33819714656290534,
+      "grad_norm": 0.43935951590538025,
+      "learning_rate": 3e-05,
+      "loss": 2.6203,
+      "step": 1043
+    },
+    {
+      "epoch": 0.33852140077821014,
+      "grad_norm": 0.4273548126220703,
+      "learning_rate": 3e-05,
+      "loss": 2.4994,
+      "step": 1044
+    },
+    {
+      "epoch": 0.33884565499351493,
+      "grad_norm": 0.4483807682991028,
+      "learning_rate": 3e-05,
+      "loss": 2.3896,
+      "step": 1045
+    },
+    {
+      "epoch": 0.33916990920881973,
+      "grad_norm": 0.37659937143325806,
+      "learning_rate": 3e-05,
+      "loss": 2.3228,
+      "step": 1046
+    },
+    {
+      "epoch": 0.33949416342412453,
+      "grad_norm": 0.49458393454551697,
+      "learning_rate": 3e-05,
+      "loss": 2.3325,
+      "step": 1047
+    },
+    {
+      "epoch": 0.3398184176394293,
+      "grad_norm": 0.6099212169647217,
+      "learning_rate": 3e-05,
+      "loss": 2.8235,
+      "step": 1048
+    },
+    {
+      "epoch": 0.3401426718547341,
+      "grad_norm": 0.3730524182319641,
+      "learning_rate": 3e-05,
+      "loss": 2.3643,
+      "step": 1049
+    },
+    {
+      "epoch": 0.3404669260700389,
+      "grad_norm": 0.9170535206794739,
+      "learning_rate": 3e-05,
+      "loss": 3.2429,
+      "step": 1050
+    },
+    {
+      "epoch": 0.3407911802853437,
+      "grad_norm": 0.6434488296508789,
+      "learning_rate": 3e-05,
+      "loss": 2.8765,
+      "step": 1051
+    },
+    {
+      "epoch": 0.3411154345006485,
+      "grad_norm": 0.39307060837745667,
+      "learning_rate": 3e-05,
+      "loss": 1.9922,
+      "step": 1052
+    },
+    {
+      "epoch": 0.3414396887159533,
+      "grad_norm": 0.5425589680671692,
+      "learning_rate": 3e-05,
+      "loss": 2.5337,
+      "step": 1053
+    },
+    {
+      "epoch": 0.3417639429312581,
+      "grad_norm": 0.5323801636695862,
+      "learning_rate": 3e-05,
+      "loss": 2.6037,
+      "step": 1054
+    },
+    {
+      "epoch": 0.3420881971465629,
+      "grad_norm": 0.5544849634170532,
+      "learning_rate": 3e-05,
+      "loss": 2.2955,
+      "step": 1055
+    },
+    {
+      "epoch": 0.3424124513618677,
+      "grad_norm": 0.49616187810897827,
+      "learning_rate": 3e-05,
+      "loss": 2.4746,
+      "step": 1056
+    },
+    {
+      "epoch": 0.3427367055771725,
+      "grad_norm": 0.41139456629753113,
+      "learning_rate": 3e-05,
+      "loss": 2.3273,
+      "step": 1057
+    },
+    {
+      "epoch": 0.3430609597924773,
+      "grad_norm": 1.1053458452224731,
+      "learning_rate": 3e-05,
+      "loss": 3.647,
+      "step": 1058
+    },
+    {
+      "epoch": 0.3433852140077821,
+      "grad_norm": 0.5757237672805786,
+      "learning_rate": 3e-05,
+      "loss": 2.7478,
+      "step": 1059
+    },
+    {
+      "epoch": 0.3437094682230869,
+      "grad_norm": 0.6032525897026062,
+      "learning_rate": 3e-05,
+      "loss": 3.1086,
+      "step": 1060
+    },
+    {
+      "epoch": 0.3440337224383917,
+      "grad_norm": 0.5692119002342224,
+      "learning_rate": 3e-05,
+      "loss": 3.4831,
+      "step": 1061
+    },
+    {
+      "epoch": 0.3443579766536965,
+      "grad_norm": 0.3667396008968353,
+      "learning_rate": 3e-05,
+      "loss": 2.3661,
+      "step": 1062
+    },
+    {
+      "epoch": 0.3446822308690013,
+      "grad_norm": 0.5311639308929443,
+      "learning_rate": 3e-05,
+      "loss": 2.5222,
+      "step": 1063
+    },
+    {
+      "epoch": 0.3450064850843061,
+      "grad_norm": 0.7020514607429504,
+      "learning_rate": 3e-05,
+      "loss": 2.3863,
+      "step": 1064
+    },
+    {
+      "epoch": 0.3453307392996109,
+      "grad_norm": 0.5091455578804016,
+      "learning_rate": 3e-05,
+      "loss": 2.4381,
+      "step": 1065
+    },
+    {
+      "epoch": 0.3456549935149157,
+      "grad_norm": 0.40224525332450867,
+      "learning_rate": 3e-05,
+      "loss": 2.8565,
+      "step": 1066
+    },
+    {
+      "epoch": 0.3459792477302205,
+      "grad_norm": 0.49794888496398926,
+      "learning_rate": 3e-05,
+      "loss": 2.9381,
+      "step": 1067
+    },
+    {
+      "epoch": 0.3463035019455253,
+      "grad_norm": 0.3929412364959717,
+      "learning_rate": 3e-05,
+      "loss": 2.8378,
+      "step": 1068
+    },
+    {
+      "epoch": 0.3466277561608301,
+      "grad_norm": 0.43183690309524536,
+      "learning_rate": 3e-05,
+      "loss": 2.5184,
+      "step": 1069
+    },
+    {
+      "epoch": 0.3469520103761349,
+      "grad_norm": 0.40891388058662415,
+      "learning_rate": 3e-05,
+      "loss": 2.3518,
+      "step": 1070
+    },
+    {
+      "epoch": 0.3472762645914397,
+      "grad_norm": 0.33023518323898315,
+      "learning_rate": 3e-05,
+      "loss": 2.1037,
+      "step": 1071
+    },
+    {
+      "epoch": 0.3476005188067445,
+      "grad_norm": 0.45663583278656006,
+      "learning_rate": 3e-05,
+      "loss": 3.0986,
+      "step": 1072
+    },
+    {
+      "epoch": 0.34792477302204927,
+      "grad_norm": 0.41894203424453735,
+      "learning_rate": 3e-05,
+      "loss": 2.7865,
+      "step": 1073
+    },
+    {
+      "epoch": 0.34824902723735407,
+      "grad_norm": 0.4957062900066376,
+      "learning_rate": 3e-05,
+      "loss": 3.2916,
+      "step": 1074
+    },
+    {
+      "epoch": 0.34857328145265887,
+      "grad_norm": 0.668289303779602,
+      "learning_rate": 3e-05,
+      "loss": 3.1633,
+      "step": 1075
+    },
+    {
+      "epoch": 0.34889753566796367,
+      "grad_norm": 0.4546116292476654,
+      "learning_rate": 3e-05,
+      "loss": 2.474,
+      "step": 1076
+    },
+    {
+      "epoch": 0.34922178988326846,
+      "grad_norm": 0.6163711547851562,
+      "learning_rate": 3e-05,
+      "loss": 2.4273,
+      "step": 1077
+    },
+    {
+      "epoch": 0.34954604409857326,
+      "grad_norm": 0.5655884146690369,
+      "learning_rate": 3e-05,
+      "loss": 2.8581,
+      "step": 1078
+    },
+    {
+      "epoch": 0.34987029831387806,
+      "grad_norm": 0.5377897024154663,
+      "learning_rate": 3e-05,
+      "loss": 2.5624,
+      "step": 1079
+    },
+    {
+      "epoch": 0.35019455252918286,
+      "grad_norm": 0.7209580540657043,
+      "learning_rate": 3e-05,
+      "loss": 3.1228,
+      "step": 1080
+    },
+    {
+      "epoch": 0.35051880674448765,
+      "grad_norm": 0.42296501994132996,
+      "learning_rate": 3e-05,
+      "loss": 2.5872,
+      "step": 1081
+    },
+    {
+      "epoch": 0.35084306095979245,
+      "grad_norm": 0.3331563472747803,
+      "learning_rate": 3e-05,
+      "loss": 2.1882,
+      "step": 1082
+    },
+    {
+      "epoch": 0.35116731517509725,
+      "grad_norm": 0.4110640287399292,
+      "learning_rate": 3e-05,
+      "loss": 2.6078,
+      "step": 1083
+    },
+    {
+      "epoch": 0.35149156939040205,
+      "grad_norm": 0.3631129562854767,
+      "learning_rate": 3e-05,
+      "loss": 2.245,
+      "step": 1084
+    },
+    {
+      "epoch": 0.3518158236057069,
+      "grad_norm": 0.867616593837738,
+      "learning_rate": 3e-05,
+      "loss": 3.3282,
+      "step": 1085
+    },
+    {
+      "epoch": 0.3521400778210117,
+      "grad_norm": 0.4596884846687317,
+      "learning_rate": 3e-05,
+      "loss": 2.5886,
+      "step": 1086
+    },
+    {
+      "epoch": 0.3524643320363165,
+      "grad_norm": 0.5417409539222717,
+      "learning_rate": 3e-05,
+      "loss": 3.023,
+      "step": 1087
+    },
+    {
+      "epoch": 0.3527885862516213,
+      "grad_norm": 0.2734490633010864,
+      "learning_rate": 3e-05,
+      "loss": 2.1037,
+      "step": 1088
+    },
+    {
+      "epoch": 0.3531128404669261,
+      "grad_norm": 0.5335866808891296,
+      "learning_rate": 3e-05,
+      "loss": 3.0935,
+      "step": 1089
+    },
+    {
+      "epoch": 0.3534370946822309,
+      "grad_norm": 0.3714607357978821,
+      "learning_rate": 3e-05,
+      "loss": 2.4409,
+      "step": 1090
+    },
+    {
+      "epoch": 0.3537613488975357,
+      "grad_norm": 0.4027988612651825,
+      "learning_rate": 3e-05,
+      "loss": 2.3341,
+      "step": 1091
+    },
+    {
+      "epoch": 0.3540856031128405,
+      "grad_norm": 0.48395007848739624,
+      "learning_rate": 3e-05,
+      "loss": 3.3463,
+      "step": 1092
+    },
+    {
+      "epoch": 0.3544098573281453,
+      "grad_norm": 0.48810601234436035,
+      "learning_rate": 3e-05,
+      "loss": 2.8271,
+      "step": 1093
+    },
+    {
+      "epoch": 0.3547341115434501,
+      "grad_norm": 0.5159446597099304,
+      "learning_rate": 3e-05,
+      "loss": 3.2485,
+      "step": 1094
+    },
+    {
+      "epoch": 0.3550583657587549,
+      "grad_norm": 0.4849570691585541,
+      "learning_rate": 3e-05,
+      "loss": 2.4579,
+      "step": 1095
+    },
+    {
+      "epoch": 0.3553826199740597,
+      "grad_norm": 0.38589388132095337,
+      "learning_rate": 3e-05,
+      "loss": 2.6571,
+      "step": 1096
+    },
+    {
+      "epoch": 0.3557068741893645,
+      "grad_norm": 0.5119343400001526,
+      "learning_rate": 3e-05,
+      "loss": 2.6776,
+      "step": 1097
+    },
+    {
+      "epoch": 0.3560311284046693,
+      "grad_norm": 0.47529423236846924,
+      "learning_rate": 3e-05,
+      "loss": 2.9236,
+      "step": 1098
+    },
+    {
+      "epoch": 0.35635538261997407,
+      "grad_norm": 0.4101375937461853,
+      "learning_rate": 3e-05,
+      "loss": 2.4818,
+      "step": 1099
+    },
+    {
+      "epoch": 0.35667963683527887,
+      "grad_norm": 0.48603275418281555,
+      "learning_rate": 3e-05,
+      "loss": 2.8755,
+      "step": 1100
+    },
+    {
+      "epoch": 0.35700389105058367,
+      "grad_norm": 0.3327399790287018,
+      "learning_rate": 3e-05,
+      "loss": 2.341,
+      "step": 1101
+    },
+    {
+      "epoch": 0.35732814526588846,
+      "grad_norm": 0.30409353971481323,
+      "learning_rate": 3e-05,
+      "loss": 2.0685,
+      "step": 1102
+    },
+    {
+      "epoch": 0.35765239948119326,
+      "grad_norm": 0.5947271585464478,
+      "learning_rate": 3e-05,
+      "loss": 3.2741,
+      "step": 1103
+    },
+    {
+      "epoch": 0.35797665369649806,
+      "grad_norm": 0.3342430591583252,
+      "learning_rate": 3e-05,
+      "loss": 2.1088,
+      "step": 1104
+    },
+    {
+      "epoch": 0.35830090791180286,
+      "grad_norm": 0.4716241657733917,
+      "learning_rate": 3e-05,
+      "loss": 3.1407,
+      "step": 1105
+    },
+    {
+      "epoch": 0.35862516212710766,
+      "grad_norm": 0.4366692900657654,
+      "learning_rate": 3e-05,
+      "loss": 2.5613,
+      "step": 1106
+    },
+    {
+      "epoch": 0.35894941634241245,
+      "grad_norm": 0.5341940522193909,
+      "learning_rate": 3e-05,
+      "loss": 2.4217,
+      "step": 1107
+    },
+    {
+      "epoch": 0.35927367055771725,
+      "grad_norm": 0.3775080144405365,
+      "learning_rate": 3e-05,
+      "loss": 2.6518,
+      "step": 1108
+    },
+    {
+      "epoch": 0.35959792477302205,
+      "grad_norm": 0.3918530344963074,
+      "learning_rate": 3e-05,
+      "loss": 2.495,
+      "step": 1109
+    },
+    {
+      "epoch": 0.35992217898832685,
+      "grad_norm": 0.665174126625061,
+      "learning_rate": 3e-05,
+      "loss": 3.1565,
+      "step": 1110
+    },
+    {
+      "epoch": 0.36024643320363164,
+      "grad_norm": 0.3993547856807709,
+      "learning_rate": 3e-05,
+      "loss": 2.5792,
+      "step": 1111
+    },
+    {
+      "epoch": 0.36057068741893644,
+      "grad_norm": 0.5850374698638916,
+      "learning_rate": 3e-05,
+      "loss": 3.298,
+      "step": 1112
+    },
+    {
+      "epoch": 0.36089494163424124,
+      "grad_norm": 0.526045560836792,
+      "learning_rate": 3e-05,
+      "loss": 3.0676,
+      "step": 1113
+    },
+    {
+      "epoch": 0.36121919584954604,
+      "grad_norm": 0.42036351561546326,
+      "learning_rate": 3e-05,
+      "loss": 2.5101,
+      "step": 1114
+    },
+    {
+      "epoch": 0.36154345006485084,
+      "grad_norm": 0.5715998411178589,
+      "learning_rate": 3e-05,
+      "loss": 2.8431,
+      "step": 1115
+    },
+    {
+      "epoch": 0.36186770428015563,
+      "grad_norm": 0.3066345751285553,
+      "learning_rate": 3e-05,
+      "loss": 2.2358,
+      "step": 1116
+    },
+    {
+      "epoch": 0.36219195849546043,
+      "grad_norm": 0.7204238772392273,
+      "learning_rate": 3e-05,
+      "loss": 3.0858,
+      "step": 1117
+    },
+    {
+      "epoch": 0.36251621271076523,
+      "grad_norm": 0.7254467010498047,
+      "learning_rate": 3e-05,
+      "loss": 3.4572,
+      "step": 1118
+    },
+    {
+      "epoch": 0.36284046692607,
+      "grad_norm": 0.5289313793182373,
+      "learning_rate": 3e-05,
+      "loss": 3.1484,
+      "step": 1119
+    },
+    {
+      "epoch": 0.3631647211413748,
+      "grad_norm": 0.5013135075569153,
+      "learning_rate": 3e-05,
+      "loss": 2.9647,
+      "step": 1120
+    },
+    {
+      "epoch": 0.3634889753566796,
+      "grad_norm": 0.4979635775089264,
+      "learning_rate": 3e-05,
+      "loss": 2.6101,
+      "step": 1121
+    },
+    {
+      "epoch": 0.3638132295719844,
+      "grad_norm": 0.3286263942718506,
+      "learning_rate": 3e-05,
+      "loss": 2.2395,
+      "step": 1122
+    },
+    {
+      "epoch": 0.3641374837872892,
+      "grad_norm": 0.6337347626686096,
+      "learning_rate": 3e-05,
+      "loss": 2.8806,
+      "step": 1123
+    },
+    {
+      "epoch": 0.364461738002594,
+      "grad_norm": 0.44174328446388245,
+      "learning_rate": 3e-05,
+      "loss": 2.3623,
+      "step": 1124
+    },
+    {
+      "epoch": 0.3647859922178988,
+      "grad_norm": 0.4823744297027588,
+      "learning_rate": 3e-05,
+      "loss": 2.9009,
+      "step": 1125
+    },
+    {
+      "epoch": 0.3651102464332036,
+      "grad_norm": 0.3640483319759369,
+      "learning_rate": 3e-05,
+      "loss": 2.2467,
+      "step": 1126
+    },
+    {
+      "epoch": 0.3654345006485084,
+      "grad_norm": 0.5577488541603088,
+      "learning_rate": 3e-05,
+      "loss": 2.7575,
+      "step": 1127
+    },
+    {
+      "epoch": 0.3657587548638132,
+      "grad_norm": 0.7975355386734009,
+      "learning_rate": 3e-05,
+      "loss": 3.6062,
+      "step": 1128
+    },
+    {
+      "epoch": 0.366083009079118,
+      "grad_norm": 0.39164331555366516,
+      "learning_rate": 3e-05,
+      "loss": 2.2235,
+      "step": 1129
+    },
+    {
+      "epoch": 0.3664072632944228,
+      "grad_norm": 0.5401096940040588,
+      "learning_rate": 3e-05,
+      "loss": 3.1331,
+      "step": 1130
+    },
+    {
+      "epoch": 0.3667315175097276,
+      "grad_norm": 0.38144341111183167,
+      "learning_rate": 3e-05,
+      "loss": 2.1808,
+      "step": 1131
+    },
+    {
+      "epoch": 0.3670557717250324,
+      "grad_norm": 0.4602196216583252,
+      "learning_rate": 3e-05,
+      "loss": 3.0369,
+      "step": 1132
+    },
+    {
+      "epoch": 0.36738002594033725,
+      "grad_norm": 0.4089469015598297,
+      "learning_rate": 3e-05,
+      "loss": 2.6529,
+      "step": 1133
+    },
+    {
+      "epoch": 0.36770428015564205,
+      "grad_norm": 0.47341832518577576,
+      "learning_rate": 3e-05,
+      "loss": 3.026,
+      "step": 1134
+    },
+    {
+      "epoch": 0.36802853437094685,
+      "grad_norm": 0.5747660994529724,
+      "learning_rate": 3e-05,
+      "loss": 2.9182,
+      "step": 1135
+    },
+    {
+      "epoch": 0.36835278858625164,
+      "grad_norm": 0.685478687286377,
+      "learning_rate": 3e-05,
+      "loss": 3.3285,
+      "step": 1136
+    },
+    {
+      "epoch": 0.36867704280155644,
+      "grad_norm": 0.3735675513744354,
+      "learning_rate": 3e-05,
+      "loss": 2.463,
+      "step": 1137
+    },
+    {
+      "epoch": 0.36900129701686124,
+      "grad_norm": 0.4058416187763214,
+      "learning_rate": 3e-05,
+      "loss": 2.5977,
+      "step": 1138
+    },
+    {
+      "epoch": 0.36932555123216604,
+      "grad_norm": 0.49682101607322693,
+      "learning_rate": 3e-05,
+      "loss": 3.0208,
+      "step": 1139
+    },
+    {
+      "epoch": 0.36964980544747084,
+      "grad_norm": 0.49674996733665466,
+      "learning_rate": 3e-05,
+      "loss": 3.1694,
+      "step": 1140
+    },
+    {
+      "epoch": 0.36997405966277563,
+      "grad_norm": 0.4904869794845581,
+      "learning_rate": 3e-05,
+      "loss": 2.9541,
+      "step": 1141
+    },
+    {
+      "epoch": 0.37029831387808043,
+      "grad_norm": 0.46412914991378784,
+      "learning_rate": 3e-05,
+      "loss": 2.4984,
+      "step": 1142
+    },
+    {
+      "epoch": 0.37062256809338523,
+      "grad_norm": 0.787523627281189,
+      "learning_rate": 3e-05,
+      "loss": 3.3755,
+      "step": 1143
+    },
+    {
+      "epoch": 0.37094682230869,
+      "grad_norm": 0.48954299092292786,
+      "learning_rate": 3e-05,
+      "loss": 2.9016,
+      "step": 1144
+    },
+    {
+      "epoch": 0.3712710765239948,
+      "grad_norm": 0.7552998661994934,
+      "learning_rate": 3e-05,
+      "loss": 3.2479,
+      "step": 1145
+    },
+    {
+      "epoch": 0.3715953307392996,
+      "grad_norm": 0.389386385679245,
+      "learning_rate": 3e-05,
+      "loss": 2.3852,
+      "step": 1146
+    },
+    {
+      "epoch": 0.3719195849546044,
+      "grad_norm": 0.5825744271278381,
+      "learning_rate": 3e-05,
+      "loss": 3.0039,
+      "step": 1147
+    },
+    {
+      "epoch": 0.3722438391699092,
+      "grad_norm": 0.33788254857063293,
+      "learning_rate": 3e-05,
+      "loss": 2.4419,
+      "step": 1148
+    },
+    {
+      "epoch": 0.372568093385214,
+      "grad_norm": 0.8153560161590576,
+      "learning_rate": 3e-05,
+      "loss": 3.088,
+      "step": 1149
+    },
+    {
+      "epoch": 0.3728923476005188,
+      "grad_norm": 0.6026387810707092,
+      "learning_rate": 3e-05,
+      "loss": 3.0117,
+      "step": 1150
+    },
+    {
+      "epoch": 0.3732166018158236,
+      "grad_norm": 0.4858814477920532,
+      "learning_rate": 3e-05,
+      "loss": 3.0646,
+      "step": 1151
+    },
+    {
+      "epoch": 0.3735408560311284,
+      "grad_norm": 0.6753942370414734,
+      "learning_rate": 3e-05,
+      "loss": 2.8571,
+      "step": 1152
+    },
+    {
+      "epoch": 0.3738651102464332,
+      "grad_norm": 0.5363312363624573,
+      "learning_rate": 3e-05,
+      "loss": 2.9143,
+      "step": 1153
+    },
+    {
+      "epoch": 0.374189364461738,
+      "grad_norm": 0.6066983938217163,
+      "learning_rate": 3e-05,
+      "loss": 2.9015,
+      "step": 1154
+    },
+    {
+      "epoch": 0.3745136186770428,
+      "grad_norm": 0.403451532125473,
+      "learning_rate": 3e-05,
+      "loss": 2.5729,
+      "step": 1155
+    },
+    {
+      "epoch": 0.3748378728923476,
+      "grad_norm": 0.36861416697502136,
+      "learning_rate": 3e-05,
+      "loss": 2.3866,
+      "step": 1156
+    },
+    {
+      "epoch": 0.3751621271076524,
+      "grad_norm": 0.35353830456733704,
+      "learning_rate": 3e-05,
+      "loss": 2.411,
+      "step": 1157
+    },
+    {
+      "epoch": 0.3754863813229572,
+      "grad_norm": 0.4445866048336029,
+      "learning_rate": 3e-05,
+      "loss": 2.3595,
+      "step": 1158
+    },
+    {
+      "epoch": 0.375810635538262,
+      "grad_norm": 0.36166608333587646,
+      "learning_rate": 3e-05,
+      "loss": 2.4702,
+      "step": 1159
+    },
+    {
+      "epoch": 0.3761348897535668,
+      "grad_norm": 0.40041977167129517,
+      "learning_rate": 3e-05,
+      "loss": 2.2538,
+      "step": 1160
+    },
+    {
+      "epoch": 0.3764591439688716,
+      "grad_norm": 0.6481019258499146,
+      "learning_rate": 3e-05,
+      "loss": 2.9463,
+      "step": 1161
+    },
+    {
+      "epoch": 0.3767833981841764,
+      "grad_norm": 0.3755876421928406,
+      "learning_rate": 3e-05,
+      "loss": 2.447,
+      "step": 1162
+    },
+    {
+      "epoch": 0.3771076523994812,
+      "grad_norm": 0.4395991563796997,
+      "learning_rate": 3e-05,
+      "loss": 2.7674,
+      "step": 1163
+    },
+    {
+      "epoch": 0.377431906614786,
+      "grad_norm": 0.4455874562263489,
+      "learning_rate": 3e-05,
+      "loss": 2.7883,
+      "step": 1164
+    },
+    {
+      "epoch": 0.3777561608300908,
+      "grad_norm": 0.4750497341156006,
+      "learning_rate": 3e-05,
+      "loss": 3.0573,
+      "step": 1165
+    },
+    {
+      "epoch": 0.3780804150453956,
+      "grad_norm": 0.5970101356506348,
+      "learning_rate": 3e-05,
+      "loss": 3.2215,
+      "step": 1166
+    },
+    {
+      "epoch": 0.3784046692607004,
+      "grad_norm": 0.42842766642570496,
+      "learning_rate": 3e-05,
+      "loss": 2.5102,
+      "step": 1167
+    },
+    {
+      "epoch": 0.3787289234760052,
+      "grad_norm": 0.4478355348110199,
+      "learning_rate": 3e-05,
+      "loss": 2.7123,
+      "step": 1168
+    },
+    {
+      "epoch": 0.37905317769131,
+      "grad_norm": 0.4109838306903839,
+      "learning_rate": 3e-05,
+      "loss": 2.5863,
+      "step": 1169
+    },
+    {
+      "epoch": 0.37937743190661477,
+      "grad_norm": 0.40779000520706177,
+      "learning_rate": 3e-05,
+      "loss": 2.6776,
+      "step": 1170
+    },
+    {
+      "epoch": 0.37970168612191957,
+      "grad_norm": 0.6468441486358643,
+      "learning_rate": 3e-05,
+      "loss": 2.8764,
+      "step": 1171
+    },
+    {
+      "epoch": 0.38002594033722437,
+      "grad_norm": 0.4315808415412903,
+      "learning_rate": 3e-05,
+      "loss": 2.8712,
+      "step": 1172
+    },
+    {
+      "epoch": 0.38035019455252916,
+      "grad_norm": 0.3235749304294586,
+      "learning_rate": 3e-05,
+      "loss": 2.0483,
+      "step": 1173
+    },
+    {
+      "epoch": 0.38067444876783396,
+      "grad_norm": 0.5061773061752319,
+      "learning_rate": 3e-05,
+      "loss": 2.9846,
+      "step": 1174
+    },
+    {
+      "epoch": 0.38099870298313876,
+      "grad_norm": 0.3318590819835663,
+      "learning_rate": 3e-05,
+      "loss": 2.4555,
+      "step": 1175
+    },
+    {
+      "epoch": 0.38132295719844356,
+      "grad_norm": 0.39517441391944885,
+      "learning_rate": 3e-05,
+      "loss": 2.7271,
+      "step": 1176
+    },
+    {
+      "epoch": 0.38164721141374836,
+      "grad_norm": 0.30384594202041626,
+      "learning_rate": 3e-05,
+      "loss": 2.1948,
+      "step": 1177
+    },
+    {
+      "epoch": 0.38197146562905315,
+      "grad_norm": 0.4711206555366516,
+      "learning_rate": 3e-05,
+      "loss": 2.7144,
+      "step": 1178
+    },
+    {
+      "epoch": 0.38229571984435795,
+      "grad_norm": 0.3744785189628601,
+      "learning_rate": 3e-05,
+      "loss": 2.1661,
+      "step": 1179
+    },
+    {
+      "epoch": 0.38261997405966275,
+      "grad_norm": 0.694517195224762,
+      "learning_rate": 3e-05,
+      "loss": 3.2496,
+      "step": 1180
+    },
+    {
+      "epoch": 0.3829442282749676,
+      "grad_norm": 0.4082977771759033,
+      "learning_rate": 3e-05,
+      "loss": 2.9685,
+      "step": 1181
+    },
+    {
+      "epoch": 0.3832684824902724,
+      "grad_norm": 0.5835530161857605,
+      "learning_rate": 3e-05,
+      "loss": 3.6929,
+      "step": 1182
+    },
+    {
+      "epoch": 0.3835927367055772,
+      "grad_norm": 0.35771310329437256,
+      "learning_rate": 3e-05,
+      "loss": 2.2912,
+      "step": 1183
+    },
+    {
+      "epoch": 0.383916990920882,
+      "grad_norm": 0.4869777262210846,
+      "learning_rate": 3e-05,
+      "loss": 2.5066,
+      "step": 1184
+    },
+    {
+      "epoch": 0.3842412451361868,
+      "grad_norm": 0.4702158272266388,
+      "learning_rate": 3e-05,
+      "loss": 2.7082,
+      "step": 1185
+    },
+    {
+      "epoch": 0.3845654993514916,
+      "grad_norm": 0.5295785665512085,
+      "learning_rate": 3e-05,
+      "loss": 3.3733,
+      "step": 1186
+    },
+    {
+      "epoch": 0.3848897535667964,
+      "grad_norm": 0.7204185128211975,
+      "learning_rate": 3e-05,
+      "loss": 3.4639,
+      "step": 1187
+    },
+    {
+      "epoch": 0.3852140077821012,
+      "grad_norm": 0.37787601351737976,
+      "learning_rate": 3e-05,
+      "loss": 2.6276,
+      "step": 1188
+    },
+    {
+      "epoch": 0.385538261997406,
+      "grad_norm": 0.47816357016563416,
+      "learning_rate": 3e-05,
+      "loss": 2.5893,
+      "step": 1189
+    },
+    {
+      "epoch": 0.3858625162127108,
+      "grad_norm": 0.5038324594497681,
+      "learning_rate": 3e-05,
+      "loss": 2.7321,
+      "step": 1190
+    },
+    {
+      "epoch": 0.3861867704280156,
+      "grad_norm": 0.4702116549015045,
+      "learning_rate": 3e-05,
+      "loss": 2.8524,
+      "step": 1191
+    },
+    {
+      "epoch": 0.3865110246433204,
+      "grad_norm": 0.3695956766605377,
+      "learning_rate": 3e-05,
+      "loss": 2.4769,
+      "step": 1192
+    },
+    {
+      "epoch": 0.3868352788586252,
+      "grad_norm": 0.562861442565918,
+      "learning_rate": 3e-05,
+      "loss": 3.0819,
+      "step": 1193
+    },
+    {
+      "epoch": 0.38715953307393,
+      "grad_norm": 0.3139483630657196,
+      "learning_rate": 3e-05,
+      "loss": 2.3152,
+      "step": 1194
+    },
+    {
+      "epoch": 0.38748378728923477,
+      "grad_norm": 0.34880053997039795,
+      "learning_rate": 3e-05,
+      "loss": 2.4476,
+      "step": 1195
+    },
+    {
+      "epoch": 0.38780804150453957,
+      "grad_norm": 0.40673908591270447,
+      "learning_rate": 3e-05,
+      "loss": 2.5512,
+      "step": 1196
+    },
+    {
+      "epoch": 0.38813229571984437,
+      "grad_norm": 0.3829664885997772,
+      "learning_rate": 3e-05,
+      "loss": 2.6248,
+      "step": 1197
+    },
+    {
+      "epoch": 0.38845654993514916,
+      "grad_norm": 0.39750412106513977,
+      "learning_rate": 3e-05,
+      "loss": 2.6482,
+      "step": 1198
+    },
+    {
+      "epoch": 0.38878080415045396,
+      "grad_norm": 0.5255913138389587,
+      "learning_rate": 3e-05,
+      "loss": 2.8366,
+      "step": 1199
+    },
+    {
+      "epoch": 0.38910505836575876,
+      "grad_norm": 0.3051549196243286,
+      "learning_rate": 3e-05,
+      "loss": 2.2319,
+      "step": 1200
+    },
+    {
+      "epoch": 0.38942931258106356,
+      "grad_norm": 0.49127820134162903,
+      "learning_rate": 3e-05,
+      "loss": 3.1835,
+      "step": 1201
+    },
+    {
+      "epoch": 0.38975356679636836,
+      "grad_norm": 0.6422721743583679,
+      "learning_rate": 3e-05,
+      "loss": 2.6301,
+      "step": 1202
+    },
+    {
+      "epoch": 0.39007782101167315,
+      "grad_norm": 0.5232359170913696,
+      "learning_rate": 3e-05,
+      "loss": 3.4633,
+      "step": 1203
+    },
+    {
+      "epoch": 0.39040207522697795,
+      "grad_norm": 0.41936224699020386,
+      "learning_rate": 3e-05,
+      "loss": 2.4517,
+      "step": 1204
+    },
+    {
+      "epoch": 0.39072632944228275,
+      "grad_norm": 0.4282357692718506,
+      "learning_rate": 3e-05,
+      "loss": 2.372,
+      "step": 1205
+    },
+    {
+      "epoch": 0.39105058365758755,
+      "grad_norm": 0.43828555941581726,
+      "learning_rate": 3e-05,
+      "loss": 2.6144,
+      "step": 1206
+    },
+    {
+      "epoch": 0.39137483787289234,
+      "grad_norm": 0.44104626774787903,
+      "learning_rate": 3e-05,
+      "loss": 2.4558,
+      "step": 1207
+    },
+    {
+      "epoch": 0.39169909208819714,
+      "grad_norm": 0.37094348669052124,
+      "learning_rate": 3e-05,
+      "loss": 2.5248,
+      "step": 1208
+    },
+    {
+      "epoch": 0.39202334630350194,
+      "grad_norm": 0.3644539415836334,
+      "learning_rate": 3e-05,
+      "loss": 2.6154,
+      "step": 1209
+    },
+    {
+      "epoch": 0.39234760051880674,
+      "grad_norm": 0.45678427815437317,
+      "learning_rate": 3e-05,
+      "loss": 2.9957,
+      "step": 1210
+    },
+    {
+      "epoch": 0.39267185473411154,
+      "grad_norm": 0.46496909856796265,
+      "learning_rate": 3e-05,
+      "loss": 2.681,
+      "step": 1211
+    },
+    {
+      "epoch": 0.39299610894941633,
+      "grad_norm": 0.46573999524116516,
+      "learning_rate": 3e-05,
+      "loss": 2.4703,
+      "step": 1212
+    },
+    {
+      "epoch": 0.39332036316472113,
+      "grad_norm": 0.3165050446987152,
+      "learning_rate": 3e-05,
+      "loss": 2.2791,
+      "step": 1213
+    },
+    {
+      "epoch": 0.39364461738002593,
+      "grad_norm": 1.0312039852142334,
+      "learning_rate": 3e-05,
+      "loss": 2.9006,
+      "step": 1214
+    },
+    {
+      "epoch": 0.3939688715953307,
+      "grad_norm": 0.5054008960723877,
+      "learning_rate": 3e-05,
+      "loss": 3.2303,
+      "step": 1215
+    },
+    {
+      "epoch": 0.3942931258106355,
+      "grad_norm": 0.38189128041267395,
+      "learning_rate": 3e-05,
+      "loss": 2.4339,
+      "step": 1216
+    },
+    {
+      "epoch": 0.3946173800259403,
+      "grad_norm": 0.5029484629631042,
+      "learning_rate": 3e-05,
+      "loss": 2.7584,
+      "step": 1217
+    },
+    {
+      "epoch": 0.3949416342412451,
+      "grad_norm": 0.4837813377380371,
+      "learning_rate": 3e-05,
+      "loss": 2.6057,
+      "step": 1218
+    },
+    {
+      "epoch": 0.3952658884565499,
+      "grad_norm": 0.46959659457206726,
+      "learning_rate": 3e-05,
+      "loss": 3.1883,
+      "step": 1219
+    },
+    {
+      "epoch": 0.3955901426718547,
+      "grad_norm": 0.4483717381954193,
+      "learning_rate": 3e-05,
+      "loss": 2.6358,
+      "step": 1220
+    },
+    {
+      "epoch": 0.3959143968871595,
+      "grad_norm": 0.6217966079711914,
+      "learning_rate": 3e-05,
+      "loss": 3.3714,
+      "step": 1221
+    },
+    {
+      "epoch": 0.3962386511024643,
+      "grad_norm": 0.35853639245033264,
+      "learning_rate": 3e-05,
+      "loss": 2.3294,
+      "step": 1222
+    },
+    {
+      "epoch": 0.3965629053177691,
+      "grad_norm": 0.3663962185382843,
+      "learning_rate": 3e-05,
+      "loss": 2.6091,
+      "step": 1223
+    },
+    {
+      "epoch": 0.3968871595330739,
+      "grad_norm": 0.439182311296463,
+      "learning_rate": 3e-05,
+      "loss": 2.7324,
+      "step": 1224
+    },
+    {
+      "epoch": 0.3972114137483787,
+      "grad_norm": 0.3938923478126526,
+      "learning_rate": 3e-05,
+      "loss": 2.4912,
+      "step": 1225
+    },
+    {
+      "epoch": 0.3975356679636835,
+      "grad_norm": 0.5637237429618835,
+      "learning_rate": 3e-05,
+      "loss": 2.8629,
+      "step": 1226
+    },
+    {
+      "epoch": 0.3978599221789883,
+      "grad_norm": 0.4105498194694519,
+      "learning_rate": 3e-05,
+      "loss": 2.8985,
+      "step": 1227
+    },
+    {
+      "epoch": 0.3981841763942931,
+      "grad_norm": 0.37009793519973755,
+      "learning_rate": 3e-05,
+      "loss": 2.2248,
+      "step": 1228
+    },
+    {
+      "epoch": 0.39850843060959795,
+      "grad_norm": 0.40683454275131226,
+      "learning_rate": 3e-05,
+      "loss": 2.7436,
+      "step": 1229
+    },
+    {
+      "epoch": 0.39883268482490275,
+      "grad_norm": 0.757985532283783,
+      "learning_rate": 3e-05,
+      "loss": 3.8614,
+      "step": 1230
+    },
+    {
+      "epoch": 0.39915693904020755,
+      "grad_norm": 0.5932989120483398,
+      "learning_rate": 3e-05,
+      "loss": 3.0287,
+      "step": 1231
+    },
+    {
+      "epoch": 0.39948119325551235,
+      "grad_norm": 0.2971871793270111,
+      "learning_rate": 3e-05,
+      "loss": 1.8914,
+      "step": 1232
+    },
+    {
+      "epoch": 0.39980544747081714,
+      "grad_norm": 0.5163520574569702,
+      "learning_rate": 3e-05,
+      "loss": 3.4371,
+      "step": 1233
+    },
+    {
+      "epoch": 0.40012970168612194,
+      "grad_norm": 0.3348129093647003,
+      "learning_rate": 3e-05,
+      "loss": 2.3299,
+      "step": 1234
+    },
+    {
+      "epoch": 0.40045395590142674,
+      "grad_norm": 0.7140020132064819,
+      "learning_rate": 3e-05,
+      "loss": 3.3482,
+      "step": 1235
+    },
+    {
+      "epoch": 0.40077821011673154,
+      "grad_norm": 0.4553595185279846,
+      "learning_rate": 3e-05,
+      "loss": 2.463,
+      "step": 1236
+    },
+    {
+      "epoch": 0.40110246433203633,
+      "grad_norm": 0.7101486325263977,
+      "learning_rate": 3e-05,
+      "loss": 3.0836,
+      "step": 1237
+    },
+    {
+      "epoch": 0.40142671854734113,
+      "grad_norm": 0.4686880111694336,
+      "learning_rate": 3e-05,
+      "loss": 2.713,
+      "step": 1238
+    },
+    {
+      "epoch": 0.40175097276264593,
+      "grad_norm": 0.4260486364364624,
+      "learning_rate": 3e-05,
+      "loss": 2.7954,
+      "step": 1239
+    },
+    {
+      "epoch": 0.40207522697795073,
+      "grad_norm": 0.42883118987083435,
+      "learning_rate": 3e-05,
+      "loss": 2.9867,
+      "step": 1240
+    },
+    {
+      "epoch": 0.4023994811932555,
+      "grad_norm": 0.7001662254333496,
+      "learning_rate": 3e-05,
+      "loss": 3.3524,
+      "step": 1241
+    },
+    {
+      "epoch": 0.4027237354085603,
+      "grad_norm": 0.4103189706802368,
+      "learning_rate": 3e-05,
+      "loss": 2.5026,
+      "step": 1242
+    },
+    {
+      "epoch": 0.4030479896238651,
+      "grad_norm": 0.4714316129684448,
+      "learning_rate": 3e-05,
+      "loss": 2.5942,
+      "step": 1243
+    },
+    {
+      "epoch": 0.4033722438391699,
+      "grad_norm": 0.4474298357963562,
+      "learning_rate": 3e-05,
+      "loss": 2.4,
+      "step": 1244
+    },
+    {
+      "epoch": 0.4036964980544747,
+      "grad_norm": 0.333217054605484,
+      "learning_rate": 3e-05,
+      "loss": 2.213,
+      "step": 1245
+    },
+    {
+      "epoch": 0.4040207522697795,
+      "grad_norm": 0.7456501722335815,
+      "learning_rate": 3e-05,
+      "loss": 3.4724,
+      "step": 1246
+    },
+    {
+      "epoch": 0.4043450064850843,
+      "grad_norm": 0.5679290890693665,
+      "learning_rate": 3e-05,
+      "loss": 2.5526,
+      "step": 1247
+    },
+    {
+      "epoch": 0.4046692607003891,
+      "grad_norm": 0.29385557770729065,
+      "learning_rate": 3e-05,
+      "loss": 2.0063,
+      "step": 1248
+    },
+    {
+      "epoch": 0.4049935149156939,
+      "grad_norm": 0.7953370809555054,
+      "learning_rate": 3e-05,
+      "loss": 3.8684,
+      "step": 1249
+    },
+    {
+      "epoch": 0.4053177691309987,
+      "grad_norm": 0.47878435254096985,
+      "learning_rate": 3e-05,
+      "loss": 3.0572,
+      "step": 1250
+    },
+    {
+      "epoch": 0.4056420233463035,
+      "grad_norm": 0.47467169165611267,
+      "learning_rate": 3e-05,
+      "loss": 2.7855,
+      "step": 1251
+    },
+    {
+      "epoch": 0.4059662775616083,
+      "grad_norm": 0.8314107656478882,
+      "learning_rate": 3e-05,
+      "loss": 3.2953,
+      "step": 1252
+    },
+    {
+      "epoch": 0.4062905317769131,
+      "grad_norm": 0.401339590549469,
+      "learning_rate": 3e-05,
+      "loss": 2.3345,
+      "step": 1253
+    },
+    {
+      "epoch": 0.4066147859922179,
+      "grad_norm": 0.33110007643699646,
+      "learning_rate": 3e-05,
+      "loss": 2.0508,
+      "step": 1254
+    },
+    {
+      "epoch": 0.4069390402075227,
+      "grad_norm": 0.43991991877555847,
+      "learning_rate": 3e-05,
+      "loss": 2.2924,
+      "step": 1255
+    },
+    {
+      "epoch": 0.4072632944228275,
+      "grad_norm": 0.38228699564933777,
+      "learning_rate": 3e-05,
+      "loss": 2.5407,
+      "step": 1256
+    },
+    {
+      "epoch": 0.4075875486381323,
+      "grad_norm": 0.3687131404876709,
+      "learning_rate": 3e-05,
+      "loss": 2.6156,
+      "step": 1257
+    },
+    {
+      "epoch": 0.4079118028534371,
+      "grad_norm": 0.3883078098297119,
+      "learning_rate": 3e-05,
+      "loss": 2.3502,
+      "step": 1258
+    },
+    {
+      "epoch": 0.4082360570687419,
+      "grad_norm": 0.5462496876716614,
+      "learning_rate": 3e-05,
+      "loss": 2.835,
+      "step": 1259
+    },
+    {
+      "epoch": 0.4085603112840467,
+      "grad_norm": 0.3653942048549652,
+      "learning_rate": 3e-05,
+      "loss": 2.2981,
+      "step": 1260
+    },
+    {
+      "epoch": 0.4088845654993515,
+      "grad_norm": 0.4542219638824463,
+      "learning_rate": 3e-05,
+      "loss": 2.5139,
+      "step": 1261
+    },
+    {
+      "epoch": 0.4092088197146563,
+      "grad_norm": 0.3420630395412445,
+      "learning_rate": 3e-05,
+      "loss": 2.4067,
+      "step": 1262
+    },
+    {
+      "epoch": 0.4095330739299611,
+      "grad_norm": 0.5145528316497803,
+      "learning_rate": 3e-05,
+      "loss": 2.5782,
+      "step": 1263
+    },
+    {
+      "epoch": 0.4098573281452659,
+      "grad_norm": 0.4506787061691284,
+      "learning_rate": 3e-05,
+      "loss": 2.8229,
+      "step": 1264
+    },
+    {
+      "epoch": 0.4101815823605707,
+      "grad_norm": 0.6560398936271667,
+      "learning_rate": 3e-05,
+      "loss": 3.5327,
+      "step": 1265
+    },
+    {
+      "epoch": 0.41050583657587547,
+      "grad_norm": 0.5179259777069092,
+      "learning_rate": 3e-05,
+      "loss": 3.0605,
+      "step": 1266
+    },
+    {
+      "epoch": 0.41083009079118027,
+      "grad_norm": 0.6469601392745972,
+      "learning_rate": 3e-05,
+      "loss": 3.3562,
+      "step": 1267
+    },
+    {
+      "epoch": 0.41115434500648507,
+      "grad_norm": 0.5125694274902344,
+      "learning_rate": 3e-05,
+      "loss": 2.6768,
+      "step": 1268
+    },
+    {
+      "epoch": 0.41147859922178986,
+      "grad_norm": 0.4260328710079193,
+      "learning_rate": 3e-05,
+      "loss": 2.1707,
+      "step": 1269
+    },
+    {
+      "epoch": 0.41180285343709466,
+      "grad_norm": 0.3886091411113739,
+      "learning_rate": 3e-05,
+      "loss": 2.638,
+      "step": 1270
+    },
+    {
+      "epoch": 0.41212710765239946,
+      "grad_norm": 0.5818714499473572,
+      "learning_rate": 3e-05,
+      "loss": 2.8095,
+      "step": 1271
+    },
+    {
+      "epoch": 0.41245136186770426,
+      "grad_norm": 0.35401269793510437,
+      "learning_rate": 3e-05,
+      "loss": 2.4448,
+      "step": 1272
+    },
+    {
+      "epoch": 0.41277561608300906,
+      "grad_norm": 0.5642257928848267,
+      "learning_rate": 3e-05,
+      "loss": 2.7551,
+      "step": 1273
+    },
+    {
+      "epoch": 0.41309987029831385,
+      "grad_norm": 0.4301935136318207,
+      "learning_rate": 3e-05,
+      "loss": 2.6733,
+      "step": 1274
+    },
+    {
+      "epoch": 0.41342412451361865,
+      "grad_norm": 0.362395316362381,
+      "learning_rate": 3e-05,
+      "loss": 2.2983,
+      "step": 1275
+    },
+    {
+      "epoch": 0.41374837872892345,
+      "grad_norm": 0.7523649334907532,
+      "learning_rate": 3e-05,
+      "loss": 3.4924,
+      "step": 1276
+    },
+    {
+      "epoch": 0.4140726329442283,
+      "grad_norm": 0.4860130250453949,
+      "learning_rate": 3e-05,
+      "loss": 2.6089,
+      "step": 1277
+    },
+    {
+      "epoch": 0.4143968871595331,
+      "grad_norm": 0.3518575131893158,
+      "learning_rate": 3e-05,
+      "loss": 2.4822,
+      "step": 1278
+    },
+    {
+      "epoch": 0.4147211413748379,
+      "grad_norm": 0.4253925085067749,
+      "learning_rate": 3e-05,
+      "loss": 2.7645,
+      "step": 1279
+    },
+    {
+      "epoch": 0.4150453955901427,
+      "grad_norm": 0.4553059935569763,
+      "learning_rate": 3e-05,
+      "loss": 2.9169,
+      "step": 1280
+    },
+    {
+      "epoch": 0.4153696498054475,
+      "grad_norm": 0.3683117926120758,
+      "learning_rate": 3e-05,
+      "loss": 2.4195,
+      "step": 1281
+    },
+    {
+      "epoch": 0.4156939040207523,
+      "grad_norm": 0.35151320695877075,
+      "learning_rate": 3e-05,
+      "loss": 2.4472,
+      "step": 1282
+    },
+    {
+      "epoch": 0.4160181582360571,
+      "grad_norm": 0.3002657890319824,
+      "learning_rate": 3e-05,
+      "loss": 2.2218,
+      "step": 1283
+    },
+    {
+      "epoch": 0.4163424124513619,
+      "grad_norm": 0.5835521817207336,
+      "learning_rate": 3e-05,
+      "loss": 2.7724,
+      "step": 1284
+    },
+    {
+      "epoch": 0.4166666666666667,
+      "grad_norm": 0.4063780903816223,
+      "learning_rate": 3e-05,
+      "loss": 2.7307,
+      "step": 1285
+    },
+    {
+      "epoch": 0.4169909208819715,
+      "grad_norm": 0.5606931447982788,
+      "learning_rate": 3e-05,
+      "loss": 3.5915,
+      "step": 1286
+    },
+    {
+      "epoch": 0.4173151750972763,
+      "grad_norm": 0.3228316009044647,
+      "learning_rate": 3e-05,
+      "loss": 2.1989,
+      "step": 1287
+    },
+    {
+      "epoch": 0.4176394293125811,
+      "grad_norm": 0.4255701005458832,
+      "learning_rate": 3e-05,
+      "loss": 2.3091,
+      "step": 1288
+    },
+    {
+      "epoch": 0.4179636835278859,
+      "grad_norm": 0.6979694366455078,
+      "learning_rate": 3e-05,
+      "loss": 3.143,
+      "step": 1289
+    },
+    {
+      "epoch": 0.4182879377431907,
+      "grad_norm": 0.48234227299690247,
+      "learning_rate": 3e-05,
+      "loss": 2.7816,
+      "step": 1290
+    },
+    {
+      "epoch": 0.41861219195849547,
+      "grad_norm": 0.5596135258674622,
+      "learning_rate": 3e-05,
+      "loss": 3.1807,
+      "step": 1291
+    },
+    {
+      "epoch": 0.41893644617380027,
+      "grad_norm": 0.6686919927597046,
+      "learning_rate": 3e-05,
+      "loss": 3.3306,
+      "step": 1292
+    },
+    {
+      "epoch": 0.41926070038910507,
+      "grad_norm": 0.40640145540237427,
+      "learning_rate": 3e-05,
+      "loss": 2.8603,
+      "step": 1293
+    },
+    {
+      "epoch": 0.41958495460440987,
+      "grad_norm": 0.6944930553436279,
+      "learning_rate": 3e-05,
+      "loss": 3.3913,
+      "step": 1294
+    },
+    {
+      "epoch": 0.41990920881971466,
+      "grad_norm": 0.30683672428131104,
+      "learning_rate": 3e-05,
+      "loss": 2.1408,
+      "step": 1295
+    },
+    {
+      "epoch": 0.42023346303501946,
+      "grad_norm": 0.4688516855239868,
+      "learning_rate": 3e-05,
+      "loss": 3.0321,
+      "step": 1296
+    },
+    {
+      "epoch": 0.42055771725032426,
+      "grad_norm": 0.471914678812027,
+      "learning_rate": 3e-05,
+      "loss": 3.0188,
+      "step": 1297
+    },
+    {
+      "epoch": 0.42088197146562906,
+      "grad_norm": 0.5236289501190186,
+      "learning_rate": 3e-05,
+      "loss": 2.6374,
+      "step": 1298
+    },
+    {
+      "epoch": 0.42120622568093385,
+      "grad_norm": 0.3588879704475403,
+      "learning_rate": 3e-05,
+      "loss": 2.483,
+      "step": 1299
+    },
+    {
+      "epoch": 0.42153047989623865,
+      "grad_norm": 0.8393073678016663,
+      "learning_rate": 3e-05,
+      "loss": 2.993,
+      "step": 1300
+    },
+    {
+      "epoch": 0.42185473411154345,
+      "grad_norm": 0.4818573296070099,
+      "learning_rate": 3e-05,
+      "loss": 2.9426,
+      "step": 1301
+    },
+    {
+      "epoch": 0.42217898832684825,
+      "grad_norm": 0.4233144223690033,
+      "learning_rate": 3e-05,
+      "loss": 3.1478,
+      "step": 1302
+    },
+    {
+      "epoch": 0.42250324254215305,
+      "grad_norm": 0.43216243386268616,
+      "learning_rate": 3e-05,
+      "loss": 2.8966,
+      "step": 1303
+    },
+    {
+      "epoch": 0.42282749675745784,
+      "grad_norm": 0.44994592666625977,
+      "learning_rate": 3e-05,
+      "loss": 2.8186,
+      "step": 1304
+    },
+    {
+      "epoch": 0.42315175097276264,
+      "grad_norm": 0.4815821349620819,
+      "learning_rate": 3e-05,
+      "loss": 2.6149,
+      "step": 1305
+    },
+    {
+      "epoch": 0.42347600518806744,
+      "grad_norm": 0.4435069262981415,
+      "learning_rate": 3e-05,
+      "loss": 2.8004,
+      "step": 1306
+    },
+    {
+      "epoch": 0.42380025940337224,
+      "grad_norm": 0.43064776062965393,
+      "learning_rate": 3e-05,
+      "loss": 2.5921,
+      "step": 1307
+    },
+    {
+      "epoch": 0.42412451361867703,
+      "grad_norm": 0.33596208691596985,
+      "learning_rate": 3e-05,
+      "loss": 2.3825,
+      "step": 1308
+    },
+    {
+      "epoch": 0.42444876783398183,
+      "grad_norm": 0.5525981783866882,
+      "learning_rate": 3e-05,
+      "loss": 3.2822,
+      "step": 1309
+    },
+    {
+      "epoch": 0.42477302204928663,
+      "grad_norm": 0.4408959746360779,
+      "learning_rate": 3e-05,
+      "loss": 2.9952,
+      "step": 1310
+    },
+    {
+      "epoch": 0.42509727626459143,
+      "grad_norm": 0.35831210017204285,
+      "learning_rate": 3e-05,
+      "loss": 2.2358,
+      "step": 1311
+    },
+    {
+      "epoch": 0.4254215304798962,
+      "grad_norm": 0.4098672568798065,
+      "learning_rate": 3e-05,
+      "loss": 2.9703,
+      "step": 1312
+    },
+    {
+      "epoch": 0.425745784695201,
+      "grad_norm": 0.6087233424186707,
+      "learning_rate": 3e-05,
+      "loss": 2.783,
+      "step": 1313
+    },
+    {
+      "epoch": 0.4260700389105058,
+      "grad_norm": 0.40153101086616516,
+      "learning_rate": 3e-05,
+      "loss": 2.3813,
+      "step": 1314
+    },
+    {
+      "epoch": 0.4263942931258106,
+      "grad_norm": 0.41631171107292175,
+      "learning_rate": 3e-05,
+      "loss": 2.7475,
+      "step": 1315
+    },
+    {
+      "epoch": 0.4267185473411154,
+      "grad_norm": 0.31214582920074463,
+      "learning_rate": 3e-05,
+      "loss": 1.8072,
+      "step": 1316
+    },
+    {
+      "epoch": 0.4270428015564202,
+      "grad_norm": 0.40959328413009644,
+      "learning_rate": 3e-05,
+      "loss": 2.4605,
+      "step": 1317
+    },
+    {
+      "epoch": 0.427367055771725,
+      "grad_norm": 0.47141560912132263,
+      "learning_rate": 3e-05,
+      "loss": 2.9569,
+      "step": 1318
+    },
+    {
+      "epoch": 0.4276913099870298,
+      "grad_norm": 0.5316165685653687,
+      "learning_rate": 3e-05,
+      "loss": 3.0977,
+      "step": 1319
+    },
+    {
+      "epoch": 0.4280155642023346,
+      "grad_norm": 0.38060620427131653,
+      "learning_rate": 3e-05,
+      "loss": 2.5191,
+      "step": 1320
+    },
+    {
+      "epoch": 0.4283398184176394,
+      "grad_norm": 0.608597457408905,
+      "learning_rate": 3e-05,
+      "loss": 2.9659,
+      "step": 1321
+    },
+    {
+      "epoch": 0.4286640726329442,
+      "grad_norm": 0.5037148594856262,
+      "learning_rate": 3e-05,
+      "loss": 2.7705,
+      "step": 1322
+    },
+    {
+      "epoch": 0.428988326848249,
+      "grad_norm": 0.4762224853038788,
+      "learning_rate": 3e-05,
+      "loss": 2.961,
+      "step": 1323
+    },
+    {
+      "epoch": 0.4293125810635538,
+      "grad_norm": 0.46620506048202515,
+      "learning_rate": 3e-05,
+      "loss": 3.139,
+      "step": 1324
+    },
+    {
+      "epoch": 0.4296368352788586,
+      "grad_norm": 0.5525588989257812,
+      "learning_rate": 3e-05,
+      "loss": 3.1544,
+      "step": 1325
+    },
+    {
+      "epoch": 0.42996108949416345,
+      "grad_norm": 0.36874231696128845,
+      "learning_rate": 3e-05,
+      "loss": 2.4578,
+      "step": 1326
+    },
+    {
+      "epoch": 0.43028534370946825,
+      "grad_norm": 0.3143581449985504,
+      "learning_rate": 3e-05,
+      "loss": 2.2709,
+      "step": 1327
+    },
+    {
+      "epoch": 0.43060959792477305,
+      "grad_norm": 0.4403678774833679,
+      "learning_rate": 3e-05,
+      "loss": 2.7529,
+      "step": 1328
+    },
+    {
+      "epoch": 0.43093385214007784,
+      "grad_norm": 0.34060239791870117,
+      "learning_rate": 3e-05,
+      "loss": 2.3986,
+      "step": 1329
+    },
+    {
+      "epoch": 0.43125810635538264,
+      "grad_norm": 0.5207977294921875,
+      "learning_rate": 3e-05,
+      "loss": 2.7795,
+      "step": 1330
+    },
+    {
+      "epoch": 0.43158236057068744,
+      "grad_norm": 0.628577470779419,
+      "learning_rate": 3e-05,
+      "loss": 2.9631,
+      "step": 1331
+    },
+    {
+      "epoch": 0.43190661478599224,
+      "grad_norm": 0.5758695602416992,
+      "learning_rate": 3e-05,
+      "loss": 3.164,
+      "step": 1332
+    },
+    {
+      "epoch": 0.43223086900129704,
+      "grad_norm": 0.38502439856529236,
+      "learning_rate": 3e-05,
+      "loss": 2.3858,
+      "step": 1333
+    },
+    {
+      "epoch": 0.43255512321660183,
+      "grad_norm": 0.3623266816139221,
+      "learning_rate": 3e-05,
+      "loss": 2.2386,
+      "step": 1334
+    },
+    {
+      "epoch": 0.43287937743190663,
+      "grad_norm": 0.4279859960079193,
+      "learning_rate": 3e-05,
+      "loss": 2.5055,
+      "step": 1335
+    },
+    {
+      "epoch": 0.43320363164721143,
+      "grad_norm": 0.49045270681381226,
+      "learning_rate": 3e-05,
+      "loss": 3.2138,
+      "step": 1336
+    },
+    {
+      "epoch": 0.4335278858625162,
+      "grad_norm": 0.43378517031669617,
+      "learning_rate": 3e-05,
+      "loss": 2.3826,
+      "step": 1337
+    },
+    {
+      "epoch": 0.433852140077821,
+      "grad_norm": 0.4253461956977844,
+      "learning_rate": 3e-05,
+      "loss": 2.7932,
+      "step": 1338
+    },
+    {
+      "epoch": 0.4341763942931258,
+      "grad_norm": 0.672945499420166,
+      "learning_rate": 3e-05,
+      "loss": 3.0461,
+      "step": 1339
+    },
+    {
+      "epoch": 0.4345006485084306,
+      "grad_norm": 0.3908237814903259,
+      "learning_rate": 3e-05,
+      "loss": 2.9459,
+      "step": 1340
+    },
+    {
+      "epoch": 0.4348249027237354,
+      "grad_norm": 0.46041566133499146,
+      "learning_rate": 3e-05,
+      "loss": 3.112,
+      "step": 1341
+    },
+    {
+      "epoch": 0.4351491569390402,
+      "grad_norm": 0.5440371036529541,
+      "learning_rate": 3e-05,
+      "loss": 2.6633,
+      "step": 1342
+    },
+    {
+      "epoch": 0.435473411154345,
+      "grad_norm": 1.6437742710113525,
+      "learning_rate": 3e-05,
+      "loss": 3.0575,
+      "step": 1343
+    },
+    {
+      "epoch": 0.4357976653696498,
+      "grad_norm": 0.5014978647232056,
+      "learning_rate": 3e-05,
+      "loss": 2.9013,
+      "step": 1344
+    },
+    {
+      "epoch": 0.4361219195849546,
+      "grad_norm": 0.3617139458656311,
+      "learning_rate": 3e-05,
+      "loss": 2.3277,
+      "step": 1345
+    },
+    {
+      "epoch": 0.4364461738002594,
+      "grad_norm": 0.5884495377540588,
+      "learning_rate": 3e-05,
+      "loss": 3.4448,
+      "step": 1346
+    },
+    {
+      "epoch": 0.4367704280155642,
+      "grad_norm": 0.3666825294494629,
+      "learning_rate": 3e-05,
+      "loss": 2.4477,
+      "step": 1347
+    },
+    {
+      "epoch": 0.437094682230869,
+      "grad_norm": 0.5433440208435059,
+      "learning_rate": 3e-05,
+      "loss": 3.0549,
+      "step": 1348
+    },
+    {
+      "epoch": 0.4374189364461738,
+      "grad_norm": 0.2934304177761078,
+      "learning_rate": 3e-05,
+      "loss": 2.1213,
+      "step": 1349
+    },
+    {
+      "epoch": 0.4377431906614786,
+      "grad_norm": 0.349470853805542,
+      "learning_rate": 3e-05,
+      "loss": 2.239,
+      "step": 1350
+    },
+    {
+      "epoch": 0.4380674448767834,
+      "grad_norm": 0.44675442576408386,
+      "learning_rate": 3e-05,
+      "loss": 2.5493,
+      "step": 1351
+    },
+    {
+      "epoch": 0.4383916990920882,
+      "grad_norm": 0.4730742871761322,
+      "learning_rate": 3e-05,
+      "loss": 2.3578,
+      "step": 1352
+    },
+    {
+      "epoch": 0.438715953307393,
+      "grad_norm": 0.45542681217193604,
+      "learning_rate": 3e-05,
+      "loss": 2.9669,
+      "step": 1353
+    },
+    {
+      "epoch": 0.4390402075226978,
+      "grad_norm": 0.5599362850189209,
+      "learning_rate": 3e-05,
+      "loss": 3.5582,
+      "step": 1354
+    },
+    {
+      "epoch": 0.4393644617380026,
+      "grad_norm": 0.5028124451637268,
+      "learning_rate": 3e-05,
+      "loss": 2.8632,
+      "step": 1355
+    },
+    {
+      "epoch": 0.4396887159533074,
+      "grad_norm": 0.47607025504112244,
+      "learning_rate": 3e-05,
+      "loss": 2.9986,
+      "step": 1356
+    },
+    {
+      "epoch": 0.4400129701686122,
+      "grad_norm": 0.3373994529247284,
+      "learning_rate": 3e-05,
+      "loss": 2.2068,
+      "step": 1357
+    },
+    {
+      "epoch": 0.440337224383917,
+      "grad_norm": 0.4815230667591095,
+      "learning_rate": 3e-05,
+      "loss": 3.3317,
+      "step": 1358
+    },
+    {
+      "epoch": 0.4406614785992218,
+      "grad_norm": 0.4606270492076874,
+      "learning_rate": 3e-05,
+      "loss": 2.3561,
+      "step": 1359
+    },
+    {
+      "epoch": 0.4409857328145266,
+      "grad_norm": 0.4815424978733063,
+      "learning_rate": 3e-05,
+      "loss": 2.4376,
+      "step": 1360
+    },
+    {
+      "epoch": 0.4413099870298314,
+      "grad_norm": 0.39359530806541443,
+      "learning_rate": 3e-05,
+      "loss": 2.4507,
+      "step": 1361
+    },
+    {
+      "epoch": 0.44163424124513617,
+      "grad_norm": 0.547265887260437,
+      "learning_rate": 3e-05,
+      "loss": 2.9728,
+      "step": 1362
+    },
+    {
+      "epoch": 0.44195849546044097,
+      "grad_norm": 0.3750714659690857,
+      "learning_rate": 3e-05,
+      "loss": 2.8336,
+      "step": 1363
+    },
+    {
+      "epoch": 0.44228274967574577,
+      "grad_norm": 0.5661170482635498,
+      "learning_rate": 3e-05,
+      "loss": 2.7491,
+      "step": 1364
+    },
+    {
+      "epoch": 0.44260700389105057,
+      "grad_norm": 0.6730261445045471,
+      "learning_rate": 3e-05,
+      "loss": 3.4711,
+      "step": 1365
+    },
+    {
+      "epoch": 0.44293125810635536,
+      "grad_norm": 0.3979703485965729,
+      "learning_rate": 3e-05,
+      "loss": 2.7666,
+      "step": 1366
+    },
+    {
+      "epoch": 0.44325551232166016,
+      "grad_norm": 0.33864128589630127,
+      "learning_rate": 3e-05,
+      "loss": 2.2496,
+      "step": 1367
+    },
+    {
+      "epoch": 0.44357976653696496,
+      "grad_norm": 0.4510902464389801,
+      "learning_rate": 3e-05,
+      "loss": 2.8811,
+      "step": 1368
+    },
+    {
+      "epoch": 0.44390402075226976,
+      "grad_norm": 0.4215259552001953,
+      "learning_rate": 3e-05,
+      "loss": 2.7451,
+      "step": 1369
+    },
+    {
+      "epoch": 0.44422827496757455,
+      "grad_norm": 0.7172505855560303,
+      "learning_rate": 3e-05,
+      "loss": 3.2969,
+      "step": 1370
+    },
+    {
+      "epoch": 0.44455252918287935,
+      "grad_norm": 0.3875070810317993,
+      "learning_rate": 3e-05,
+      "loss": 2.7454,
+      "step": 1371
+    },
+    {
+      "epoch": 0.44487678339818415,
+      "grad_norm": 0.34435129165649414,
+      "learning_rate": 3e-05,
+      "loss": 2.5205,
+      "step": 1372
+    },
+    {
+      "epoch": 0.44520103761348895,
+      "grad_norm": 0.5346459150314331,
+      "learning_rate": 3e-05,
+      "loss": 2.6797,
+      "step": 1373
+    },
+    {
+      "epoch": 0.4455252918287938,
+      "grad_norm": 0.8478299379348755,
+      "learning_rate": 3e-05,
+      "loss": 2.9691,
+      "step": 1374
+    },
+    {
+      "epoch": 0.4458495460440986,
+      "grad_norm": 0.39494001865386963,
+      "learning_rate": 3e-05,
+      "loss": 2.2573,
+      "step": 1375
+    },
+    {
+      "epoch": 0.4461738002594034,
+      "grad_norm": 0.9179035425186157,
+      "learning_rate": 3e-05,
+      "loss": 3.4616,
+      "step": 1376
+    },
+    {
+      "epoch": 0.4464980544747082,
+      "grad_norm": 0.4647447466850281,
+      "learning_rate": 3e-05,
+      "loss": 2.4352,
+      "step": 1377
+    },
+    {
+      "epoch": 0.446822308690013,
+      "grad_norm": 0.7937042713165283,
+      "learning_rate": 3e-05,
+      "loss": 3.4843,
+      "step": 1378
+    },
+    {
+      "epoch": 0.4471465629053178,
+      "grad_norm": 0.7792840600013733,
+      "learning_rate": 3e-05,
+      "loss": 2.9194,
+      "step": 1379
+    },
+    {
+      "epoch": 0.4474708171206226,
+      "grad_norm": 0.8572617173194885,
+      "learning_rate": 3e-05,
+      "loss": 3.1852,
+      "step": 1380
+    },
+    {
+      "epoch": 0.4477950713359274,
+      "grad_norm": 0.4945402443408966,
+      "learning_rate": 3e-05,
+      "loss": 3.0933,
+      "step": 1381
+    },
+    {
+      "epoch": 0.4481193255512322,
+      "grad_norm": 0.6436623930931091,
+      "learning_rate": 3e-05,
+      "loss": 3.1908,
+      "step": 1382
+    },
+    {
+      "epoch": 0.448443579766537,
+      "grad_norm": 0.5378843545913696,
+      "learning_rate": 3e-05,
+      "loss": 2.6296,
+      "step": 1383
+    },
+    {
+      "epoch": 0.4487678339818418,
+      "grad_norm": 0.7814305424690247,
+      "learning_rate": 3e-05,
+      "loss": 2.6397,
+      "step": 1384
+    },
+    {
+      "epoch": 0.4490920881971466,
+      "grad_norm": 1.4011073112487793,
+      "learning_rate": 3e-05,
+      "loss": 3.4079,
+      "step": 1385
+    },
+    {
+      "epoch": 0.4494163424124514,
+      "grad_norm": 0.5338678956031799,
+      "learning_rate": 3e-05,
+      "loss": 2.7638,
+      "step": 1386
+    },
+    {
+      "epoch": 0.4497405966277562,
+      "grad_norm": 0.40464258193969727,
+      "learning_rate": 3e-05,
+      "loss": 2.9796,
+      "step": 1387
+    },
+    {
+      "epoch": 0.45006485084306097,
+      "grad_norm": 0.7425667643547058,
+      "learning_rate": 3e-05,
+      "loss": 3.1244,
+      "step": 1388
+    },
+    {
+      "epoch": 0.45038910505836577,
+      "grad_norm": 1.0161622762680054,
+      "learning_rate": 3e-05,
+      "loss": 3.0155,
+      "step": 1389
+    },
+    {
+      "epoch": 0.45071335927367057,
+      "grad_norm": 0.638938844203949,
+      "learning_rate": 3e-05,
+      "loss": 2.655,
+      "step": 1390
+    },
+    {
+      "epoch": 0.45103761348897536,
+      "grad_norm": 0.49087902903556824,
+      "learning_rate": 3e-05,
+      "loss": 3.0173,
+      "step": 1391
+    },
+    {
+      "epoch": 0.45136186770428016,
+      "grad_norm": 0.4924505650997162,
+      "learning_rate": 3e-05,
+      "loss": 2.8628,
+      "step": 1392
+    },
+    {
+      "epoch": 0.45168612191958496,
+      "grad_norm": 0.36778524518013,
+      "learning_rate": 3e-05,
+      "loss": 2.396,
+      "step": 1393
+    },
+    {
+      "epoch": 0.45201037613488976,
+      "grad_norm": 0.42188313603401184,
+      "learning_rate": 3e-05,
+      "loss": 2.4605,
+      "step": 1394
+    },
+    {
+      "epoch": 0.45233463035019456,
+      "grad_norm": 0.38293787837028503,
+      "learning_rate": 3e-05,
+      "loss": 2.3092,
+      "step": 1395
+    },
+    {
+      "epoch": 0.45265888456549935,
+      "grad_norm": 1.4900380373001099,
+      "learning_rate": 3e-05,
+      "loss": 3.5263,
+      "step": 1396
+    },
+    {
+      "epoch": 0.45298313878080415,
+      "grad_norm": 0.4871014356613159,
+      "learning_rate": 3e-05,
+      "loss": 3.2936,
+      "step": 1397
+    },
+    {
+      "epoch": 0.45330739299610895,
+      "grad_norm": 0.3634713888168335,
+      "learning_rate": 3e-05,
+      "loss": 2.4907,
+      "step": 1398
+    },
+    {
+      "epoch": 0.45363164721141375,
+      "grad_norm": 0.48739272356033325,
+      "learning_rate": 3e-05,
+      "loss": 2.5709,
+      "step": 1399
+    },
+    {
+      "epoch": 0.45395590142671854,
+      "grad_norm": 0.6873285174369812,
+      "learning_rate": 3e-05,
+      "loss": 2.5855,
+      "step": 1400
+    },
+    {
+      "epoch": 0.45428015564202334,
+      "grad_norm": 0.6144210696220398,
+      "learning_rate": 3e-05,
+      "loss": 2.2754,
+      "step": 1401
+    },
+    {
+      "epoch": 0.45460440985732814,
+      "grad_norm": 0.5977807641029358,
+      "learning_rate": 3e-05,
+      "loss": 3.3163,
+      "step": 1402
+    },
+    {
+      "epoch": 0.45492866407263294,
+      "grad_norm": 0.6249608993530273,
+      "learning_rate": 3e-05,
+      "loss": 3.6663,
+      "step": 1403
+    },
+    {
+      "epoch": 0.45525291828793774,
+      "grad_norm": 0.3463802635669708,
+      "learning_rate": 3e-05,
+      "loss": 2.2947,
+      "step": 1404
+    },
+    {
+      "epoch": 0.45557717250324253,
+      "grad_norm": 0.5028194189071655,
+      "learning_rate": 3e-05,
+      "loss": 2.7745,
+      "step": 1405
+    },
+    {
+      "epoch": 0.45590142671854733,
+      "grad_norm": 0.6974214911460876,
+      "learning_rate": 3e-05,
+      "loss": 2.8927,
+      "step": 1406
+    },
+    {
+      "epoch": 0.45622568093385213,
+      "grad_norm": 0.39795762300491333,
+      "learning_rate": 3e-05,
+      "loss": 2.1744,
+      "step": 1407
+    },
+    {
+      "epoch": 0.4565499351491569,
+      "grad_norm": 0.5468930602073669,
+      "learning_rate": 3e-05,
+      "loss": 2.71,
+      "step": 1408
+    },
+    {
+      "epoch": 0.4568741893644617,
+      "grad_norm": 0.610824465751648,
+      "learning_rate": 3e-05,
+      "loss": 2.7361,
+      "step": 1409
+    },
+    {
+      "epoch": 0.4571984435797665,
+      "grad_norm": 0.8177030086517334,
+      "learning_rate": 3e-05,
+      "loss": 3.3845,
+      "step": 1410
+    },
+    {
+      "epoch": 0.4575226977950713,
+      "grad_norm": 0.5949739217758179,
+      "learning_rate": 3e-05,
+      "loss": 2.8594,
+      "step": 1411
+    },
+    {
+      "epoch": 0.4578469520103761,
+      "grad_norm": 0.4874008297920227,
+      "learning_rate": 3e-05,
+      "loss": 2.6241,
+      "step": 1412
+    },
+    {
+      "epoch": 0.4581712062256809,
+      "grad_norm": 0.584190845489502,
+      "learning_rate": 3e-05,
+      "loss": 2.4912,
+      "step": 1413
+    },
+    {
+      "epoch": 0.4584954604409857,
+      "grad_norm": 0.731478214263916,
+      "learning_rate": 3e-05,
+      "loss": 2.6471,
+      "step": 1414
+    },
+    {
+      "epoch": 0.4588197146562905,
+      "grad_norm": 0.5334786772727966,
+      "learning_rate": 3e-05,
+      "loss": 2.6216,
+      "step": 1415
+    },
+    {
+      "epoch": 0.4591439688715953,
+      "grad_norm": 0.34439969062805176,
+      "learning_rate": 3e-05,
+      "loss": 2.5622,
+      "step": 1416
+    },
+    {
+      "epoch": 0.4594682230869001,
+      "grad_norm": 0.6263260245323181,
+      "learning_rate": 3e-05,
+      "loss": 2.2789,
+      "step": 1417
+    },
+    {
+      "epoch": 0.4597924773022049,
+      "grad_norm": 0.9479245543479919,
+      "learning_rate": 3e-05,
+      "loss": 3.3231,
+      "step": 1418
+    },
+    {
+      "epoch": 0.4601167315175097,
+      "grad_norm": 0.7419570684432983,
+      "learning_rate": 3e-05,
+      "loss": 3.4357,
+      "step": 1419
+    },
+    {
+      "epoch": 0.4604409857328145,
+      "grad_norm": 0.373055636882782,
+      "learning_rate": 3e-05,
+      "loss": 2.4995,
+      "step": 1420
+    },
+    {
+      "epoch": 0.4607652399481193,
+      "grad_norm": 0.4185406267642975,
+      "learning_rate": 3e-05,
+      "loss": 2.9419,
+      "step": 1421
+    },
+    {
+      "epoch": 0.46108949416342415,
+      "grad_norm": 0.6370330452919006,
+      "learning_rate": 3e-05,
+      "loss": 2.8992,
+      "step": 1422
+    },
+    {
+      "epoch": 0.46141374837872895,
+      "grad_norm": 0.535131573677063,
+      "learning_rate": 3e-05,
+      "loss": 2.7543,
+      "step": 1423
+    },
+    {
+      "epoch": 0.46173800259403375,
+      "grad_norm": 0.8282862305641174,
+      "learning_rate": 3e-05,
+      "loss": 2.9977,
+      "step": 1424
+    },
+    {
+      "epoch": 0.46206225680933855,
+      "grad_norm": 0.404869019985199,
+      "learning_rate": 3e-05,
+      "loss": 2.7734,
+      "step": 1425
+    },
+    {
+      "epoch": 0.46238651102464334,
+      "grad_norm": 0.4121834933757782,
+      "learning_rate": 3e-05,
+      "loss": 2.5102,
+      "step": 1426
+    },
+    {
+      "epoch": 0.46271076523994814,
+      "grad_norm": 0.38208016753196716,
+      "learning_rate": 3e-05,
+      "loss": 2.5539,
+      "step": 1427
+    },
+    {
+      "epoch": 0.46303501945525294,
+      "grad_norm": 0.7146673202514648,
+      "learning_rate": 3e-05,
+      "loss": 2.9726,
+      "step": 1428
+    },
+    {
+      "epoch": 0.46335927367055774,
+      "grad_norm": 0.4644511640071869,
+      "learning_rate": 3e-05,
+      "loss": 2.6881,
+      "step": 1429
+    },
+    {
+      "epoch": 0.46368352788586253,
+      "grad_norm": 0.2972213923931122,
+      "learning_rate": 3e-05,
+      "loss": 2.1091,
+      "step": 1430
+    },
+    {
+      "epoch": 0.46400778210116733,
+      "grad_norm": 0.43376424908638,
+      "learning_rate": 3e-05,
+      "loss": 2.4048,
+      "step": 1431
+    },
+    {
+      "epoch": 0.46433203631647213,
+      "grad_norm": 0.821345329284668,
+      "learning_rate": 3e-05,
+      "loss": 3.1541,
+      "step": 1432
+    },
+    {
+      "epoch": 0.4646562905317769,
+      "grad_norm": 0.3683786690235138,
+      "learning_rate": 3e-05,
+      "loss": 2.3332,
+      "step": 1433
+    },
+    {
+      "epoch": 0.4649805447470817,
+      "grad_norm": 0.4221295118331909,
+      "learning_rate": 3e-05,
+      "loss": 2.5774,
+      "step": 1434
+    },
+    {
+      "epoch": 0.4653047989623865,
+      "grad_norm": 0.3510437309741974,
+      "learning_rate": 3e-05,
+      "loss": 2.4986,
+      "step": 1435
+    },
+    {
+      "epoch": 0.4656290531776913,
+      "grad_norm": 0.2961505353450775,
+      "learning_rate": 3e-05,
+      "loss": 1.8969,
+      "step": 1436
+    },
+    {
+      "epoch": 0.4659533073929961,
+      "grad_norm": 1.047551155090332,
+      "learning_rate": 3e-05,
+      "loss": 3.7049,
+      "step": 1437
+    },
+    {
+      "epoch": 0.4662775616083009,
+      "grad_norm": 0.38310396671295166,
+      "learning_rate": 3e-05,
+      "loss": 2.6552,
+      "step": 1438
+    },
+    {
+      "epoch": 0.4666018158236057,
+      "grad_norm": 0.4215133786201477,
+      "learning_rate": 3e-05,
+      "loss": 2.9461,
+      "step": 1439
+    },
+    {
+      "epoch": 0.4669260700389105,
+      "grad_norm": 0.4853152632713318,
+      "learning_rate": 3e-05,
+      "loss": 2.9026,
+      "step": 1440
+    },
+    {
+      "epoch": 0.4672503242542153,
+      "grad_norm": 0.462440550327301,
+      "learning_rate": 3e-05,
+      "loss": 3.0153,
+      "step": 1441
+    },
+    {
+      "epoch": 0.4675745784695201,
+      "grad_norm": 0.5902437567710876,
+      "learning_rate": 3e-05,
+      "loss": 3.3771,
+      "step": 1442
+    },
+    {
+      "epoch": 0.4678988326848249,
+      "grad_norm": 0.36659523844718933,
+      "learning_rate": 3e-05,
+      "loss": 2.5558,
+      "step": 1443
+    },
+    {
+      "epoch": 0.4682230869001297,
+      "grad_norm": 0.4044749140739441,
+      "learning_rate": 3e-05,
+      "loss": 2.5599,
+      "step": 1444
+    },
+    {
+      "epoch": 0.4685473411154345,
+      "grad_norm": 0.6034001708030701,
+      "learning_rate": 3e-05,
+      "loss": 3.0598,
+      "step": 1445
+    },
+    {
+      "epoch": 0.4688715953307393,
+      "grad_norm": 0.6464588642120361,
+      "learning_rate": 3e-05,
+      "loss": 3.2781,
+      "step": 1446
+    },
+    {
+      "epoch": 0.4691958495460441,
+      "grad_norm": 0.4417009949684143,
+      "learning_rate": 3e-05,
+      "loss": 2.5014,
+      "step": 1447
+    },
+    {
+      "epoch": 0.4695201037613489,
+      "grad_norm": 0.39509180188179016,
+      "learning_rate": 3e-05,
+      "loss": 2.4572,
+      "step": 1448
+    },
+    {
+      "epoch": 0.4698443579766537,
+      "grad_norm": 0.5031090378761292,
+      "learning_rate": 3e-05,
+      "loss": 3.0305,
+      "step": 1449
+    },
+    {
+      "epoch": 0.4701686121919585,
+      "grad_norm": 0.4663790464401245,
+      "learning_rate": 3e-05,
+      "loss": 2.4601,
+      "step": 1450
+    },
+    {
+      "epoch": 0.4704928664072633,
+      "grad_norm": 0.3405059278011322,
+      "learning_rate": 3e-05,
+      "loss": 2.2562,
+      "step": 1451
+    },
+    {
+      "epoch": 0.4708171206225681,
+      "grad_norm": 0.39420798420906067,
+      "learning_rate": 3e-05,
+      "loss": 2.6245,
+      "step": 1452
+    },
+    {
+      "epoch": 0.4711413748378729,
+      "grad_norm": 0.4079931676387787,
+      "learning_rate": 3e-05,
+      "loss": 2.5249,
+      "step": 1453
+    },
+    {
+      "epoch": 0.4714656290531777,
+      "grad_norm": 0.3166867196559906,
+      "learning_rate": 3e-05,
+      "loss": 2.0335,
+      "step": 1454
+    },
+    {
+      "epoch": 0.4717898832684825,
+      "grad_norm": 0.5444930195808411,
+      "learning_rate": 3e-05,
+      "loss": 2.6228,
+      "step": 1455
+    },
+    {
+      "epoch": 0.4721141374837873,
+      "grad_norm": 0.6555557250976562,
+      "learning_rate": 3e-05,
+      "loss": 2.8901,
+      "step": 1456
+    },
+    {
+      "epoch": 0.4724383916990921,
+      "grad_norm": 0.4597679078578949,
+      "learning_rate": 3e-05,
+      "loss": 2.9277,
+      "step": 1457
+    },
+    {
+      "epoch": 0.4727626459143969,
+      "grad_norm": 0.32781898975372314,
+      "learning_rate": 3e-05,
+      "loss": 2.2405,
+      "step": 1458
+    },
+    {
+      "epoch": 0.47308690012970167,
+      "grad_norm": 0.5637408494949341,
+      "learning_rate": 3e-05,
+      "loss": 3.1885,
+      "step": 1459
+    },
+    {
+      "epoch": 0.47341115434500647,
+      "grad_norm": 0.3839605748653412,
+      "learning_rate": 3e-05,
+      "loss": 2.4895,
+      "step": 1460
+    },
+    {
+      "epoch": 0.47373540856031127,
+      "grad_norm": 0.4510838985443115,
+      "learning_rate": 3e-05,
+      "loss": 2.806,
+      "step": 1461
+    },
+    {
+      "epoch": 0.47405966277561606,
+      "grad_norm": 0.4191063344478607,
+      "learning_rate": 3e-05,
+      "loss": 2.4798,
+      "step": 1462
+    },
+    {
+      "epoch": 0.47438391699092086,
+      "grad_norm": 0.3906472325325012,
+      "learning_rate": 3e-05,
+      "loss": 2.4179,
+      "step": 1463
+    },
+    {
+      "epoch": 0.47470817120622566,
+      "grad_norm": 0.5385026931762695,
+      "learning_rate": 3e-05,
+      "loss": 2.9937,
+      "step": 1464
+    },
+    {
+      "epoch": 0.47503242542153046,
+      "grad_norm": 0.4429912269115448,
+      "learning_rate": 3e-05,
+      "loss": 2.7959,
+      "step": 1465
+    },
+    {
+      "epoch": 0.47535667963683526,
+      "grad_norm": 0.32905665040016174,
+      "learning_rate": 3e-05,
+      "loss": 2.3311,
+      "step": 1466
+    },
+    {
+      "epoch": 0.47568093385214005,
+      "grad_norm": 0.4660322666168213,
+      "learning_rate": 3e-05,
+      "loss": 2.5318,
+      "step": 1467
+    },
+    {
+      "epoch": 0.47600518806744485,
+      "grad_norm": 0.49521404504776,
+      "learning_rate": 3e-05,
+      "loss": 3.5317,
+      "step": 1468
+    },
+    {
+      "epoch": 0.47632944228274965,
+      "grad_norm": 0.3586205840110779,
+      "learning_rate": 3e-05,
+      "loss": 2.6325,
+      "step": 1469
+    },
+    {
+      "epoch": 0.4766536964980545,
+      "grad_norm": 0.49939578771591187,
+      "learning_rate": 3e-05,
+      "loss": 2.9014,
+      "step": 1470
+    },
+    {
+      "epoch": 0.4769779507133593,
+      "grad_norm": 0.3470568060874939,
+      "learning_rate": 3e-05,
+      "loss": 2.4612,
+      "step": 1471
+    },
+    {
+      "epoch": 0.4773022049286641,
+      "grad_norm": 0.6364990472793579,
+      "learning_rate": 3e-05,
+      "loss": 3.0098,
+      "step": 1472
+    },
+    {
+      "epoch": 0.4776264591439689,
+      "grad_norm": 0.4464481472969055,
+      "learning_rate": 3e-05,
+      "loss": 2.9066,
+      "step": 1473
+    },
+    {
+      "epoch": 0.4779507133592737,
+      "grad_norm": 0.35837432742118835,
+      "learning_rate": 3e-05,
+      "loss": 2.4751,
+      "step": 1474
+    },
+    {
+      "epoch": 0.4782749675745785,
+      "grad_norm": 0.7242171764373779,
+      "learning_rate": 3e-05,
+      "loss": 3.0895,
+      "step": 1475
+    },
+    {
+      "epoch": 0.4785992217898833,
+      "grad_norm": 0.480854332447052,
+      "learning_rate": 3e-05,
+      "loss": 2.9142,
+      "step": 1476
+    },
+    {
+      "epoch": 0.4789234760051881,
+      "grad_norm": 0.3646979033946991,
+      "learning_rate": 3e-05,
+      "loss": 2.5176,
+      "step": 1477
+    },
+    {
+      "epoch": 0.4792477302204929,
+      "grad_norm": 0.3305288553237915,
+      "learning_rate": 3e-05,
+      "loss": 2.3094,
+      "step": 1478
+    },
+    {
+      "epoch": 0.4795719844357977,
+      "grad_norm": 0.37247246503829956,
+      "learning_rate": 3e-05,
+      "loss": 2.4714,
+      "step": 1479
+    },
+    {
+      "epoch": 0.4798962386511025,
+      "grad_norm": 0.5258364677429199,
+      "learning_rate": 3e-05,
+      "loss": 2.6068,
+      "step": 1480
+    },
+    {
+      "epoch": 0.4802204928664073,
+      "grad_norm": 0.4988524317741394,
+      "learning_rate": 3e-05,
+      "loss": 2.9439,
+      "step": 1481
+    },
+    {
+      "epoch": 0.4805447470817121,
+      "grad_norm": 0.3806118369102478,
+      "learning_rate": 3e-05,
+      "loss": 2.5603,
+      "step": 1482
+    },
+    {
+      "epoch": 0.4808690012970169,
+      "grad_norm": 0.5143681764602661,
+      "learning_rate": 3e-05,
+      "loss": 2.6449,
+      "step": 1483
+    },
+    {
+      "epoch": 0.48119325551232167,
+      "grad_norm": 0.4168810546398163,
+      "learning_rate": 3e-05,
+      "loss": 2.7116,
+      "step": 1484
+    },
+    {
+      "epoch": 0.48151750972762647,
+      "grad_norm": 0.573697030544281,
+      "learning_rate": 3e-05,
+      "loss": 3.376,
+      "step": 1485
+    },
+    {
+      "epoch": 0.48184176394293127,
+      "grad_norm": 0.5606452226638794,
+      "learning_rate": 3e-05,
+      "loss": 2.6717,
+      "step": 1486
+    },
+    {
+      "epoch": 0.48216601815823606,
+      "grad_norm": 0.5274271965026855,
+      "learning_rate": 3e-05,
+      "loss": 2.8205,
+      "step": 1487
+    },
+    {
+      "epoch": 0.48249027237354086,
+      "grad_norm": 0.49651384353637695,
+      "learning_rate": 3e-05,
+      "loss": 3.3531,
+      "step": 1488
+    },
+    {
+      "epoch": 0.48281452658884566,
+      "grad_norm": 0.3106611371040344,
+      "learning_rate": 3e-05,
+      "loss": 2.0689,
+      "step": 1489
+    },
+    {
+      "epoch": 0.48313878080415046,
+      "grad_norm": 0.40287265181541443,
+      "learning_rate": 3e-05,
+      "loss": 2.3919,
+      "step": 1490
+    },
+    {
+      "epoch": 0.48346303501945526,
+      "grad_norm": 0.3279760777950287,
+      "learning_rate": 3e-05,
+      "loss": 2.1631,
+      "step": 1491
+    },
+    {
+      "epoch": 0.48378728923476005,
+      "grad_norm": 0.3447646200656891,
+      "learning_rate": 3e-05,
+      "loss": 2.2309,
+      "step": 1492
+    },
+    {
+      "epoch": 0.48411154345006485,
+      "grad_norm": 0.3916889727115631,
+      "learning_rate": 3e-05,
+      "loss": 2.2117,
+      "step": 1493
+    },
+    {
+      "epoch": 0.48443579766536965,
+      "grad_norm": 0.8013365268707275,
+      "learning_rate": 3e-05,
+      "loss": 3.1189,
+      "step": 1494
+    },
+    {
+      "epoch": 0.48476005188067445,
+      "grad_norm": 0.7343838214874268,
+      "learning_rate": 3e-05,
+      "loss": 3.3568,
+      "step": 1495
+    },
+    {
+      "epoch": 0.48508430609597925,
+      "grad_norm": 0.39398375153541565,
+      "learning_rate": 3e-05,
+      "loss": 2.5131,
+      "step": 1496
+    },
+    {
+      "epoch": 0.48540856031128404,
+      "grad_norm": 0.7150238156318665,
+      "learning_rate": 3e-05,
+      "loss": 3.5032,
+      "step": 1497
+    },
+    {
+      "epoch": 0.48573281452658884,
+      "grad_norm": 0.6845331788063049,
+      "learning_rate": 3e-05,
+      "loss": 2.5845,
+      "step": 1498
+    },
+    {
+      "epoch": 0.48605706874189364,
+      "grad_norm": 0.7601929306983948,
+      "learning_rate": 3e-05,
+      "loss": 3.056,
+      "step": 1499
+    },
+    {
+      "epoch": 0.48638132295719844,
+      "grad_norm": 0.44406095147132874,
+      "learning_rate": 3e-05,
+      "loss": 2.3207,
+      "step": 1500
+    },
+    {
+      "epoch": 0.48670557717250323,
+      "grad_norm": 0.7749094367027283,
+      "learning_rate": 3e-05,
+      "loss": 3.4891,
+      "step": 1501
+    },
+    {
+      "epoch": 0.48702983138780803,
+      "grad_norm": 0.37267374992370605,
+      "learning_rate": 3e-05,
+      "loss": 2.1362,
+      "step": 1502
+    },
+    {
+      "epoch": 0.48735408560311283,
+      "grad_norm": 0.5236530900001526,
+      "learning_rate": 3e-05,
+      "loss": 2.532,
+      "step": 1503
+    },
+    {
+      "epoch": 0.4876783398184176,
+      "grad_norm": 0.8633908629417419,
+      "learning_rate": 3e-05,
+      "loss": 3.4516,
+      "step": 1504
+    },
+    {
+      "epoch": 0.4880025940337224,
+      "grad_norm": 0.6296999454498291,
+      "learning_rate": 3e-05,
+      "loss": 2.8468,
+      "step": 1505
+    },
+    {
+      "epoch": 0.4883268482490272,
+      "grad_norm": 0.4389503598213196,
+      "learning_rate": 3e-05,
+      "loss": 2.643,
+      "step": 1506
+    },
+    {
+      "epoch": 0.488651102464332,
+      "grad_norm": 0.4152829051017761,
+      "learning_rate": 3e-05,
+      "loss": 2.7105,
+      "step": 1507
+    },
+    {
+      "epoch": 0.4889753566796368,
+      "grad_norm": 0.4336642622947693,
+      "learning_rate": 3e-05,
+      "loss": 2.4211,
+      "step": 1508
+    },
+    {
+      "epoch": 0.4892996108949416,
+      "grad_norm": 0.5087767243385315,
+      "learning_rate": 3e-05,
+      "loss": 2.2356,
+      "step": 1509
+    },
+    {
+      "epoch": 0.4896238651102464,
+      "grad_norm": 0.6433772444725037,
+      "learning_rate": 3e-05,
+      "loss": 2.365,
+      "step": 1510
+    },
+    {
+      "epoch": 0.4899481193255512,
+      "grad_norm": 0.4201391935348511,
+      "learning_rate": 3e-05,
+      "loss": 2.273,
+      "step": 1511
+    },
+    {
+      "epoch": 0.490272373540856,
+      "grad_norm": 0.5611066222190857,
+      "learning_rate": 3e-05,
+      "loss": 3.379,
+      "step": 1512
+    },
+    {
+      "epoch": 0.4905966277561608,
+      "grad_norm": 0.39032283425331116,
+      "learning_rate": 3e-05,
+      "loss": 2.5926,
+      "step": 1513
+    },
+    {
+      "epoch": 0.4909208819714656,
+      "grad_norm": 0.28694209456443787,
+      "learning_rate": 3e-05,
+      "loss": 2.149,
+      "step": 1514
+    },
+    {
+      "epoch": 0.4912451361867704,
+      "grad_norm": 0.7721209526062012,
+      "learning_rate": 3e-05,
+      "loss": 3.2021,
+      "step": 1515
+    },
+    {
+      "epoch": 0.4915693904020752,
+      "grad_norm": 0.42571043968200684,
+      "learning_rate": 3e-05,
+      "loss": 2.614,
+      "step": 1516
+    },
+    {
+      "epoch": 0.49189364461738,
+      "grad_norm": 0.3823404014110565,
+      "learning_rate": 3e-05,
+      "loss": 2.9804,
+      "step": 1517
+    },
+    {
+      "epoch": 0.49221789883268485,
+      "grad_norm": 0.510025143623352,
+      "learning_rate": 3e-05,
+      "loss": 3.0635,
+      "step": 1518
+    },
+    {
+      "epoch": 0.49254215304798965,
+      "grad_norm": 0.36896640062332153,
+      "learning_rate": 3e-05,
+      "loss": 2.783,
+      "step": 1519
+    },
+    {
+      "epoch": 0.49286640726329445,
+      "grad_norm": 0.3996583819389343,
+      "learning_rate": 3e-05,
+      "loss": 2.6826,
+      "step": 1520
+    },
+    {
+      "epoch": 0.49319066147859925,
+      "grad_norm": 0.3318382799625397,
+      "learning_rate": 3e-05,
+      "loss": 2.0427,
+      "step": 1521
+    },
+    {
+      "epoch": 0.49351491569390404,
+      "grad_norm": 0.5450424551963806,
+      "learning_rate": 3e-05,
+      "loss": 2.9083,
+      "step": 1522
+    },
+    {
+      "epoch": 0.49383916990920884,
+      "grad_norm": 0.4182894229888916,
+      "learning_rate": 3e-05,
+      "loss": 3.0421,
+      "step": 1523
+    },
+    {
+      "epoch": 0.49416342412451364,
+      "grad_norm": 0.5747566819190979,
+      "learning_rate": 3e-05,
+      "loss": 2.9887,
+      "step": 1524
+    },
+    {
+      "epoch": 0.49448767833981844,
+      "grad_norm": 0.584333598613739,
+      "learning_rate": 3e-05,
+      "loss": 3.0604,
+      "step": 1525
+    },
+    {
+      "epoch": 0.49481193255512324,
+      "grad_norm": 0.3816571831703186,
+      "learning_rate": 3e-05,
+      "loss": 2.3817,
+      "step": 1526
+    },
+    {
+      "epoch": 0.49513618677042803,
+      "grad_norm": 0.3406500518321991,
+      "learning_rate": 3e-05,
+      "loss": 2.5421,
+      "step": 1527
+    },
+    {
+      "epoch": 0.49546044098573283,
+      "grad_norm": 0.40909838676452637,
+      "learning_rate": 3e-05,
+      "loss": 2.7494,
+      "step": 1528
+    },
+    {
+      "epoch": 0.49578469520103763,
+      "grad_norm": 0.4781501591205597,
+      "learning_rate": 3e-05,
+      "loss": 3.1833,
+      "step": 1529
+    },
+    {
+      "epoch": 0.4961089494163424,
+      "grad_norm": 0.3463471531867981,
+      "learning_rate": 3e-05,
+      "loss": 2.4984,
+      "step": 1530
+    },
+    {
+      "epoch": 0.4964332036316472,
+      "grad_norm": 0.3697502017021179,
+      "learning_rate": 3e-05,
+      "loss": 2.7134,
+      "step": 1531
+    },
+    {
+      "epoch": 0.496757457846952,
+      "grad_norm": 0.691720724105835,
+      "learning_rate": 3e-05,
+      "loss": 3.1719,
+      "step": 1532
+    },
+    {
+      "epoch": 0.4970817120622568,
+      "grad_norm": 0.32337120175361633,
+      "learning_rate": 3e-05,
+      "loss": 2.3225,
+      "step": 1533
+    },
+    {
+      "epoch": 0.4974059662775616,
+      "grad_norm": 0.4715064465999603,
+      "learning_rate": 3e-05,
+      "loss": 3.1133,
+      "step": 1534
+    },
+    {
+      "epoch": 0.4977302204928664,
+      "grad_norm": 0.40829581022262573,
+      "learning_rate": 3e-05,
+      "loss": 2.3561,
+      "step": 1535
+    },
+    {
+      "epoch": 0.4980544747081712,
+      "grad_norm": 0.4673785865306854,
+      "learning_rate": 3e-05,
+      "loss": 3.1168,
+      "step": 1536
+    },
+    {
+      "epoch": 0.498378728923476,
+      "grad_norm": 0.36511221528053284,
+      "learning_rate": 3e-05,
+      "loss": 2.318,
+      "step": 1537
+    },
+    {
+      "epoch": 0.4987029831387808,
+      "grad_norm": 0.4440150856971741,
+      "learning_rate": 3e-05,
+      "loss": 2.8249,
+      "step": 1538
+    },
+    {
+      "epoch": 0.4990272373540856,
+      "grad_norm": 0.4282509386539459,
+      "learning_rate": 3e-05,
+      "loss": 2.5061,
+      "step": 1539
+    },
+    {
+      "epoch": 0.4993514915693904,
+      "grad_norm": 0.3973773121833801,
+      "learning_rate": 3e-05,
+      "loss": 2.7204,
+      "step": 1540
+    },
+    {
+      "epoch": 0.4996757457846952,
+      "grad_norm": 0.6284478902816772,
+      "learning_rate": 3e-05,
+      "loss": 3.6118,
+      "step": 1541
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.35517528653144836,
+      "learning_rate": 3e-05,
+      "loss": 2.1701,
+      "step": 1542
+    },
+    {
+      "epoch": 0.5003242542153048,
+      "grad_norm": 0.36391380429267883,
+      "learning_rate": 3e-05,
+      "loss": 2.2818,
+      "step": 1543
+    },
+    {
+      "epoch": 0.5006485084306096,
+      "grad_norm": 0.6255760192871094,
+      "learning_rate": 3e-05,
+      "loss": 2.7875,
+      "step": 1544
+    },
+    {
+      "epoch": 0.5009727626459144,
+      "grad_norm": 0.495789498090744,
+      "learning_rate": 3e-05,
+      "loss": 2.9377,
+      "step": 1545
+    },
+    {
+      "epoch": 0.5012970168612192,
+      "grad_norm": 0.5592718720436096,
+      "learning_rate": 3e-05,
+      "loss": 2.864,
+      "step": 1546
+    },
+    {
+      "epoch": 0.501621271076524,
+      "grad_norm": 0.5450345277786255,
+      "learning_rate": 3e-05,
+      "loss": 3.1544,
+      "step": 1547
+    },
+    {
+      "epoch": 0.5019455252918288,
+      "grad_norm": 0.6737727522850037,
+      "learning_rate": 3e-05,
+      "loss": 2.8924,
+      "step": 1548
+    },
+    {
+      "epoch": 0.5022697795071336,
+      "grad_norm": 0.4801010489463806,
+      "learning_rate": 3e-05,
+      "loss": 2.6758,
+      "step": 1549
+    },
+    {
+      "epoch": 0.5025940337224384,
+      "grad_norm": 0.3288918435573578,
+      "learning_rate": 3e-05,
+      "loss": 2.2878,
+      "step": 1550
+    },
+    {
+      "epoch": 0.5029182879377432,
+      "grad_norm": 0.37064945697784424,
+      "learning_rate": 3e-05,
+      "loss": 2.5604,
+      "step": 1551
+    },
+    {
+      "epoch": 0.503242542153048,
+      "grad_norm": 0.3913347125053406,
+      "learning_rate": 3e-05,
+      "loss": 2.1779,
+      "step": 1552
+    },
+    {
+      "epoch": 0.5035667963683528,
+      "grad_norm": 0.33827364444732666,
+      "learning_rate": 3e-05,
+      "loss": 2.5615,
+      "step": 1553
+    },
+    {
+      "epoch": 0.5038910505836576,
+      "grad_norm": 0.4642769396305084,
+      "learning_rate": 3e-05,
+      "loss": 2.4098,
+      "step": 1554
+    },
+    {
+      "epoch": 0.5042153047989624,
+      "grad_norm": 0.3588915467262268,
+      "learning_rate": 3e-05,
+      "loss": 2.6982,
+      "step": 1555
+    },
+    {
+      "epoch": 0.5045395590142672,
+      "grad_norm": 0.4696119725704193,
+      "learning_rate": 3e-05,
+      "loss": 2.7863,
+      "step": 1556
+    },
+    {
+      "epoch": 0.504863813229572,
+      "grad_norm": 0.4686877429485321,
+      "learning_rate": 3e-05,
+      "loss": 2.5368,
+      "step": 1557
+    },
+    {
+      "epoch": 0.5051880674448768,
+      "grad_norm": 0.30904966592788696,
+      "learning_rate": 3e-05,
+      "loss": 2.1912,
+      "step": 1558
+    },
+    {
+      "epoch": 0.5055123216601816,
+      "grad_norm": 0.37136951088905334,
+      "learning_rate": 3e-05,
+      "loss": 2.3522,
+      "step": 1559
+    },
+    {
+      "epoch": 0.5058365758754864,
+      "grad_norm": 0.34488239884376526,
+      "learning_rate": 3e-05,
+      "loss": 2.3018,
+      "step": 1560
+    },
+    {
+      "epoch": 0.5061608300907912,
+      "grad_norm": 0.9845595359802246,
+      "learning_rate": 3e-05,
+      "loss": 3.7051,
+      "step": 1561
+    },
+    {
+      "epoch": 0.506485084306096,
+      "grad_norm": 0.49596714973449707,
+      "learning_rate": 3e-05,
+      "loss": 2.7525,
+      "step": 1562
+    },
+    {
+      "epoch": 0.5068093385214008,
+      "grad_norm": 0.3978366255760193,
+      "learning_rate": 3e-05,
+      "loss": 2.6047,
+      "step": 1563
+    },
+    {
+      "epoch": 0.5071335927367056,
+      "grad_norm": 0.3728798031806946,
+      "learning_rate": 3e-05,
+      "loss": 2.6105,
+      "step": 1564
+    },
+    {
+      "epoch": 0.5074578469520103,
+      "grad_norm": 0.47106871008872986,
+      "learning_rate": 3e-05,
+      "loss": 2.7951,
+      "step": 1565
+    },
+    {
+      "epoch": 0.5077821011673151,
+      "grad_norm": 0.5367615818977356,
+      "learning_rate": 3e-05,
+      "loss": 3.0043,
+      "step": 1566
+    },
+    {
+      "epoch": 0.50810635538262,
+      "grad_norm": 0.4026597738265991,
+      "learning_rate": 3e-05,
+      "loss": 2.4515,
+      "step": 1567
+    },
+    {
+      "epoch": 0.5084306095979247,
+      "grad_norm": 0.5449621677398682,
+      "learning_rate": 3e-05,
+      "loss": 2.8683,
+      "step": 1568
+    },
+    {
+      "epoch": 0.5087548638132295,
+      "grad_norm": 0.401913583278656,
+      "learning_rate": 3e-05,
+      "loss": 2.7447,
+      "step": 1569
+    },
+    {
+      "epoch": 0.5090791180285343,
+      "grad_norm": 0.3578577935695648,
+      "learning_rate": 3e-05,
+      "loss": 2.4737,
+      "step": 1570
+    },
+    {
+      "epoch": 0.5094033722438391,
+      "grad_norm": 0.8233593702316284,
+      "learning_rate": 3e-05,
+      "loss": 3.7094,
+      "step": 1571
+    },
+    {
+      "epoch": 0.5097276264591439,
+      "grad_norm": 0.608200192451477,
+      "learning_rate": 3e-05,
+      "loss": 2.9583,
+      "step": 1572
+    },
+    {
+      "epoch": 0.5100518806744487,
+      "grad_norm": 0.3789304792881012,
+      "learning_rate": 3e-05,
+      "loss": 2.5961,
+      "step": 1573
+    },
+    {
+      "epoch": 0.5103761348897535,
+      "grad_norm": 0.39314743876457214,
+      "learning_rate": 3e-05,
+      "loss": 2.3075,
+      "step": 1574
+    },
+    {
+      "epoch": 0.5107003891050583,
+      "grad_norm": 0.5736052989959717,
+      "learning_rate": 3e-05,
+      "loss": 3.4523,
+      "step": 1575
+    },
+    {
+      "epoch": 0.5110246433203631,
+      "grad_norm": 0.45256417989730835,
+      "learning_rate": 3e-05,
+      "loss": 2.5069,
+      "step": 1576
+    },
+    {
+      "epoch": 0.5113488975356679,
+      "grad_norm": 0.59746915102005,
+      "learning_rate": 3e-05,
+      "loss": 2.6464,
+      "step": 1577
+    },
+    {
+      "epoch": 0.5116731517509727,
+      "grad_norm": 0.4065306484699249,
+      "learning_rate": 3e-05,
+      "loss": 2.2571,
+      "step": 1578
+    },
+    {
+      "epoch": 0.5119974059662775,
+      "grad_norm": 0.5983695983886719,
+      "learning_rate": 3e-05,
+      "loss": 3.3427,
+      "step": 1579
+    },
+    {
+      "epoch": 0.5123216601815823,
+      "grad_norm": 0.5653676390647888,
+      "learning_rate": 3e-05,
+      "loss": 3.0278,
+      "step": 1580
+    },
+    {
+      "epoch": 0.5126459143968871,
+      "grad_norm": 0.3666940927505493,
+      "learning_rate": 3e-05,
+      "loss": 2.3337,
+      "step": 1581
+    },
+    {
+      "epoch": 0.5129701686121919,
+      "grad_norm": 0.3877595365047455,
+      "learning_rate": 3e-05,
+      "loss": 2.4451,
+      "step": 1582
+    },
+    {
+      "epoch": 0.5132944228274967,
+      "grad_norm": 0.7416172623634338,
+      "learning_rate": 3e-05,
+      "loss": 2.9989,
+      "step": 1583
+    },
+    {
+      "epoch": 0.5136186770428015,
+      "grad_norm": 0.8620609045028687,
+      "learning_rate": 3e-05,
+      "loss": 3.1589,
+      "step": 1584
+    },
+    {
+      "epoch": 0.5139429312581063,
+      "grad_norm": 0.40080952644348145,
+      "learning_rate": 3e-05,
+      "loss": 2.5461,
+      "step": 1585
+    },
+    {
+      "epoch": 0.5142671854734111,
+      "grad_norm": 0.47496768832206726,
+      "learning_rate": 3e-05,
+      "loss": 3.2906,
+      "step": 1586
+    },
+    {
+      "epoch": 0.5145914396887159,
+      "grad_norm": 0.6676542162895203,
+      "learning_rate": 3e-05,
+      "loss": 3.0221,
+      "step": 1587
+    },
+    {
+      "epoch": 0.5149156939040207,
+      "grad_norm": 0.4993215501308441,
+      "learning_rate": 3e-05,
+      "loss": 2.1856,
+      "step": 1588
+    },
+    {
+      "epoch": 0.5152399481193255,
+      "grad_norm": 0.8144398331642151,
+      "learning_rate": 3e-05,
+      "loss": 3.1028,
+      "step": 1589
+    },
+    {
+      "epoch": 0.5155642023346303,
+      "grad_norm": 0.6495314836502075,
+      "learning_rate": 3e-05,
+      "loss": 3.0959,
+      "step": 1590
+    },
+    {
+      "epoch": 0.5158884565499352,
+      "grad_norm": 0.6900297999382019,
+      "learning_rate": 3e-05,
+      "loss": 2.638,
+      "step": 1591
+    },
+    {
+      "epoch": 0.51621271076524,
+      "grad_norm": 0.42997896671295166,
+      "learning_rate": 3e-05,
+      "loss": 2.9518,
+      "step": 1592
+    },
+    {
+      "epoch": 0.5165369649805448,
+      "grad_norm": 0.33789771795272827,
+      "learning_rate": 3e-05,
+      "loss": 2.2792,
+      "step": 1593
+    },
+    {
+      "epoch": 0.5168612191958496,
+      "grad_norm": 0.493896484375,
+      "learning_rate": 3e-05,
+      "loss": 2.7617,
+      "step": 1594
+    },
+    {
+      "epoch": 0.5171854734111544,
+      "grad_norm": 0.6016855239868164,
+      "learning_rate": 3e-05,
+      "loss": 2.7146,
+      "step": 1595
+    },
+    {
+      "epoch": 0.5175097276264592,
+      "grad_norm": 0.4950004518032074,
+      "learning_rate": 3e-05,
+      "loss": 2.3644,
+      "step": 1596
+    },
+    {
+      "epoch": 0.517833981841764,
+      "grad_norm": 0.557538628578186,
+      "learning_rate": 3e-05,
+      "loss": 2.4337,
+      "step": 1597
+    },
+    {
+      "epoch": 0.5181582360570688,
+      "grad_norm": 0.4362228512763977,
+      "learning_rate": 3e-05,
+      "loss": 2.6149,
+      "step": 1598
+    },
+    {
+      "epoch": 0.5184824902723736,
+      "grad_norm": 0.47829321026802063,
+      "learning_rate": 3e-05,
+      "loss": 2.4533,
+      "step": 1599
+    },
+    {
+      "epoch": 0.5188067444876784,
+      "grad_norm": 0.9308983683586121,
+      "learning_rate": 3e-05,
+      "loss": 3.8209,
+      "step": 1600
+    },
+    {
+      "epoch": 0.5191309987029832,
+      "grad_norm": 0.566008448600769,
+      "learning_rate": 3e-05,
+      "loss": 2.971,
+      "step": 1601
+    },
+    {
+      "epoch": 0.519455252918288,
+      "grad_norm": 0.5582615733146667,
+      "learning_rate": 3e-05,
+      "loss": 3.2749,
+      "step": 1602
+    },
+    {
+      "epoch": 0.5197795071335928,
+      "grad_norm": 0.4735746681690216,
+      "learning_rate": 3e-05,
+      "loss": 2.1173,
+      "step": 1603
+    },
+    {
+      "epoch": 0.5201037613488976,
+      "grad_norm": 0.4919450581073761,
+      "learning_rate": 3e-05,
+      "loss": 2.5435,
+      "step": 1604
+    },
+    {
+      "epoch": 0.5204280155642024,
+      "grad_norm": 0.3929174542427063,
+      "learning_rate": 3e-05,
+      "loss": 2.4017,
+      "step": 1605
+    },
+    {
+      "epoch": 0.5207522697795072,
+      "grad_norm": 0.3712298572063446,
+      "learning_rate": 3e-05,
+      "loss": 2.4539,
+      "step": 1606
+    },
+    {
+      "epoch": 0.521076523994812,
+      "grad_norm": 0.5985029339790344,
+      "learning_rate": 3e-05,
+      "loss": 3.2479,
+      "step": 1607
+    },
+    {
+      "epoch": 0.5214007782101168,
+      "grad_norm": 0.43268021941185,
+      "learning_rate": 3e-05,
+      "loss": 2.8035,
+      "step": 1608
+    },
+    {
+      "epoch": 0.5217250324254216,
+      "grad_norm": 0.410976380109787,
+      "learning_rate": 3e-05,
+      "loss": 2.7911,
+      "step": 1609
+    },
+    {
+      "epoch": 0.5220492866407264,
+      "grad_norm": 0.4774320125579834,
+      "learning_rate": 3e-05,
+      "loss": 2.7602,
+      "step": 1610
+    },
+    {
+      "epoch": 0.5223735408560312,
+      "grad_norm": 0.366381973028183,
+      "learning_rate": 3e-05,
+      "loss": 2.6268,
+      "step": 1611
+    },
+    {
+      "epoch": 0.522697795071336,
+      "grad_norm": 0.5201402306556702,
+      "learning_rate": 3e-05,
+      "loss": 2.4534,
+      "step": 1612
+    },
+    {
+      "epoch": 0.5230220492866408,
+      "grad_norm": 0.5599924921989441,
+      "learning_rate": 3e-05,
+      "loss": 2.4028,
+      "step": 1613
+    },
+    {
+      "epoch": 0.5233463035019456,
+      "grad_norm": 0.41001096367836,
+      "learning_rate": 3e-05,
+      "loss": 2.9518,
+      "step": 1614
+    },
+    {
+      "epoch": 0.5236705577172504,
+      "grad_norm": 0.3847190737724304,
+      "learning_rate": 3e-05,
+      "loss": 2.4178,
+      "step": 1615
+    },
+    {
+      "epoch": 0.5239948119325551,
+      "grad_norm": 0.4514461159706116,
+      "learning_rate": 3e-05,
+      "loss": 2.8661,
+      "step": 1616
+    },
+    {
+      "epoch": 0.52431906614786,
+      "grad_norm": 0.3520987629890442,
+      "learning_rate": 3e-05,
+      "loss": 2.1206,
+      "step": 1617
+    },
+    {
+      "epoch": 0.5246433203631647,
+      "grad_norm": 0.405545175075531,
+      "learning_rate": 3e-05,
+      "loss": 2.3369,
+      "step": 1618
+    },
+    {
+      "epoch": 0.5249675745784695,
+      "grad_norm": 0.43646809458732605,
+      "learning_rate": 3e-05,
+      "loss": 2.5736,
+      "step": 1619
+    },
+    {
+      "epoch": 0.5252918287937743,
+      "grad_norm": 0.4421340227127075,
+      "learning_rate": 3e-05,
+      "loss": 2.8307,
+      "step": 1620
+    },
+    {
+      "epoch": 0.5256160830090791,
+      "grad_norm": 0.3402283787727356,
+      "learning_rate": 3e-05,
+      "loss": 2.2424,
+      "step": 1621
+    },
+    {
+      "epoch": 0.5259403372243839,
+      "grad_norm": 0.3970811367034912,
+      "learning_rate": 3e-05,
+      "loss": 2.6557,
+      "step": 1622
+    },
+    {
+      "epoch": 0.5262645914396887,
+      "grad_norm": 0.7435619831085205,
+      "learning_rate": 3e-05,
+      "loss": 3.4775,
+      "step": 1623
+    },
+    {
+      "epoch": 0.5265888456549935,
+      "grad_norm": 0.4696846306324005,
+      "learning_rate": 3e-05,
+      "loss": 2.8065,
+      "step": 1624
+    },
+    {
+      "epoch": 0.5269130998702983,
+      "grad_norm": 0.4474312663078308,
+      "learning_rate": 3e-05,
+      "loss": 3.0697,
+      "step": 1625
+    },
+    {
+      "epoch": 0.5272373540856031,
+      "grad_norm": 0.5841845273971558,
+      "learning_rate": 3e-05,
+      "loss": 3.6618,
+      "step": 1626
+    },
+    {
+      "epoch": 0.5275616083009079,
+      "grad_norm": 0.47695136070251465,
+      "learning_rate": 3e-05,
+      "loss": 2.6901,
+      "step": 1627
+    },
+    {
+      "epoch": 0.5278858625162127,
+      "grad_norm": 0.31721436977386475,
+      "learning_rate": 3e-05,
+      "loss": 2.1941,
+      "step": 1628
+    },
+    {
+      "epoch": 0.5282101167315175,
+      "grad_norm": 0.4641149044036865,
+      "learning_rate": 3e-05,
+      "loss": 2.8193,
+      "step": 1629
+    },
+    {
+      "epoch": 0.5285343709468223,
+      "grad_norm": 0.4010216295719147,
+      "learning_rate": 3e-05,
+      "loss": 2.5173,
+      "step": 1630
+    },
+    {
+      "epoch": 0.5288586251621271,
+      "grad_norm": 0.3801226317882538,
+      "learning_rate": 3e-05,
+      "loss": 2.4485,
+      "step": 1631
+    },
+    {
+      "epoch": 0.5291828793774319,
+      "grad_norm": 0.4622304439544678,
+      "learning_rate": 3e-05,
+      "loss": 2.7582,
+      "step": 1632
+    },
+    {
+      "epoch": 0.5295071335927367,
+      "grad_norm": 0.30398088693618774,
+      "learning_rate": 3e-05,
+      "loss": 2.3716,
+      "step": 1633
+    },
+    {
+      "epoch": 0.5298313878080415,
+      "grad_norm": 0.3690205216407776,
+      "learning_rate": 3e-05,
+      "loss": 2.4884,
+      "step": 1634
+    },
+    {
+      "epoch": 0.5301556420233463,
+      "grad_norm": 0.3345027565956116,
+      "learning_rate": 3e-05,
+      "loss": 2.1385,
+      "step": 1635
+    },
+    {
+      "epoch": 0.5304798962386511,
+      "grad_norm": 0.8159152865409851,
+      "learning_rate": 3e-05,
+      "loss": 2.7499,
+      "step": 1636
+    },
+    {
+      "epoch": 0.5308041504539559,
+      "grad_norm": 0.42214223742485046,
+      "learning_rate": 3e-05,
+      "loss": 2.6324,
+      "step": 1637
+    },
+    {
+      "epoch": 0.5311284046692607,
+      "grad_norm": 0.8278467655181885,
+      "learning_rate": 3e-05,
+      "loss": 3.7615,
+      "step": 1638
+    },
+    {
+      "epoch": 0.5314526588845655,
+      "grad_norm": 0.38480085134506226,
+      "learning_rate": 3e-05,
+      "loss": 2.3364,
+      "step": 1639
+    },
+    {
+      "epoch": 0.5317769130998703,
+      "grad_norm": 0.7166218161582947,
+      "learning_rate": 3e-05,
+      "loss": 3.9661,
+      "step": 1640
+    },
+    {
+      "epoch": 0.5321011673151751,
+      "grad_norm": 0.5015415549278259,
+      "learning_rate": 3e-05,
+      "loss": 3.1387,
+      "step": 1641
+    },
+    {
+      "epoch": 0.5324254215304799,
+      "grad_norm": 0.5076246857643127,
+      "learning_rate": 3e-05,
+      "loss": 2.8369,
+      "step": 1642
+    },
+    {
+      "epoch": 0.5327496757457847,
+      "grad_norm": 0.6185001730918884,
+      "learning_rate": 3e-05,
+      "loss": 3.9486,
+      "step": 1643
+    },
+    {
+      "epoch": 0.5330739299610895,
+      "grad_norm": 0.44848597049713135,
+      "learning_rate": 3e-05,
+      "loss": 2.4443,
+      "step": 1644
+    },
+    {
+      "epoch": 0.5333981841763943,
+      "grad_norm": 0.3550649881362915,
+      "learning_rate": 3e-05,
+      "loss": 2.6932,
+      "step": 1645
+    },
+    {
+      "epoch": 0.5337224383916991,
+      "grad_norm": 0.4024414122104645,
+      "learning_rate": 3e-05,
+      "loss": 2.3745,
+      "step": 1646
+    },
+    {
+      "epoch": 0.5340466926070039,
+      "grad_norm": 0.4165817201137543,
+      "learning_rate": 3e-05,
+      "loss": 2.7411,
+      "step": 1647
+    },
+    {
+      "epoch": 0.5343709468223087,
+      "grad_norm": 0.42525801062583923,
+      "learning_rate": 3e-05,
+      "loss": 2.4873,
+      "step": 1648
+    },
+    {
+      "epoch": 0.5346952010376135,
+      "grad_norm": 0.45017313957214355,
+      "learning_rate": 3e-05,
+      "loss": 2.9892,
+      "step": 1649
+    },
+    {
+      "epoch": 0.5350194552529183,
+      "grad_norm": 0.3523767292499542,
+      "learning_rate": 3e-05,
+      "loss": 2.3971,
+      "step": 1650
+    },
+    {
+      "epoch": 0.5353437094682231,
+      "grad_norm": 1.7728818655014038,
+      "learning_rate": 3e-05,
+      "loss": 4.0499,
+      "step": 1651
+    },
+    {
+      "epoch": 0.5356679636835279,
+      "grad_norm": 0.6242770552635193,
+      "learning_rate": 3e-05,
+      "loss": 3.2351,
+      "step": 1652
+    },
+    {
+      "epoch": 0.5359922178988327,
+      "grad_norm": 0.35867708921432495,
+      "learning_rate": 3e-05,
+      "loss": 2.6586,
+      "step": 1653
+    },
+    {
+      "epoch": 0.5363164721141375,
+      "grad_norm": 0.5451458096504211,
+      "learning_rate": 3e-05,
+      "loss": 3.0225,
+      "step": 1654
+    },
+    {
+      "epoch": 0.5366407263294423,
+      "grad_norm": 1.045674443244934,
+      "learning_rate": 3e-05,
+      "loss": 4.2454,
+      "step": 1655
+    },
+    {
+      "epoch": 0.5369649805447471,
+      "grad_norm": 0.5935481190681458,
+      "learning_rate": 3e-05,
+      "loss": 2.3324,
+      "step": 1656
+    },
+    {
+      "epoch": 0.5372892347600519,
+      "grad_norm": 0.5565251708030701,
+      "learning_rate": 3e-05,
+      "loss": 2.6423,
+      "step": 1657
+    },
+    {
+      "epoch": 0.5376134889753567,
+      "grad_norm": 0.3937702178955078,
+      "learning_rate": 3e-05,
+      "loss": 2.455,
+      "step": 1658
+    },
+    {
+      "epoch": 0.5379377431906615,
+      "grad_norm": 0.644116997718811,
+      "learning_rate": 3e-05,
+      "loss": 2.695,
+      "step": 1659
+    },
+    {
+      "epoch": 0.5382619974059663,
+      "grad_norm": 0.6612233519554138,
+      "learning_rate": 3e-05,
+      "loss": 2.9128,
+      "step": 1660
+    },
+    {
+      "epoch": 0.538586251621271,
+      "grad_norm": 0.48336008191108704,
+      "learning_rate": 3e-05,
+      "loss": 3.1201,
+      "step": 1661
+    },
+    {
+      "epoch": 0.5389105058365758,
+      "grad_norm": 0.42623984813690186,
+      "learning_rate": 3e-05,
+      "loss": 2.5355,
+      "step": 1662
+    },
+    {
+      "epoch": 0.5392347600518806,
+      "grad_norm": 0.2757989168167114,
+      "learning_rate": 3e-05,
+      "loss": 1.7706,
+      "step": 1663
+    },
+    {
+      "epoch": 0.5395590142671854,
+      "grad_norm": 0.3477553725242615,
+      "learning_rate": 3e-05,
+      "loss": 2.5057,
+      "step": 1664
+    },
+    {
+      "epoch": 0.5398832684824902,
+      "grad_norm": 0.42022421956062317,
+      "learning_rate": 3e-05,
+      "loss": 2.8608,
+      "step": 1665
+    },
+    {
+      "epoch": 0.540207522697795,
+      "grad_norm": 0.650000274181366,
+      "learning_rate": 3e-05,
+      "loss": 3.3154,
+      "step": 1666
+    },
+    {
+      "epoch": 0.5405317769130998,
+      "grad_norm": 0.5534403324127197,
+      "learning_rate": 3e-05,
+      "loss": 2.3317,
+      "step": 1667
+    },
+    {
+      "epoch": 0.5408560311284046,
+      "grad_norm": 0.41957032680511475,
+      "learning_rate": 3e-05,
+      "loss": 2.8595,
+      "step": 1668
+    },
+    {
+      "epoch": 0.5411802853437094,
+      "grad_norm": 0.4904481768608093,
+      "learning_rate": 3e-05,
+      "loss": 2.8035,
+      "step": 1669
+    },
+    {
+      "epoch": 0.5415045395590142,
+      "grad_norm": 0.4693237543106079,
+      "learning_rate": 3e-05,
+      "loss": 3.0356,
+      "step": 1670
+    },
+    {
+      "epoch": 0.541828793774319,
+      "grad_norm": 0.41217854619026184,
+      "learning_rate": 3e-05,
+      "loss": 2.6187,
+      "step": 1671
+    },
+    {
+      "epoch": 0.5421530479896238,
+      "grad_norm": 0.38579264283180237,
+      "learning_rate": 3e-05,
+      "loss": 2.9645,
+      "step": 1672
+    },
+    {
+      "epoch": 0.5424773022049286,
+      "grad_norm": 0.313135027885437,
+      "learning_rate": 3e-05,
+      "loss": 2.2618,
+      "step": 1673
+    },
+    {
+      "epoch": 0.5428015564202334,
+      "grad_norm": 0.3807741403579712,
+      "learning_rate": 3e-05,
+      "loss": 2.7795,
+      "step": 1674
+    },
+    {
+      "epoch": 0.5431258106355382,
+      "grad_norm": 0.2820696234703064,
+      "learning_rate": 3e-05,
+      "loss": 2.0973,
+      "step": 1675
+    },
+    {
+      "epoch": 0.543450064850843,
+      "grad_norm": 0.3380673825740814,
+      "learning_rate": 3e-05,
+      "loss": 1.9946,
+      "step": 1676
+    },
+    {
+      "epoch": 0.5437743190661478,
+      "grad_norm": 0.4555770456790924,
+      "learning_rate": 3e-05,
+      "loss": 2.5309,
+      "step": 1677
+    },
+    {
+      "epoch": 0.5440985732814526,
+      "grad_norm": 0.4609755277633667,
+      "learning_rate": 3e-05,
+      "loss": 2.5356,
+      "step": 1678
+    },
+    {
+      "epoch": 0.5444228274967574,
+      "grad_norm": 0.3884716033935547,
+      "learning_rate": 3e-05,
+      "loss": 2.5982,
+      "step": 1679
+    },
+    {
+      "epoch": 0.5447470817120622,
+      "grad_norm": 0.4033474922180176,
+      "learning_rate": 3e-05,
+      "loss": 2.3892,
+      "step": 1680
+    },
+    {
+      "epoch": 0.545071335927367,
+      "grad_norm": 0.36535418033599854,
+      "learning_rate": 3e-05,
+      "loss": 2.4818,
+      "step": 1681
+    },
+    {
+      "epoch": 0.5453955901426718,
+      "grad_norm": 0.378195583820343,
+      "learning_rate": 3e-05,
+      "loss": 2.3789,
+      "step": 1682
+    },
+    {
+      "epoch": 0.5457198443579766,
+      "grad_norm": 0.631031334400177,
+      "learning_rate": 3e-05,
+      "loss": 3.2104,
+      "step": 1683
+    },
+    {
+      "epoch": 0.5460440985732814,
+      "grad_norm": 1.0410749912261963,
+      "learning_rate": 3e-05,
+      "loss": 3.7595,
+      "step": 1684
+    },
+    {
+      "epoch": 0.5463683527885862,
+      "grad_norm": 0.40788111090660095,
+      "learning_rate": 3e-05,
+      "loss": 2.9906,
+      "step": 1685
+    },
+    {
+      "epoch": 0.546692607003891,
+      "grad_norm": 0.5050503015518188,
+      "learning_rate": 3e-05,
+      "loss": 3.0116,
+      "step": 1686
+    },
+    {
+      "epoch": 0.5470168612191959,
+      "grad_norm": 0.4566037058830261,
+      "learning_rate": 3e-05,
+      "loss": 2.5176,
+      "step": 1687
+    },
+    {
+      "epoch": 0.5473411154345007,
+      "grad_norm": 0.4748525619506836,
+      "learning_rate": 3e-05,
+      "loss": 2.358,
+      "step": 1688
+    },
+    {
+      "epoch": 0.5476653696498055,
+      "grad_norm": 0.5324022769927979,
+      "learning_rate": 3e-05,
+      "loss": 2.6698,
+      "step": 1689
+    },
+    {
+      "epoch": 0.5479896238651103,
+      "grad_norm": 0.5735905170440674,
+      "learning_rate": 3e-05,
+      "loss": 2.6275,
+      "step": 1690
+    },
+    {
+      "epoch": 0.5483138780804151,
+      "grad_norm": 0.3671328127384186,
+      "learning_rate": 3e-05,
+      "loss": 2.2247,
+      "step": 1691
+    },
+    {
+      "epoch": 0.5486381322957199,
+      "grad_norm": 0.45348644256591797,
+      "learning_rate": 3e-05,
+      "loss": 2.4851,
+      "step": 1692
+    },
+    {
+      "epoch": 0.5489623865110247,
+      "grad_norm": 0.5420578718185425,
+      "learning_rate": 3e-05,
+      "loss": 2.9634,
+      "step": 1693
+    },
+    {
+      "epoch": 0.5492866407263295,
+      "grad_norm": 0.6672835350036621,
+      "learning_rate": 3e-05,
+      "loss": 3.0313,
+      "step": 1694
+    },
+    {
+      "epoch": 0.5496108949416343,
+      "grad_norm": 0.40458494424819946,
+      "learning_rate": 3e-05,
+      "loss": 2.7476,
+      "step": 1695
+    },
+    {
+      "epoch": 0.5499351491569391,
+      "grad_norm": 0.524838924407959,
+      "learning_rate": 3e-05,
+      "loss": 2.8705,
+      "step": 1696
+    },
+    {
+      "epoch": 0.5502594033722439,
+      "grad_norm": 0.4901329278945923,
+      "learning_rate": 3e-05,
+      "loss": 3.2133,
+      "step": 1697
+    },
+    {
+      "epoch": 0.5505836575875487,
+      "grad_norm": 0.5130648612976074,
+      "learning_rate": 3e-05,
+      "loss": 3.3287,
+      "step": 1698
+    },
+    {
+      "epoch": 0.5509079118028535,
+      "grad_norm": 0.48373404145240784,
+      "learning_rate": 3e-05,
+      "loss": 2.3679,
+      "step": 1699
+    },
+    {
+      "epoch": 0.5512321660181583,
+      "grad_norm": 0.6387945413589478,
+      "learning_rate": 3e-05,
+      "loss": 2.8311,
+      "step": 1700
+    },
+    {
+      "epoch": 0.5515564202334631,
+      "grad_norm": 0.33760374784469604,
+      "learning_rate": 3e-05,
+      "loss": 2.4218,
+      "step": 1701
+    },
+    {
+      "epoch": 0.5518806744487679,
+      "grad_norm": 0.3635815382003784,
+      "learning_rate": 3e-05,
+      "loss": 2.4925,
+      "step": 1702
+    },
+    {
+      "epoch": 0.5522049286640727,
+      "grad_norm": 0.3842930197715759,
+      "learning_rate": 3e-05,
+      "loss": 2.667,
+      "step": 1703
+    },
+    {
+      "epoch": 0.5525291828793775,
+      "grad_norm": 0.5629712343215942,
+      "learning_rate": 3e-05,
+      "loss": 2.9096,
+      "step": 1704
+    },
+    {
+      "epoch": 0.5528534370946823,
+      "grad_norm": 0.6862927079200745,
+      "learning_rate": 3e-05,
+      "loss": 3.3433,
+      "step": 1705
+    },
+    {
+      "epoch": 0.5531776913099871,
+      "grad_norm": 0.4313713610172272,
+      "learning_rate": 3e-05,
+      "loss": 2.7659,
+      "step": 1706
+    },
+    {
+      "epoch": 0.5535019455252919,
+      "grad_norm": 0.30821502208709717,
+      "learning_rate": 3e-05,
+      "loss": 2.3357,
+      "step": 1707
+    },
+    {
+      "epoch": 0.5538261997405967,
+      "grad_norm": 0.4664706885814667,
+      "learning_rate": 3e-05,
+      "loss": 2.8406,
+      "step": 1708
+    },
+    {
+      "epoch": 0.5541504539559015,
+      "grad_norm": 0.39621731638908386,
+      "learning_rate": 3e-05,
+      "loss": 2.4691,
+      "step": 1709
+    },
+    {
+      "epoch": 0.5544747081712063,
+      "grad_norm": 0.46711283922195435,
+      "learning_rate": 3e-05,
+      "loss": 2.8851,
+      "step": 1710
+    },
+    {
+      "epoch": 0.554798962386511,
+      "grad_norm": 0.3194664418697357,
+      "learning_rate": 3e-05,
+      "loss": 1.9736,
+      "step": 1711
+    },
+    {
+      "epoch": 0.5551232166018158,
+      "grad_norm": 0.4700626730918884,
+      "learning_rate": 3e-05,
+      "loss": 2.3882,
+      "step": 1712
+    },
+    {
+      "epoch": 0.5554474708171206,
+      "grad_norm": 0.4381219744682312,
+      "learning_rate": 3e-05,
+      "loss": 2.5392,
+      "step": 1713
+    },
+    {
+      "epoch": 0.5557717250324254,
+      "grad_norm": 0.595716118812561,
+      "learning_rate": 3e-05,
+      "loss": 2.6654,
+      "step": 1714
+    },
+    {
+      "epoch": 0.5560959792477302,
+      "grad_norm": 0.39745011925697327,
+      "learning_rate": 3e-05,
+      "loss": 2.3864,
+      "step": 1715
+    },
+    {
+      "epoch": 0.556420233463035,
+      "grad_norm": 0.3404266834259033,
+      "learning_rate": 3e-05,
+      "loss": 2.39,
+      "step": 1716
+    },
+    {
+      "epoch": 0.5567444876783398,
+      "grad_norm": 0.532381534576416,
+      "learning_rate": 3e-05,
+      "loss": 3.2491,
+      "step": 1717
+    },
+    {
+      "epoch": 0.5570687418936446,
+      "grad_norm": 0.4522727429866791,
+      "learning_rate": 3e-05,
+      "loss": 3.6271,
+      "step": 1718
+    },
+    {
+      "epoch": 0.5573929961089494,
+      "grad_norm": 0.4577166736125946,
+      "learning_rate": 3e-05,
+      "loss": 3.0033,
+      "step": 1719
+    },
+    {
+      "epoch": 0.5577172503242542,
+      "grad_norm": 0.34769439697265625,
+      "learning_rate": 3e-05,
+      "loss": 2.447,
+      "step": 1720
+    },
+    {
+      "epoch": 0.558041504539559,
+      "grad_norm": 0.3181087374687195,
+      "learning_rate": 3e-05,
+      "loss": 2.2277,
+      "step": 1721
+    },
+    {
+      "epoch": 0.5583657587548638,
+      "grad_norm": 0.4336315393447876,
+      "learning_rate": 3e-05,
+      "loss": 2.4345,
+      "step": 1722
+    },
+    {
+      "epoch": 0.5586900129701686,
+      "grad_norm": 0.34905102849006653,
+      "learning_rate": 3e-05,
+      "loss": 2.2933,
+      "step": 1723
+    },
+    {
+      "epoch": 0.5590142671854734,
+      "grad_norm": 0.31294018030166626,
+      "learning_rate": 3e-05,
+      "loss": 2.2684,
+      "step": 1724
+    },
+    {
+      "epoch": 0.5593385214007782,
+      "grad_norm": 0.7225369811058044,
+      "learning_rate": 3e-05,
+      "loss": 3.391,
+      "step": 1725
+    },
+    {
+      "epoch": 0.559662775616083,
+      "grad_norm": 0.3601735532283783,
+      "learning_rate": 3e-05,
+      "loss": 2.3511,
+      "step": 1726
+    },
+    {
+      "epoch": 0.5599870298313878,
+      "grad_norm": 0.31936317682266235,
+      "learning_rate": 3e-05,
+      "loss": 2.3933,
+      "step": 1727
+    },
+    {
+      "epoch": 0.5603112840466926,
+      "grad_norm": 0.3414732813835144,
+      "learning_rate": 3e-05,
+      "loss": 2.244,
+      "step": 1728
+    },
+    {
+      "epoch": 0.5606355382619974,
+      "grad_norm": 0.33823007345199585,
+      "learning_rate": 3e-05,
+      "loss": 2.3737,
+      "step": 1729
+    },
+    {
+      "epoch": 0.5609597924773022,
+      "grad_norm": 0.6727493405342102,
+      "learning_rate": 3e-05,
+      "loss": 3.4171,
+      "step": 1730
+    },
+    {
+      "epoch": 0.561284046692607,
+      "grad_norm": 0.4312712550163269,
+      "learning_rate": 3e-05,
+      "loss": 2.8316,
+      "step": 1731
+    },
+    {
+      "epoch": 0.5616083009079118,
+      "grad_norm": 0.3770003914833069,
+      "learning_rate": 3e-05,
+      "loss": 2.3613,
+      "step": 1732
+    },
+    {
+      "epoch": 0.5619325551232166,
+      "grad_norm": 0.5660895109176636,
+      "learning_rate": 3e-05,
+      "loss": 2.7777,
+      "step": 1733
+    },
+    {
+      "epoch": 0.5622568093385214,
+      "grad_norm": 0.5565730929374695,
+      "learning_rate": 3e-05,
+      "loss": 3.1668,
+      "step": 1734
+    },
+    {
+      "epoch": 0.5625810635538262,
+      "grad_norm": 0.4311894178390503,
+      "learning_rate": 3e-05,
+      "loss": 2.7355,
+      "step": 1735
+    },
+    {
+      "epoch": 0.562905317769131,
+      "grad_norm": 0.5821871161460876,
+      "learning_rate": 3e-05,
+      "loss": 3.2736,
+      "step": 1736
+    },
+    {
+      "epoch": 0.5632295719844358,
+      "grad_norm": 0.4946500062942505,
+      "learning_rate": 3e-05,
+      "loss": 3.2485,
+      "step": 1737
+    },
+    {
+      "epoch": 0.5635538261997406,
+      "grad_norm": 0.6859528422355652,
+      "learning_rate": 3e-05,
+      "loss": 2.9404,
+      "step": 1738
+    },
+    {
+      "epoch": 0.5638780804150454,
+      "grad_norm": 0.600695788860321,
+      "learning_rate": 3e-05,
+      "loss": 2.8698,
+      "step": 1739
+    },
+    {
+      "epoch": 0.5642023346303502,
+      "grad_norm": 0.300672709941864,
+      "learning_rate": 3e-05,
+      "loss": 2.2543,
+      "step": 1740
+    },
+    {
+      "epoch": 0.564526588845655,
+      "grad_norm": 0.5719174742698669,
+      "learning_rate": 3e-05,
+      "loss": 3.5741,
+      "step": 1741
+    },
+    {
+      "epoch": 0.5648508430609598,
+      "grad_norm": 0.38101133704185486,
+      "learning_rate": 3e-05,
+      "loss": 2.4432,
+      "step": 1742
+    },
+    {
+      "epoch": 0.5651750972762646,
+      "grad_norm": 0.4650246798992157,
+      "learning_rate": 3e-05,
+      "loss": 2.354,
+      "step": 1743
+    },
+    {
+      "epoch": 0.5654993514915694,
+      "grad_norm": 0.43708091974258423,
+      "learning_rate": 3e-05,
+      "loss": 2.6757,
+      "step": 1744
+    },
+    {
+      "epoch": 0.5658236057068742,
+      "grad_norm": 0.5063639283180237,
+      "learning_rate": 3e-05,
+      "loss": 3.1012,
+      "step": 1745
+    },
+    {
+      "epoch": 0.566147859922179,
+      "grad_norm": 0.5364939570426941,
+      "learning_rate": 3e-05,
+      "loss": 2.6922,
+      "step": 1746
+    },
+    {
+      "epoch": 0.5664721141374838,
+      "grad_norm": 0.3901953101158142,
+      "learning_rate": 3e-05,
+      "loss": 2.8505,
+      "step": 1747
+    },
+    {
+      "epoch": 0.5667963683527886,
+      "grad_norm": 0.5631880164146423,
+      "learning_rate": 3e-05,
+      "loss": 3.3305,
+      "step": 1748
+    },
+    {
+      "epoch": 0.5671206225680934,
+      "grad_norm": 0.47534021735191345,
+      "learning_rate": 3e-05,
+      "loss": 3.0749,
+      "step": 1749
+    },
+    {
+      "epoch": 0.5674448767833982,
+      "grad_norm": 0.7674520015716553,
+      "learning_rate": 3e-05,
+      "loss": 3.543,
+      "step": 1750
+    },
+    {
+      "epoch": 0.567769130998703,
+      "grad_norm": 0.4259376525878906,
+      "learning_rate": 3e-05,
+      "loss": 3.2088,
+      "step": 1751
+    },
+    {
+      "epoch": 0.5680933852140078,
+      "grad_norm": 0.44311511516571045,
+      "learning_rate": 3e-05,
+      "loss": 2.5777,
+      "step": 1752
+    },
+    {
+      "epoch": 0.5684176394293126,
+      "grad_norm": 0.3596744239330292,
+      "learning_rate": 3e-05,
+      "loss": 2.2257,
+      "step": 1753
+    },
+    {
+      "epoch": 0.5687418936446174,
+      "grad_norm": 0.4873691499233246,
+      "learning_rate": 3e-05,
+      "loss": 2.9118,
+      "step": 1754
+    },
+    {
+      "epoch": 0.5690661478599222,
+      "grad_norm": 0.5407915711402893,
+      "learning_rate": 3e-05,
+      "loss": 3.0222,
+      "step": 1755
+    },
+    {
+      "epoch": 0.569390402075227,
+      "grad_norm": 0.509260356426239,
+      "learning_rate": 3e-05,
+      "loss": 3.0826,
+      "step": 1756
+    },
+    {
+      "epoch": 0.5697146562905318,
+      "grad_norm": 0.34407052397727966,
+      "learning_rate": 3e-05,
+      "loss": 2.3152,
+      "step": 1757
+    },
+    {
+      "epoch": 0.5700389105058365,
+      "grad_norm": 0.616728663444519,
+      "learning_rate": 3e-05,
+      "loss": 3.1423,
+      "step": 1758
+    },
+    {
+      "epoch": 0.5703631647211413,
+      "grad_norm": 0.4021288752555847,
+      "learning_rate": 3e-05,
+      "loss": 2.366,
+      "step": 1759
+    },
+    {
+      "epoch": 0.5706874189364461,
+      "grad_norm": 0.4219648838043213,
+      "learning_rate": 3e-05,
+      "loss": 2.3234,
+      "step": 1760
+    },
+    {
+      "epoch": 0.5710116731517509,
+      "grad_norm": 0.2833535373210907,
+      "learning_rate": 3e-05,
+      "loss": 2.0496,
+      "step": 1761
+    },
+    {
+      "epoch": 0.5713359273670557,
+      "grad_norm": 0.9621824026107788,
+      "learning_rate": 3e-05,
+      "loss": 2.6061,
+      "step": 1762
+    },
+    {
+      "epoch": 0.5716601815823605,
+      "grad_norm": 0.34205707907676697,
+      "learning_rate": 3e-05,
+      "loss": 2.6287,
+      "step": 1763
+    },
+    {
+      "epoch": 0.5719844357976653,
+      "grad_norm": 0.6648195385932922,
+      "learning_rate": 3e-05,
+      "loss": 3.1167,
+      "step": 1764
+    },
+    {
+      "epoch": 0.5723086900129701,
+      "grad_norm": 0.4204351007938385,
+      "learning_rate": 3e-05,
+      "loss": 2.9824,
+      "step": 1765
+    },
+    {
+      "epoch": 0.5726329442282749,
+      "grad_norm": 0.3223739266395569,
+      "learning_rate": 3e-05,
+      "loss": 2.1068,
+      "step": 1766
+    },
+    {
+      "epoch": 0.5729571984435797,
+      "grad_norm": 0.4766865670681,
+      "learning_rate": 3e-05,
+      "loss": 2.9329,
+      "step": 1767
+    },
+    {
+      "epoch": 0.5732814526588845,
+      "grad_norm": 0.5410792231559753,
+      "learning_rate": 3e-05,
+      "loss": 2.7088,
+      "step": 1768
+    },
+    {
+      "epoch": 0.5736057068741893,
+      "grad_norm": 0.4642774760723114,
+      "learning_rate": 3e-05,
+      "loss": 2.8256,
+      "step": 1769
+    },
+    {
+      "epoch": 0.5739299610894941,
+      "grad_norm": 0.43121376633644104,
+      "learning_rate": 3e-05,
+      "loss": 2.6824,
+      "step": 1770
+    },
+    {
+      "epoch": 0.5742542153047989,
+      "grad_norm": 0.5465131998062134,
+      "learning_rate": 3e-05,
+      "loss": 3.0738,
+      "step": 1771
+    },
+    {
+      "epoch": 0.5745784695201037,
+      "grad_norm": 0.3423590064048767,
+      "learning_rate": 3e-05,
+      "loss": 2.4717,
+      "step": 1772
+    },
+    {
+      "epoch": 0.5749027237354085,
+      "grad_norm": 0.5503602623939514,
+      "learning_rate": 3e-05,
+      "loss": 3.0677,
+      "step": 1773
+    },
+    {
+      "epoch": 0.5752269779507133,
+      "grad_norm": 0.847072422504425,
+      "learning_rate": 3e-05,
+      "loss": 3.4197,
+      "step": 1774
+    },
+    {
+      "epoch": 0.5755512321660181,
+      "grad_norm": 0.3998866081237793,
+      "learning_rate": 3e-05,
+      "loss": 2.6389,
+      "step": 1775
+    },
+    {
+      "epoch": 0.5758754863813229,
+      "grad_norm": 0.46230271458625793,
+      "learning_rate": 3e-05,
+      "loss": 2.5819,
+      "step": 1776
+    },
+    {
+      "epoch": 0.5761997405966277,
+      "grad_norm": 0.5279173254966736,
+      "learning_rate": 3e-05,
+      "loss": 2.8956,
+      "step": 1777
+    },
+    {
+      "epoch": 0.5765239948119325,
+      "grad_norm": 0.43178510665893555,
+      "learning_rate": 3e-05,
+      "loss": 2.9113,
+      "step": 1778
+    },
+    {
+      "epoch": 0.5768482490272373,
+      "grad_norm": 0.43116989731788635,
+      "learning_rate": 3e-05,
+      "loss": 2.575,
+      "step": 1779
+    },
+    {
+      "epoch": 0.5771725032425421,
+      "grad_norm": 0.4642339050769806,
+      "learning_rate": 3e-05,
+      "loss": 3.2779,
+      "step": 1780
+    },
+    {
+      "epoch": 0.5774967574578469,
+      "grad_norm": 0.2988036572933197,
+      "learning_rate": 3e-05,
+      "loss": 2.1559,
+      "step": 1781
+    },
+    {
+      "epoch": 0.5778210116731517,
+      "grad_norm": 0.5306640863418579,
+      "learning_rate": 3e-05,
+      "loss": 3.299,
+      "step": 1782
+    },
+    {
+      "epoch": 0.5781452658884566,
+      "grad_norm": 0.6150776147842407,
+      "learning_rate": 3e-05,
+      "loss": 3.5295,
+      "step": 1783
+    },
+    {
+      "epoch": 0.5784695201037614,
+      "grad_norm": 0.33247795701026917,
+      "learning_rate": 3e-05,
+      "loss": 2.332,
+      "step": 1784
+    },
+    {
+      "epoch": 0.5787937743190662,
+      "grad_norm": 0.37106800079345703,
+      "learning_rate": 3e-05,
+      "loss": 2.9146,
+      "step": 1785
+    },
+    {
+      "epoch": 0.579118028534371,
+      "grad_norm": 0.5441930890083313,
+      "learning_rate": 3e-05,
+      "loss": 3.0233,
+      "step": 1786
+    },
+    {
+      "epoch": 0.5794422827496758,
+      "grad_norm": 0.46233853697776794,
+      "learning_rate": 3e-05,
+      "loss": 3.3404,
+      "step": 1787
+    },
+    {
+      "epoch": 0.5797665369649806,
+      "grad_norm": 0.48225072026252747,
+      "learning_rate": 3e-05,
+      "loss": 2.9499,
+      "step": 1788
+    },
+    {
+      "epoch": 0.5800907911802854,
+      "grad_norm": 0.435072660446167,
+      "learning_rate": 3e-05,
+      "loss": 2.6183,
+      "step": 1789
+    },
+    {
+      "epoch": 0.5804150453955902,
+      "grad_norm": 0.5129862427711487,
+      "learning_rate": 3e-05,
+      "loss": 2.7404,
+      "step": 1790
+    },
+    {
+      "epoch": 0.580739299610895,
+      "grad_norm": 0.3233097791671753,
+      "learning_rate": 3e-05,
+      "loss": 2.3887,
+      "step": 1791
+    },
+    {
+      "epoch": 0.5810635538261998,
+      "grad_norm": 0.364729106426239,
+      "learning_rate": 3e-05,
+      "loss": 2.4422,
+      "step": 1792
+    },
+    {
+      "epoch": 0.5813878080415046,
+      "grad_norm": 0.33344748616218567,
+      "learning_rate": 3e-05,
+      "loss": 2.2998,
+      "step": 1793
+    },
+    {
+      "epoch": 0.5817120622568094,
+      "grad_norm": 1.658810019493103,
+      "learning_rate": 3e-05,
+      "loss": 3.2382,
+      "step": 1794
+    },
+    {
+      "epoch": 0.5820363164721142,
+      "grad_norm": 0.4670470654964447,
+      "learning_rate": 3e-05,
+      "loss": 2.7033,
+      "step": 1795
+    },
+    {
+      "epoch": 0.582360570687419,
+      "grad_norm": 0.36022329330444336,
+      "learning_rate": 3e-05,
+      "loss": 2.4773,
+      "step": 1796
+    },
+    {
+      "epoch": 0.5826848249027238,
+      "grad_norm": 0.4391596019268036,
+      "learning_rate": 3e-05,
+      "loss": 2.7309,
+      "step": 1797
+    },
+    {
+      "epoch": 0.5830090791180286,
+      "grad_norm": 0.31371477246284485,
+      "learning_rate": 3e-05,
+      "loss": 2.3109,
+      "step": 1798
+    },
+    {
+      "epoch": 0.5833333333333334,
+      "grad_norm": 0.2933622896671295,
+      "learning_rate": 3e-05,
+      "loss": 2.2397,
+      "step": 1799
+    },
+    {
+      "epoch": 0.5836575875486382,
+      "grad_norm": 0.27904602885246277,
+      "learning_rate": 3e-05,
+      "loss": 2.1983,
+      "step": 1800
+    },
+    {
+      "epoch": 0.583981841763943,
+      "grad_norm": 0.3553931713104248,
+      "learning_rate": 3e-05,
+      "loss": 2.3617,
+      "step": 1801
+    },
+    {
+      "epoch": 0.5843060959792478,
+      "grad_norm": 0.34696537256240845,
+      "learning_rate": 3e-05,
+      "loss": 2.4192,
+      "step": 1802
+    },
+    {
+      "epoch": 0.5846303501945526,
+      "grad_norm": 0.38925695419311523,
+      "learning_rate": 3e-05,
+      "loss": 2.6395,
+      "step": 1803
+    },
+    {
+      "epoch": 0.5849546044098574,
+      "grad_norm": 0.3545275926589966,
+      "learning_rate": 3e-05,
+      "loss": 2.5767,
+      "step": 1804
+    },
+    {
+      "epoch": 0.5852788586251622,
+      "grad_norm": 0.35899627208709717,
+      "learning_rate": 3e-05,
+      "loss": 2.6549,
+      "step": 1805
+    },
+    {
+      "epoch": 0.585603112840467,
+      "grad_norm": 0.3642694354057312,
+      "learning_rate": 3e-05,
+      "loss": 2.7466,
+      "step": 1806
+    },
+    {
+      "epoch": 0.5859273670557718,
+      "grad_norm": 0.48464149236679077,
+      "learning_rate": 3e-05,
+      "loss": 3.2124,
+      "step": 1807
+    },
+    {
+      "epoch": 0.5862516212710766,
+      "grad_norm": 0.3640475869178772,
+      "learning_rate": 3e-05,
+      "loss": 2.4837,
+      "step": 1808
+    },
+    {
+      "epoch": 0.5865758754863813,
+      "grad_norm": 0.3514559864997864,
+      "learning_rate": 3e-05,
+      "loss": 2.5448,
+      "step": 1809
+    },
+    {
+      "epoch": 0.5869001297016861,
+      "grad_norm": 0.39220601320266724,
+      "learning_rate": 3e-05,
+      "loss": 2.5959,
+      "step": 1810
+    },
+    {
+      "epoch": 0.5872243839169909,
+      "grad_norm": 0.3664841055870056,
+      "learning_rate": 3e-05,
+      "loss": 2.7642,
+      "step": 1811
+    },
+    {
+      "epoch": 0.5875486381322957,
+      "grad_norm": 0.44191718101501465,
+      "learning_rate": 3e-05,
+      "loss": 3.0201,
+      "step": 1812
+    },
+    {
+      "epoch": 0.5878728923476005,
+      "grad_norm": 0.3984758257865906,
+      "learning_rate": 3e-05,
+      "loss": 3.0273,
+      "step": 1813
+    },
+    {
+      "epoch": 0.5881971465629053,
+      "grad_norm": 0.4503876268863678,
+      "learning_rate": 3e-05,
+      "loss": 2.966,
+      "step": 1814
+    },
+    {
+      "epoch": 0.5885214007782101,
+      "grad_norm": 0.5097358822822571,
+      "learning_rate": 3e-05,
+      "loss": 3.1993,
+      "step": 1815
+    },
+    {
+      "epoch": 0.5888456549935149,
+      "grad_norm": 0.38068950176239014,
+      "learning_rate": 3e-05,
+      "loss": 2.3896,
+      "step": 1816
+    },
+    {
+      "epoch": 0.5891699092088197,
+      "grad_norm": 0.4630526602268219,
+      "learning_rate": 3e-05,
+      "loss": 3.2401,
+      "step": 1817
+    },
+    {
+      "epoch": 0.5894941634241245,
+      "grad_norm": 0.3317679166793823,
+      "learning_rate": 3e-05,
+      "loss": 2.1086,
+      "step": 1818
+    },
+    {
+      "epoch": 0.5898184176394293,
+      "grad_norm": 0.368918240070343,
+      "learning_rate": 3e-05,
+      "loss": 2.4813,
+      "step": 1819
+    },
+    {
+      "epoch": 0.5901426718547341,
+      "grad_norm": 0.5888726711273193,
+      "learning_rate": 3e-05,
+      "loss": 3.0169,
+      "step": 1820
+    },
+    {
+      "epoch": 0.5904669260700389,
+      "grad_norm": 0.617592453956604,
+      "learning_rate": 3e-05,
+      "loss": 2.7348,
+      "step": 1821
+    },
+    {
+      "epoch": 0.5907911802853437,
+      "grad_norm": 0.4514785706996918,
+      "learning_rate": 3e-05,
+      "loss": 2.5652,
+      "step": 1822
+    },
+    {
+      "epoch": 0.5911154345006485,
+      "grad_norm": 0.6451595425605774,
+      "learning_rate": 3e-05,
+      "loss": 3.5047,
+      "step": 1823
+    },
+    {
+      "epoch": 0.5914396887159533,
+      "grad_norm": 0.4798338711261749,
+      "learning_rate": 3e-05,
+      "loss": 3.1274,
+      "step": 1824
+    },
+    {
+      "epoch": 0.5917639429312581,
+      "grad_norm": 0.512535572052002,
+      "learning_rate": 3e-05,
+      "loss": 3.2161,
+      "step": 1825
+    },
+    {
+      "epoch": 0.5920881971465629,
+      "grad_norm": 0.41621848940849304,
+      "learning_rate": 3e-05,
+      "loss": 2.7237,
+      "step": 1826
+    },
+    {
+      "epoch": 0.5924124513618677,
+      "grad_norm": 0.5292104482650757,
+      "learning_rate": 3e-05,
+      "loss": 2.4842,
+      "step": 1827
+    },
+    {
+      "epoch": 0.5927367055771725,
+      "grad_norm": 0.7832330465316772,
+      "learning_rate": 3e-05,
+      "loss": 2.675,
+      "step": 1828
+    },
+    {
+      "epoch": 0.5930609597924773,
+      "grad_norm": 0.3440128564834595,
+      "learning_rate": 3e-05,
+      "loss": 2.1628,
+      "step": 1829
+    },
+    {
+      "epoch": 0.5933852140077821,
+      "grad_norm": 0.411664217710495,
+      "learning_rate": 3e-05,
+      "loss": 3.0656,
+      "step": 1830
+    },
+    {
+      "epoch": 0.5937094682230869,
+      "grad_norm": 0.4345646798610687,
+      "learning_rate": 3e-05,
+      "loss": 2.6729,
+      "step": 1831
+    },
+    {
+      "epoch": 0.5940337224383917,
+      "grad_norm": 0.32874011993408203,
+      "learning_rate": 3e-05,
+      "loss": 2.021,
+      "step": 1832
+    },
+    {
+      "epoch": 0.5943579766536965,
+      "grad_norm": 0.3970089554786682,
+      "learning_rate": 3e-05,
+      "loss": 2.0732,
+      "step": 1833
+    },
+    {
+      "epoch": 0.5946822308690013,
+      "grad_norm": 0.5751341581344604,
+      "learning_rate": 3e-05,
+      "loss": 2.8406,
+      "step": 1834
+    },
+    {
+      "epoch": 0.5950064850843061,
+      "grad_norm": 0.6372278928756714,
+      "learning_rate": 3e-05,
+      "loss": 2.9168,
+      "step": 1835
+    },
+    {
+      "epoch": 0.5953307392996109,
+      "grad_norm": 0.5301762223243713,
+      "learning_rate": 3e-05,
+      "loss": 3.1138,
+      "step": 1836
+    },
+    {
+      "epoch": 0.5956549935149157,
+      "grad_norm": 0.3825346529483795,
+      "learning_rate": 3e-05,
+      "loss": 2.5897,
+      "step": 1837
+    },
+    {
+      "epoch": 0.5959792477302205,
+      "grad_norm": 0.47653621435165405,
+      "learning_rate": 3e-05,
+      "loss": 2.9774,
+      "step": 1838
+    },
+    {
+      "epoch": 0.5963035019455253,
+      "grad_norm": 0.3527122437953949,
+      "learning_rate": 3e-05,
+      "loss": 2.2782,
+      "step": 1839
+    },
+    {
+      "epoch": 0.5966277561608301,
+      "grad_norm": 0.4622247517108917,
+      "learning_rate": 3e-05,
+      "loss": 2.7622,
+      "step": 1840
+    },
+    {
+      "epoch": 0.5969520103761349,
+      "grad_norm": 0.4090559482574463,
+      "learning_rate": 3e-05,
+      "loss": 3.29,
+      "step": 1841
+    },
+    {
+      "epoch": 0.5972762645914397,
+      "grad_norm": 0.35499808192253113,
+      "learning_rate": 3e-05,
+      "loss": 2.3799,
+      "step": 1842
+    },
+    {
+      "epoch": 0.5976005188067445,
+      "grad_norm": 0.43470498919487,
+      "learning_rate": 3e-05,
+      "loss": 2.8666,
+      "step": 1843
+    },
+    {
+      "epoch": 0.5979247730220493,
+      "grad_norm": 0.3549688756465912,
+      "learning_rate": 3e-05,
+      "loss": 2.3169,
+      "step": 1844
+    },
+    {
+      "epoch": 0.5982490272373541,
+      "grad_norm": 0.436036616563797,
+      "learning_rate": 3e-05,
+      "loss": 2.5633,
+      "step": 1845
+    },
+    {
+      "epoch": 0.5985732814526589,
+      "grad_norm": 0.6829612255096436,
+      "learning_rate": 3e-05,
+      "loss": 3.6151,
+      "step": 1846
+    },
+    {
+      "epoch": 0.5988975356679637,
+      "grad_norm": 0.4949396848678589,
+      "learning_rate": 3e-05,
+      "loss": 2.9184,
+      "step": 1847
+    },
+    {
+      "epoch": 0.5992217898832685,
+      "grad_norm": 0.4274238348007202,
+      "learning_rate": 3e-05,
+      "loss": 2.9201,
+      "step": 1848
+    },
+    {
+      "epoch": 0.5995460440985733,
+      "grad_norm": 0.37713590264320374,
+      "learning_rate": 3e-05,
+      "loss": 2.3683,
+      "step": 1849
+    },
+    {
+      "epoch": 0.5998702983138781,
+      "grad_norm": 0.3964006006717682,
+      "learning_rate": 3e-05,
+      "loss": 2.6579,
+      "step": 1850
+    },
+    {
+      "epoch": 0.6001945525291829,
+      "grad_norm": 0.493781715631485,
+      "learning_rate": 3e-05,
+      "loss": 2.8205,
+      "step": 1851
+    },
+    {
+      "epoch": 0.6005188067444877,
+      "grad_norm": 0.418012410402298,
+      "learning_rate": 3e-05,
+      "loss": 2.468,
+      "step": 1852
+    },
+    {
+      "epoch": 0.6008430609597925,
+      "grad_norm": 0.4089684784412384,
+      "learning_rate": 3e-05,
+      "loss": 2.3987,
+      "step": 1853
+    },
+    {
+      "epoch": 0.6011673151750972,
+      "grad_norm": 0.4211864173412323,
+      "learning_rate": 3e-05,
+      "loss": 2.7083,
+      "step": 1854
+    },
+    {
+      "epoch": 0.601491569390402,
+      "grad_norm": 0.4098812937736511,
+      "learning_rate": 3e-05,
+      "loss": 2.5147,
+      "step": 1855
+    },
+    {
+      "epoch": 0.6018158236057068,
+      "grad_norm": 0.4055812358856201,
+      "learning_rate": 3e-05,
+      "loss": 2.7007,
+      "step": 1856
+    },
+    {
+      "epoch": 0.6021400778210116,
+      "grad_norm": 0.46148404479026794,
+      "learning_rate": 3e-05,
+      "loss": 2.6531,
+      "step": 1857
+    },
+    {
+      "epoch": 0.6024643320363164,
+      "grad_norm": 0.3926507234573364,
+      "learning_rate": 3e-05,
+      "loss": 2.4829,
+      "step": 1858
+    },
+    {
+      "epoch": 0.6027885862516212,
+      "grad_norm": 0.7038570642471313,
+      "learning_rate": 3e-05,
+      "loss": 2.9812,
+      "step": 1859
+    },
+    {
+      "epoch": 0.603112840466926,
+      "grad_norm": 0.3273842930793762,
+      "learning_rate": 3e-05,
+      "loss": 2.2868,
+      "step": 1860
+    },
+    {
+      "epoch": 0.6034370946822308,
+      "grad_norm": 0.5636449456214905,
+      "learning_rate": 3e-05,
+      "loss": 3.0984,
+      "step": 1861
+    },
+    {
+      "epoch": 0.6037613488975356,
+      "grad_norm": 0.5151119232177734,
+      "learning_rate": 3e-05,
+      "loss": 3.4479,
+      "step": 1862
+    },
+    {
+      "epoch": 0.6040856031128404,
+      "grad_norm": 0.3774999976158142,
+      "learning_rate": 3e-05,
+      "loss": 2.8166,
+      "step": 1863
+    },
+    {
+      "epoch": 0.6044098573281452,
+      "grad_norm": 0.36309072375297546,
+      "learning_rate": 3e-05,
+      "loss": 2.4967,
+      "step": 1864
+    },
+    {
+      "epoch": 0.60473411154345,
+      "grad_norm": 0.43766137957572937,
+      "learning_rate": 3e-05,
+      "loss": 3.1265,
+      "step": 1865
+    },
+    {
+      "epoch": 0.6050583657587548,
+      "grad_norm": 0.41877543926239014,
+      "learning_rate": 3e-05,
+      "loss": 2.8367,
+      "step": 1866
+    },
+    {
+      "epoch": 0.6053826199740596,
+      "grad_norm": 0.4145008623600006,
+      "learning_rate": 3e-05,
+      "loss": 2.7949,
+      "step": 1867
+    },
+    {
+      "epoch": 0.6057068741893644,
+      "grad_norm": 0.33224213123321533,
+      "learning_rate": 3e-05,
+      "loss": 2.4914,
+      "step": 1868
+    },
+    {
+      "epoch": 0.6060311284046692,
+      "grad_norm": 0.375447154045105,
+      "learning_rate": 3e-05,
+      "loss": 3.2756,
+      "step": 1869
+    },
+    {
+      "epoch": 0.606355382619974,
+      "grad_norm": 0.4423779249191284,
+      "learning_rate": 3e-05,
+      "loss": 2.5862,
+      "step": 1870
+    },
+    {
+      "epoch": 0.6066796368352788,
+      "grad_norm": 0.5094466805458069,
+      "learning_rate": 3e-05,
+      "loss": 3.6062,
+      "step": 1871
+    },
+    {
+      "epoch": 0.6070038910505836,
+      "grad_norm": 0.5093324780464172,
+      "learning_rate": 3e-05,
+      "loss": 3.8937,
+      "step": 1872
+    },
+    {
+      "epoch": 0.6073281452658884,
+      "grad_norm": 0.4457301199436188,
+      "learning_rate": 3e-05,
+      "loss": 3.1561,
+      "step": 1873
+    },
+    {
+      "epoch": 0.6076523994811932,
+      "grad_norm": 0.37723642587661743,
+      "learning_rate": 3e-05,
+      "loss": 2.4267,
+      "step": 1874
+    },
+    {
+      "epoch": 0.607976653696498,
+      "grad_norm": 0.34340769052505493,
+      "learning_rate": 3e-05,
+      "loss": 2.6025,
+      "step": 1875
+    },
+    {
+      "epoch": 0.6083009079118028,
+      "grad_norm": 0.35477349162101746,
+      "learning_rate": 3e-05,
+      "loss": 2.4217,
+      "step": 1876
+    },
+    {
+      "epoch": 0.6086251621271076,
+      "grad_norm": 0.4600470960140228,
+      "learning_rate": 3e-05,
+      "loss": 2.2517,
+      "step": 1877
+    },
+    {
+      "epoch": 0.6089494163424124,
+      "grad_norm": 0.4416479170322418,
+      "learning_rate": 3e-05,
+      "loss": 2.6284,
+      "step": 1878
+    },
+    {
+      "epoch": 0.6092736705577172,
+      "grad_norm": 0.4546282887458801,
+      "learning_rate": 3e-05,
+      "loss": 3.5739,
+      "step": 1879
+    },
+    {
+      "epoch": 0.6095979247730221,
+      "grad_norm": 0.3637444078922272,
+      "learning_rate": 3e-05,
+      "loss": 2.3029,
+      "step": 1880
+    },
+    {
+      "epoch": 0.6099221789883269,
+      "grad_norm": 0.34091976284980774,
+      "learning_rate": 3e-05,
+      "loss": 2.7189,
+      "step": 1881
+    },
+    {
+      "epoch": 0.6102464332036317,
+      "grad_norm": 0.38328373432159424,
+      "learning_rate": 3e-05,
+      "loss": 2.8467,
+      "step": 1882
+    },
+    {
+      "epoch": 0.6105706874189365,
+      "grad_norm": 0.379491925239563,
+      "learning_rate": 3e-05,
+      "loss": 2.8531,
+      "step": 1883
+    },
+    {
+      "epoch": 0.6108949416342413,
+      "grad_norm": 0.49049100279808044,
+      "learning_rate": 3e-05,
+      "loss": 3.1465,
+      "step": 1884
+    },
+    {
+      "epoch": 0.6112191958495461,
+      "grad_norm": 0.3728467524051666,
+      "learning_rate": 3e-05,
+      "loss": 2.4729,
+      "step": 1885
+    },
+    {
+      "epoch": 0.6115434500648509,
+      "grad_norm": 0.3906378448009491,
+      "learning_rate": 3e-05,
+      "loss": 2.3693,
+      "step": 1886
+    },
+    {
+      "epoch": 0.6118677042801557,
+      "grad_norm": 0.34296897053718567,
+      "learning_rate": 3e-05,
+      "loss": 2.3152,
+      "step": 1887
+    },
+    {
+      "epoch": 0.6121919584954605,
+      "grad_norm": 0.32202717661857605,
+      "learning_rate": 3e-05,
+      "loss": 2.3272,
+      "step": 1888
+    },
+    {
+      "epoch": 0.6125162127107653,
+      "grad_norm": 0.3092218041419983,
+      "learning_rate": 3e-05,
+      "loss": 2.2514,
+      "step": 1889
+    },
+    {
+      "epoch": 0.6128404669260701,
+      "grad_norm": 0.36877188086509705,
+      "learning_rate": 3e-05,
+      "loss": 2.4152,
+      "step": 1890
+    },
+    {
+      "epoch": 0.6131647211413749,
+      "grad_norm": 0.3579102158546448,
+      "learning_rate": 3e-05,
+      "loss": 2.2441,
+      "step": 1891
+    },
+    {
+      "epoch": 0.6134889753566797,
+      "grad_norm": 0.4953182339668274,
+      "learning_rate": 3e-05,
+      "loss": 2.8128,
+      "step": 1892
+    },
+    {
+      "epoch": 0.6138132295719845,
+      "grad_norm": 0.36372292041778564,
+      "learning_rate": 3e-05,
+      "loss": 2.1503,
+      "step": 1893
+    },
+    {
+      "epoch": 0.6141374837872893,
+      "grad_norm": 0.5083059668540955,
+      "learning_rate": 3e-05,
+      "loss": 3.3302,
+      "step": 1894
+    },
+    {
+      "epoch": 0.6144617380025941,
+      "grad_norm": 0.4489228129386902,
+      "learning_rate": 3e-05,
+      "loss": 3.006,
+      "step": 1895
+    },
+    {
+      "epoch": 0.6147859922178989,
+      "grad_norm": 0.3655955195426941,
+      "learning_rate": 3e-05,
+      "loss": 2.2958,
+      "step": 1896
+    },
+    {
+      "epoch": 0.6151102464332037,
+      "grad_norm": 0.4298546314239502,
+      "learning_rate": 3e-05,
+      "loss": 2.9143,
+      "step": 1897
+    },
+    {
+      "epoch": 0.6154345006485085,
+      "grad_norm": 0.3257444500923157,
+      "learning_rate": 3e-05,
+      "loss": 2.5111,
+      "step": 1898
+    },
+    {
+      "epoch": 0.6157587548638133,
+      "grad_norm": 0.4322950839996338,
+      "learning_rate": 3e-05,
+      "loss": 2.5782,
+      "step": 1899
+    },
+    {
+      "epoch": 0.6160830090791181,
+      "grad_norm": 0.45007121562957764,
+      "learning_rate": 3e-05,
+      "loss": 2.7076,
+      "step": 1900
+    },
+    {
+      "epoch": 0.6164072632944229,
+      "grad_norm": 0.3785910904407501,
+      "learning_rate": 3e-05,
+      "loss": 2.6603,
+      "step": 1901
+    },
+    {
+      "epoch": 0.6167315175097277,
+      "grad_norm": 0.5024528503417969,
+      "learning_rate": 3e-05,
+      "loss": 3.314,
+      "step": 1902
+    },
+    {
+      "epoch": 0.6170557717250325,
+      "grad_norm": 0.525749683380127,
+      "learning_rate": 3e-05,
+      "loss": 2.87,
+      "step": 1903
+    },
+    {
+      "epoch": 0.6173800259403373,
+      "grad_norm": 0.4436599910259247,
+      "learning_rate": 3e-05,
+      "loss": 2.6602,
+      "step": 1904
+    },
+    {
+      "epoch": 0.617704280155642,
+      "grad_norm": 0.36924636363983154,
+      "learning_rate": 3e-05,
+      "loss": 2.8306,
+      "step": 1905
+    },
+    {
+      "epoch": 0.6180285343709468,
+      "grad_norm": 0.32437270879745483,
+      "learning_rate": 3e-05,
+      "loss": 2.6463,
+      "step": 1906
+    },
+    {
+      "epoch": 0.6183527885862516,
+      "grad_norm": 0.5442020893096924,
+      "learning_rate": 3e-05,
+      "loss": 2.7378,
+      "step": 1907
+    },
+    {
+      "epoch": 0.6186770428015564,
+      "grad_norm": 0.41076865792274475,
+      "learning_rate": 3e-05,
+      "loss": 2.7562,
+      "step": 1908
+    },
+    {
+      "epoch": 0.6190012970168612,
+      "grad_norm": 0.5626446604728699,
+      "learning_rate": 3e-05,
+      "loss": 2.7256,
+      "step": 1909
+    },
+    {
+      "epoch": 0.619325551232166,
+      "grad_norm": 0.4098798632621765,
+      "learning_rate": 3e-05,
+      "loss": 2.8628,
+      "step": 1910
+    },
+    {
+      "epoch": 0.6196498054474708,
+      "grad_norm": 0.6764887571334839,
+      "learning_rate": 3e-05,
+      "loss": 3.0568,
+      "step": 1911
+    },
+    {
+      "epoch": 0.6199740596627756,
+      "grad_norm": 0.30608415603637695,
+      "learning_rate": 3e-05,
+      "loss": 1.8309,
+      "step": 1912
+    },
+    {
+      "epoch": 0.6202983138780804,
+      "grad_norm": 0.47648394107818604,
+      "learning_rate": 3e-05,
+      "loss": 2.6326,
+      "step": 1913
+    },
+    {
+      "epoch": 0.6206225680933852,
+      "grad_norm": 1.153546929359436,
+      "learning_rate": 3e-05,
+      "loss": 3.7124,
+      "step": 1914
+    },
+    {
+      "epoch": 0.62094682230869,
+      "grad_norm": 0.624366283416748,
+      "learning_rate": 3e-05,
+      "loss": 3.2423,
+      "step": 1915
+    },
+    {
+      "epoch": 0.6212710765239948,
+      "grad_norm": 0.5034214854240417,
+      "learning_rate": 3e-05,
+      "loss": 3.0326,
+      "step": 1916
+    },
+    {
+      "epoch": 0.6215953307392996,
+      "grad_norm": 0.399652898311615,
+      "learning_rate": 3e-05,
+      "loss": 2.3419,
+      "step": 1917
+    },
+    {
+      "epoch": 0.6219195849546044,
+      "grad_norm": 0.6442016959190369,
+      "learning_rate": 3e-05,
+      "loss": 3.0426,
+      "step": 1918
+    },
+    {
+      "epoch": 0.6222438391699092,
+      "grad_norm": 0.5885382890701294,
+      "learning_rate": 3e-05,
+      "loss": 2.4713,
+      "step": 1919
+    },
+    {
+      "epoch": 0.622568093385214,
+      "grad_norm": 0.877016544342041,
+      "learning_rate": 3e-05,
+      "loss": 3.283,
+      "step": 1920
+    },
+    {
+      "epoch": 0.6228923476005188,
+      "grad_norm": 0.7744568586349487,
+      "learning_rate": 3e-05,
+      "loss": 3.2969,
+      "step": 1921
+    },
+    {
+      "epoch": 0.6232166018158236,
+      "grad_norm": 0.42940449714660645,
+      "learning_rate": 3e-05,
+      "loss": 3.0253,
+      "step": 1922
+    },
+    {
+      "epoch": 0.6235408560311284,
+      "grad_norm": 0.5053513646125793,
+      "learning_rate": 3e-05,
+      "loss": 2.7503,
+      "step": 1923
+    },
+    {
+      "epoch": 0.6238651102464332,
+      "grad_norm": 0.5622748732566833,
+      "learning_rate": 3e-05,
+      "loss": 2.7897,
+      "step": 1924
+    },
+    {
+      "epoch": 0.624189364461738,
+      "grad_norm": 0.5655689239501953,
+      "learning_rate": 3e-05,
+      "loss": 3.0692,
+      "step": 1925
+    },
+    {
+      "epoch": 0.6245136186770428,
+      "grad_norm": 0.3833586573600769,
+      "learning_rate": 3e-05,
+      "loss": 2.2008,
+      "step": 1926
+    },
+    {
+      "epoch": 0.6248378728923476,
+      "grad_norm": 0.5508374571800232,
+      "learning_rate": 3e-05,
+      "loss": 2.862,
+      "step": 1927
+    },
+    {
+      "epoch": 0.6251621271076524,
+      "grad_norm": 0.35305479168891907,
+      "learning_rate": 3e-05,
+      "loss": 2.4606,
+      "step": 1928
+    },
+    {
+      "epoch": 0.6254863813229572,
+      "grad_norm": 0.4052164554595947,
+      "learning_rate": 3e-05,
+      "loss": 3.055,
+      "step": 1929
+    },
+    {
+      "epoch": 0.625810635538262,
+      "grad_norm": 0.5047911405563354,
+      "learning_rate": 3e-05,
+      "loss": 3.1236,
+      "step": 1930
+    },
+    {
+      "epoch": 0.6261348897535668,
+      "grad_norm": 0.5616687536239624,
+      "learning_rate": 3e-05,
+      "loss": 2.6085,
+      "step": 1931
+    },
+    {
+      "epoch": 0.6264591439688716,
+      "grad_norm": 0.3760230541229248,
+      "learning_rate": 3e-05,
+      "loss": 2.5259,
+      "step": 1932
+    },
+    {
+      "epoch": 0.6267833981841764,
+      "grad_norm": 0.4421178102493286,
+      "learning_rate": 3e-05,
+      "loss": 2.9778,
+      "step": 1933
+    },
+    {
+      "epoch": 0.6271076523994812,
+      "grad_norm": 0.4268551468849182,
+      "learning_rate": 3e-05,
+      "loss": 2.7774,
+      "step": 1934
+    },
+    {
+      "epoch": 0.627431906614786,
+      "grad_norm": 0.32855528593063354,
+      "learning_rate": 3e-05,
+      "loss": 2.4505,
+      "step": 1935
+    },
+    {
+      "epoch": 0.6277561608300908,
+      "grad_norm": 0.40525022149086,
+      "learning_rate": 3e-05,
+      "loss": 2.4319,
+      "step": 1936
+    },
+    {
+      "epoch": 0.6280804150453956,
+      "grad_norm": 0.43174007534980774,
+      "learning_rate": 3e-05,
+      "loss": 2.6989,
+      "step": 1937
+    },
+    {
+      "epoch": 0.6284046692607004,
+      "grad_norm": 0.37582069635391235,
+      "learning_rate": 3e-05,
+      "loss": 2.2659,
+      "step": 1938
+    },
+    {
+      "epoch": 0.6287289234760052,
+      "grad_norm": 0.42650651931762695,
+      "learning_rate": 3e-05,
+      "loss": 2.2998,
+      "step": 1939
+    },
+    {
+      "epoch": 0.62905317769131,
+      "grad_norm": 0.527214765548706,
+      "learning_rate": 3e-05,
+      "loss": 3.0215,
+      "step": 1940
+    },
+    {
+      "epoch": 0.6293774319066148,
+      "grad_norm": 0.6160122752189636,
+      "learning_rate": 3e-05,
+      "loss": 3.2926,
+      "step": 1941
+    },
+    {
+      "epoch": 0.6297016861219196,
+      "grad_norm": 0.349858820438385,
+      "learning_rate": 3e-05,
+      "loss": 2.3421,
+      "step": 1942
+    },
+    {
+      "epoch": 0.6300259403372244,
+      "grad_norm": 0.48673489689826965,
+      "learning_rate": 3e-05,
+      "loss": 2.5182,
+      "step": 1943
+    },
+    {
+      "epoch": 0.6303501945525292,
+      "grad_norm": 0.5374829769134521,
+      "learning_rate": 3e-05,
+      "loss": 2.8476,
+      "step": 1944
+    },
+    {
+      "epoch": 0.630674448767834,
+      "grad_norm": 0.475783109664917,
+      "learning_rate": 3e-05,
+      "loss": 2.8998,
+      "step": 1945
+    },
+    {
+      "epoch": 0.6309987029831388,
+      "grad_norm": 0.5441059470176697,
+      "learning_rate": 3e-05,
+      "loss": 2.5667,
+      "step": 1946
+    },
+    {
+      "epoch": 0.6313229571984436,
+      "grad_norm": 0.409647136926651,
+      "learning_rate": 3e-05,
+      "loss": 2.8417,
+      "step": 1947
+    },
+    {
+      "epoch": 0.6316472114137484,
+      "grad_norm": 0.35691744089126587,
+      "learning_rate": 3e-05,
+      "loss": 2.5975,
+      "step": 1948
+    },
+    {
+      "epoch": 0.6319714656290532,
+      "grad_norm": 0.4158698320388794,
+      "learning_rate": 3e-05,
+      "loss": 2.9566,
+      "step": 1949
+    },
+    {
+      "epoch": 0.632295719844358,
+      "grad_norm": 0.5195173025131226,
+      "learning_rate": 3e-05,
+      "loss": 3.3483,
+      "step": 1950
+    },
+    {
+      "epoch": 0.6326199740596627,
+      "grad_norm": 0.4284685254096985,
+      "learning_rate": 3e-05,
+      "loss": 2.5387,
+      "step": 1951
+    },
+    {
+      "epoch": 0.6329442282749675,
+      "grad_norm": 0.33130428194999695,
+      "learning_rate": 3e-05,
+      "loss": 2.1006,
+      "step": 1952
+    },
+    {
+      "epoch": 0.6332684824902723,
+      "grad_norm": 0.5425959229469299,
+      "learning_rate": 3e-05,
+      "loss": 3.1368,
+      "step": 1953
+    },
+    {
+      "epoch": 0.6335927367055771,
+      "grad_norm": 0.28663796186447144,
+      "learning_rate": 3e-05,
+      "loss": 2.0441,
+      "step": 1954
+    },
+    {
+      "epoch": 0.6339169909208819,
+      "grad_norm": 0.2953278720378876,
+      "learning_rate": 3e-05,
+      "loss": 2.4129,
+      "step": 1955
+    },
+    {
+      "epoch": 0.6342412451361867,
+      "grad_norm": 0.35210683941841125,
+      "learning_rate": 3e-05,
+      "loss": 2.365,
+      "step": 1956
+    },
+    {
+      "epoch": 0.6345654993514915,
+      "grad_norm": 0.2904893457889557,
+      "learning_rate": 3e-05,
+      "loss": 2.0575,
+      "step": 1957
+    },
+    {
+      "epoch": 0.6348897535667963,
+      "grad_norm": 0.4488282799720764,
+      "learning_rate": 3e-05,
+      "loss": 2.6498,
+      "step": 1958
+    },
+    {
+      "epoch": 0.6352140077821011,
+      "grad_norm": 0.4229547381401062,
+      "learning_rate": 3e-05,
+      "loss": 2.8708,
+      "step": 1959
+    },
+    {
+      "epoch": 0.6355382619974059,
+      "grad_norm": 0.42206376791000366,
+      "learning_rate": 3e-05,
+      "loss": 2.8304,
+      "step": 1960
+    },
+    {
+      "epoch": 0.6358625162127107,
+      "grad_norm": 0.31145966053009033,
+      "learning_rate": 3e-05,
+      "loss": 2.3744,
+      "step": 1961
+    },
+    {
+      "epoch": 0.6361867704280155,
+      "grad_norm": 0.6030656695365906,
+      "learning_rate": 3e-05,
+      "loss": 3.4918,
+      "step": 1962
+    },
+    {
+      "epoch": 0.6365110246433203,
+      "grad_norm": 0.3640313744544983,
+      "learning_rate": 3e-05,
+      "loss": 2.5369,
+      "step": 1963
+    },
+    {
+      "epoch": 0.6368352788586251,
+      "grad_norm": 0.3352667987346649,
+      "learning_rate": 3e-05,
+      "loss": 2.4025,
+      "step": 1964
+    },
+    {
+      "epoch": 0.6371595330739299,
+      "grad_norm": 0.37022265791893005,
+      "learning_rate": 3e-05,
+      "loss": 2.2273,
+      "step": 1965
+    },
+    {
+      "epoch": 0.6374837872892347,
+      "grad_norm": 0.6250036358833313,
+      "learning_rate": 3e-05,
+      "loss": 3.3048,
+      "step": 1966
+    },
+    {
+      "epoch": 0.6378080415045395,
+      "grad_norm": 0.3938817083835602,
+      "learning_rate": 3e-05,
+      "loss": 2.6439,
+      "step": 1967
+    },
+    {
+      "epoch": 0.6381322957198443,
+      "grad_norm": 0.4045473635196686,
+      "learning_rate": 3e-05,
+      "loss": 2.5563,
+      "step": 1968
+    },
+    {
+      "epoch": 0.6384565499351491,
+      "grad_norm": 0.42916080355644226,
+      "learning_rate": 3e-05,
+      "loss": 2.8844,
+      "step": 1969
+    },
+    {
+      "epoch": 0.6387808041504539,
+      "grad_norm": 0.3752673268318176,
+      "learning_rate": 3e-05,
+      "loss": 2.9647,
+      "step": 1970
+    },
+    {
+      "epoch": 0.6391050583657587,
+      "grad_norm": 0.44216930866241455,
+      "learning_rate": 3e-05,
+      "loss": 2.418,
+      "step": 1971
+    },
+    {
+      "epoch": 0.6394293125810635,
+      "grad_norm": 0.3817199468612671,
+      "learning_rate": 3e-05,
+      "loss": 2.5809,
+      "step": 1972
+    },
+    {
+      "epoch": 0.6397535667963683,
+      "grad_norm": 0.4236925542354584,
+      "learning_rate": 3e-05,
+      "loss": 2.5432,
+      "step": 1973
+    },
+    {
+      "epoch": 0.6400778210116731,
+      "grad_norm": 0.37849804759025574,
+      "learning_rate": 3e-05,
+      "loss": 2.9962,
+      "step": 1974
+    },
+    {
+      "epoch": 0.6404020752269779,
+      "grad_norm": 0.3307724893093109,
+      "learning_rate": 3e-05,
+      "loss": 2.4828,
+      "step": 1975
+    },
+    {
+      "epoch": 0.6407263294422828,
+      "grad_norm": 0.3574386239051819,
+      "learning_rate": 3e-05,
+      "loss": 2.4521,
+      "step": 1976
+    },
+    {
+      "epoch": 0.6410505836575876,
+      "grad_norm": 0.5531674027442932,
+      "learning_rate": 3e-05,
+      "loss": 3.0472,
+      "step": 1977
+    },
+    {
+      "epoch": 0.6413748378728924,
+      "grad_norm": 0.6509301662445068,
+      "learning_rate": 3e-05,
+      "loss": 3.0531,
+      "step": 1978
+    },
+    {
+      "epoch": 0.6416990920881972,
+      "grad_norm": 0.3731357455253601,
+      "learning_rate": 3e-05,
+      "loss": 2.6672,
+      "step": 1979
+    },
+    {
+      "epoch": 0.642023346303502,
+      "grad_norm": 0.6078703999519348,
+      "learning_rate": 3e-05,
+      "loss": 2.9038,
+      "step": 1980
+    },
+    {
+      "epoch": 0.6423476005188068,
+      "grad_norm": 0.4128943085670471,
+      "learning_rate": 3e-05,
+      "loss": 2.9723,
+      "step": 1981
+    },
+    {
+      "epoch": 0.6426718547341116,
+      "grad_norm": 0.3851010799407959,
+      "learning_rate": 3e-05,
+      "loss": 2.5037,
+      "step": 1982
+    },
+    {
+      "epoch": 0.6429961089494164,
+      "grad_norm": 0.4384444057941437,
+      "learning_rate": 3e-05,
+      "loss": 3.0162,
+      "step": 1983
+    },
+    {
+      "epoch": 0.6433203631647212,
+      "grad_norm": 0.33411678671836853,
+      "learning_rate": 3e-05,
+      "loss": 2.5953,
+      "step": 1984
+    },
+    {
+      "epoch": 0.643644617380026,
+      "grad_norm": 0.38261306285858154,
+      "learning_rate": 3e-05,
+      "loss": 2.5873,
+      "step": 1985
+    },
+    {
+      "epoch": 0.6439688715953308,
+      "grad_norm": 0.3520139753818512,
+      "learning_rate": 3e-05,
+      "loss": 2.5693,
+      "step": 1986
+    },
+    {
+      "epoch": 0.6442931258106356,
+      "grad_norm": 0.7422581911087036,
+      "learning_rate": 3e-05,
+      "loss": 3.3573,
+      "step": 1987
+    },
+    {
+      "epoch": 0.6446173800259404,
+      "grad_norm": 0.34916576743125916,
+      "learning_rate": 3e-05,
+      "loss": 2.6753,
+      "step": 1988
+    },
+    {
+      "epoch": 0.6449416342412452,
+      "grad_norm": 0.42346319556236267,
+      "learning_rate": 3e-05,
+      "loss": 2.7855,
+      "step": 1989
+    },
+    {
+      "epoch": 0.64526588845655,
+      "grad_norm": 0.3933003842830658,
+      "learning_rate": 3e-05,
+      "loss": 2.8663,
+      "step": 1990
+    },
+    {
+      "epoch": 0.6455901426718548,
+      "grad_norm": 0.5284514427185059,
+      "learning_rate": 3e-05,
+      "loss": 3.4116,
+      "step": 1991
+    },
+    {
+      "epoch": 0.6459143968871596,
+      "grad_norm": 0.6694428324699402,
+      "learning_rate": 3e-05,
+      "loss": 2.6113,
+      "step": 1992
+    },
+    {
+      "epoch": 0.6462386511024644,
+      "grad_norm": 0.32601287961006165,
+      "learning_rate": 3e-05,
+      "loss": 2.2002,
+      "step": 1993
+    },
+    {
+      "epoch": 0.6465629053177692,
+      "grad_norm": 0.31432750821113586,
+      "learning_rate": 3e-05,
+      "loss": 2.3661,
+      "step": 1994
+    },
+    {
+      "epoch": 0.646887159533074,
+      "grad_norm": 0.5121628642082214,
+      "learning_rate": 3e-05,
+      "loss": 3.1551,
+      "step": 1995
+    },
+    {
+      "epoch": 0.6472114137483788,
+      "grad_norm": 0.35157009959220886,
+      "learning_rate": 3e-05,
+      "loss": 2.3541,
+      "step": 1996
+    },
+    {
+      "epoch": 0.6475356679636836,
+      "grad_norm": 0.33471328020095825,
+      "learning_rate": 3e-05,
+      "loss": 2.5964,
+      "step": 1997
+    },
+    {
+      "epoch": 0.6478599221789884,
+      "grad_norm": 0.3737802505493164,
+      "learning_rate": 3e-05,
+      "loss": 2.379,
+      "step": 1998
+    },
+    {
+      "epoch": 0.6481841763942932,
+      "grad_norm": 0.34244945645332336,
+      "learning_rate": 3e-05,
+      "loss": 2.5145,
+      "step": 1999
+    },
+    {
+      "epoch": 0.648508430609598,
+      "grad_norm": 0.335250586271286,
+      "learning_rate": 3e-05,
+      "loss": 2.2545,
+      "step": 2000
+    },
+    {
+      "epoch": 0.6488326848249028,
+      "grad_norm": 0.38585853576660156,
+      "learning_rate": 3e-05,
+      "loss": 2.5132,
+      "step": 2001
+    },
+    {
+      "epoch": 0.6491569390402075,
+      "grad_norm": 0.3745160698890686,
+      "learning_rate": 3e-05,
+      "loss": 2.4972,
+      "step": 2002
+    },
+    {
+      "epoch": 0.6494811932555123,
+      "grad_norm": 0.4897535741329193,
+      "learning_rate": 3e-05,
+      "loss": 3.1826,
+      "step": 2003
+    },
+    {
+      "epoch": 0.6498054474708171,
+      "grad_norm": 0.39678463339805603,
+      "learning_rate": 3e-05,
+      "loss": 2.7533,
+      "step": 2004
+    },
+    {
+      "epoch": 0.6501297016861219,
+      "grad_norm": 0.45522651076316833,
+      "learning_rate": 3e-05,
+      "loss": 2.5241,
+      "step": 2005
+    },
+    {
+      "epoch": 0.6504539559014267,
+      "grad_norm": 0.3080872893333435,
+      "learning_rate": 3e-05,
+      "loss": 2.3126,
+      "step": 2006
+    },
+    {
+      "epoch": 0.6507782101167315,
+      "grad_norm": 0.621538519859314,
+      "learning_rate": 3e-05,
+      "loss": 3.9609,
+      "step": 2007
+    },
+    {
+      "epoch": 0.6511024643320363,
+      "grad_norm": 0.5174079537391663,
+      "learning_rate": 3e-05,
+      "loss": 3.1955,
+      "step": 2008
+    },
+    {
+      "epoch": 0.6514267185473411,
+      "grad_norm": 0.4216119945049286,
+      "learning_rate": 3e-05,
+      "loss": 3.1452,
+      "step": 2009
+    },
+    {
+      "epoch": 0.6517509727626459,
+      "grad_norm": 0.38078856468200684,
+      "learning_rate": 3e-05,
+      "loss": 2.816,
+      "step": 2010
+    },
+    {
+      "epoch": 0.6520752269779507,
+      "grad_norm": 0.4917844235897064,
+      "learning_rate": 3e-05,
+      "loss": 3.5397,
+      "step": 2011
+    },
+    {
+      "epoch": 0.6523994811932555,
+      "grad_norm": 0.38255494832992554,
+      "learning_rate": 3e-05,
+      "loss": 2.3652,
+      "step": 2012
+    },
+    {
+      "epoch": 0.6527237354085603,
+      "grad_norm": 0.471364289522171,
+      "learning_rate": 3e-05,
+      "loss": 2.6526,
+      "step": 2013
+    },
+    {
+      "epoch": 0.6530479896238651,
+      "grad_norm": 0.4261862337589264,
+      "learning_rate": 3e-05,
+      "loss": 2.6427,
+      "step": 2014
+    },
+    {
+      "epoch": 0.6533722438391699,
+      "grad_norm": 0.3270340859889984,
+      "learning_rate": 3e-05,
+      "loss": 2.4173,
+      "step": 2015
+    },
+    {
+      "epoch": 0.6536964980544747,
+      "grad_norm": 0.5101332068443298,
+      "learning_rate": 3e-05,
+      "loss": 3.1262,
+      "step": 2016
+    },
+    {
+      "epoch": 0.6540207522697795,
+      "grad_norm": 0.4176621735095978,
+      "learning_rate": 3e-05,
+      "loss": 2.9402,
+      "step": 2017
+    },
+    {
+      "epoch": 0.6543450064850843,
+      "grad_norm": 0.39699506759643555,
+      "learning_rate": 3e-05,
+      "loss": 2.7203,
+      "step": 2018
+    },
+    {
+      "epoch": 0.6546692607003891,
+      "grad_norm": 0.3972765803337097,
+      "learning_rate": 3e-05,
+      "loss": 2.5155,
+      "step": 2019
+    },
+    {
+      "epoch": 0.6549935149156939,
+      "grad_norm": 0.4655422270298004,
+      "learning_rate": 3e-05,
+      "loss": 2.6305,
+      "step": 2020
+    },
+    {
+      "epoch": 0.6553177691309987,
+      "grad_norm": 0.5240328311920166,
+      "learning_rate": 3e-05,
+      "loss": 3.0925,
+      "step": 2021
+    },
+    {
+      "epoch": 0.6556420233463035,
+      "grad_norm": 0.3926469087600708,
+      "learning_rate": 3e-05,
+      "loss": 2.7062,
+      "step": 2022
+    },
+    {
+      "epoch": 0.6559662775616083,
+      "grad_norm": 0.3127439618110657,
+      "learning_rate": 3e-05,
+      "loss": 2.1179,
+      "step": 2023
+    },
+    {
+      "epoch": 0.6562905317769131,
+      "grad_norm": 0.3135978579521179,
+      "learning_rate": 3e-05,
+      "loss": 2.2009,
+      "step": 2024
+    },
+    {
+      "epoch": 0.6566147859922179,
+      "grad_norm": 0.36984071135520935,
+      "learning_rate": 3e-05,
+      "loss": 2.4296,
+      "step": 2025
+    },
+    {
+      "epoch": 0.6569390402075227,
+      "grad_norm": 0.3011988699436188,
+      "learning_rate": 3e-05,
+      "loss": 2.1664,
+      "step": 2026
+    },
+    {
+      "epoch": 0.6572632944228275,
+      "grad_norm": 0.4938504695892334,
+      "learning_rate": 3e-05,
+      "loss": 2.7952,
+      "step": 2027
+    },
+    {
+      "epoch": 0.6575875486381323,
+      "grad_norm": 0.32471299171447754,
+      "learning_rate": 3e-05,
+      "loss": 2.5011,
+      "step": 2028
+    },
+    {
+      "epoch": 0.6579118028534371,
+      "grad_norm": 0.4755270779132843,
+      "learning_rate": 3e-05,
+      "loss": 2.8034,
+      "step": 2029
+    },
+    {
+      "epoch": 0.6582360570687419,
+      "grad_norm": 0.6197144389152527,
+      "learning_rate": 3e-05,
+      "loss": 3.1553,
+      "step": 2030
+    },
+    {
+      "epoch": 0.6585603112840467,
+      "grad_norm": 0.39791038632392883,
+      "learning_rate": 3e-05,
+      "loss": 2.8475,
+      "step": 2031
+    },
+    {
+      "epoch": 0.6588845654993515,
+      "grad_norm": 0.39058470726013184,
+      "learning_rate": 3e-05,
+      "loss": 2.4291,
+      "step": 2032
+    },
+    {
+      "epoch": 0.6592088197146563,
+      "grad_norm": 0.36039480566978455,
+      "learning_rate": 3e-05,
+      "loss": 2.2671,
+      "step": 2033
+    },
+    {
+      "epoch": 0.6595330739299611,
+      "grad_norm": 0.48510482907295227,
+      "learning_rate": 3e-05,
+      "loss": 2.6367,
+      "step": 2034
+    },
+    {
+      "epoch": 0.6598573281452659,
+      "grad_norm": 0.45733609795570374,
+      "learning_rate": 3e-05,
+      "loss": 2.6456,
+      "step": 2035
+    },
+    {
+      "epoch": 0.6601815823605707,
+      "grad_norm": 0.3553083837032318,
+      "learning_rate": 3e-05,
+      "loss": 2.3556,
+      "step": 2036
+    },
+    {
+      "epoch": 0.6605058365758755,
+      "grad_norm": 0.3381214737892151,
+      "learning_rate": 3e-05,
+      "loss": 2.3199,
+      "step": 2037
+    },
+    {
+      "epoch": 0.6608300907911803,
+      "grad_norm": 0.4676477909088135,
+      "learning_rate": 3e-05,
+      "loss": 2.6896,
+      "step": 2038
+    },
+    {
+      "epoch": 0.6611543450064851,
+      "grad_norm": 0.5896043181419373,
+      "learning_rate": 3e-05,
+      "loss": 2.9777,
+      "step": 2039
+    },
+    {
+      "epoch": 0.6614785992217899,
+      "grad_norm": 0.41078928112983704,
+      "learning_rate": 3e-05,
+      "loss": 2.919,
+      "step": 2040
+    },
+    {
+      "epoch": 0.6618028534370947,
+      "grad_norm": 0.40259432792663574,
+      "learning_rate": 3e-05,
+      "loss": 2.2033,
+      "step": 2041
+    },
+    {
+      "epoch": 0.6621271076523995,
+      "grad_norm": 0.4518181383609772,
+      "learning_rate": 3e-05,
+      "loss": 2.7158,
+      "step": 2042
+    },
+    {
+      "epoch": 0.6624513618677043,
+      "grad_norm": 0.3952997028827667,
+      "learning_rate": 3e-05,
+      "loss": 3.2649,
+      "step": 2043
+    },
+    {
+      "epoch": 0.6627756160830091,
+      "grad_norm": 0.5194694995880127,
+      "learning_rate": 3e-05,
+      "loss": 3.3927,
+      "step": 2044
+    },
+    {
+      "epoch": 0.6630998702983139,
+      "grad_norm": 0.4222572147846222,
+      "learning_rate": 3e-05,
+      "loss": 2.8884,
+      "step": 2045
+    },
+    {
+      "epoch": 0.6634241245136187,
+      "grad_norm": 0.3965514302253723,
+      "learning_rate": 3e-05,
+      "loss": 2.2793,
+      "step": 2046
+    },
+    {
+      "epoch": 0.6637483787289234,
+      "grad_norm": 0.5106009840965271,
+      "learning_rate": 3e-05,
+      "loss": 2.9426,
+      "step": 2047
+    },
+    {
+      "epoch": 0.6640726329442282,
+      "grad_norm": 0.5350344777107239,
+      "learning_rate": 3e-05,
+      "loss": 2.756,
+      "step": 2048
+    },
+    {
+      "epoch": 0.664396887159533,
+      "grad_norm": 0.29620060324668884,
+      "learning_rate": 3e-05,
+      "loss": 2.2367,
+      "step": 2049
+    },
+    {
+      "epoch": 0.6647211413748378,
+      "grad_norm": 0.3392261862754822,
+      "learning_rate": 3e-05,
+      "loss": 2.0815,
+      "step": 2050
+    },
+    {
+      "epoch": 0.6650453955901426,
+      "grad_norm": 0.7223390936851501,
+      "learning_rate": 3e-05,
+      "loss": 2.9901,
+      "step": 2051
+    },
+    {
+      "epoch": 0.6653696498054474,
+      "grad_norm": 0.676921546459198,
+      "learning_rate": 3e-05,
+      "loss": 3.2757,
+      "step": 2052
+    },
+    {
+      "epoch": 0.6656939040207522,
+      "grad_norm": 0.3624151945114136,
+      "learning_rate": 3e-05,
+      "loss": 2.8266,
+      "step": 2053
+    },
+    {
+      "epoch": 0.666018158236057,
+      "grad_norm": 0.4907524883747101,
+      "learning_rate": 3e-05,
+      "loss": 2.5443,
+      "step": 2054
+    },
+    {
+      "epoch": 0.6663424124513618,
+      "grad_norm": 0.5350075960159302,
+      "learning_rate": 3e-05,
+      "loss": 2.3529,
+      "step": 2055
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.6155858039855957,
+      "learning_rate": 3e-05,
+      "loss": 2.6776,
+      "step": 2056
+    },
+    {
+      "epoch": 0.6669909208819714,
+      "grad_norm": 0.5037629008293152,
+      "learning_rate": 3e-05,
+      "loss": 2.7806,
+      "step": 2057
+    },
+    {
+      "epoch": 0.6673151750972762,
+      "grad_norm": 0.5505346059799194,
+      "learning_rate": 3e-05,
+      "loss": 3.049,
+      "step": 2058
+    },
+    {
+      "epoch": 0.667639429312581,
+      "grad_norm": 0.45099619030952454,
+      "learning_rate": 3e-05,
+      "loss": 3.0009,
+      "step": 2059
+    },
+    {
+      "epoch": 0.6679636835278858,
+      "grad_norm": 0.592972993850708,
+      "learning_rate": 3e-05,
+      "loss": 2.6653,
+      "step": 2060
+    },
+    {
+      "epoch": 0.6682879377431906,
+      "grad_norm": 0.6917620301246643,
+      "learning_rate": 3e-05,
+      "loss": 3.3586,
+      "step": 2061
+    },
+    {
+      "epoch": 0.6686121919584954,
+      "grad_norm": 0.37479451298713684,
+      "learning_rate": 3e-05,
+      "loss": 2.4369,
+      "step": 2062
+    },
+    {
+      "epoch": 0.6689364461738002,
+      "grad_norm": 0.3480730950832367,
+      "learning_rate": 3e-05,
+      "loss": 2.3885,
+      "step": 2063
+    },
+    {
+      "epoch": 0.669260700389105,
+      "grad_norm": 0.5690239071846008,
+      "learning_rate": 3e-05,
+      "loss": 2.705,
+      "step": 2064
+    },
+    {
+      "epoch": 0.6695849546044098,
+      "grad_norm": 0.4905945360660553,
+      "learning_rate": 3e-05,
+      "loss": 3.1386,
+      "step": 2065
+    },
+    {
+      "epoch": 0.6699092088197146,
+      "grad_norm": 0.41636893153190613,
+      "learning_rate": 3e-05,
+      "loss": 2.3486,
+      "step": 2066
+    },
+    {
+      "epoch": 0.6702334630350194,
+      "grad_norm": 0.3662385940551758,
+      "learning_rate": 3e-05,
+      "loss": 2.5143,
+      "step": 2067
+    },
+    {
+      "epoch": 0.6705577172503242,
+      "grad_norm": 0.4015274941921234,
+      "learning_rate": 3e-05,
+      "loss": 2.3864,
+      "step": 2068
+    },
+    {
+      "epoch": 0.670881971465629,
+      "grad_norm": 0.5971083641052246,
+      "learning_rate": 3e-05,
+      "loss": 3.4908,
+      "step": 2069
+    },
+    {
+      "epoch": 0.6712062256809338,
+      "grad_norm": 0.5574107766151428,
+      "learning_rate": 3e-05,
+      "loss": 2.9616,
+      "step": 2070
+    },
+    {
+      "epoch": 0.6715304798962386,
+      "grad_norm": 0.36029040813446045,
+      "learning_rate": 3e-05,
+      "loss": 2.3353,
+      "step": 2071
+    },
+    {
+      "epoch": 0.6718547341115434,
+      "grad_norm": 0.3967643976211548,
+      "learning_rate": 3e-05,
+      "loss": 3.0331,
+      "step": 2072
+    },
+    {
+      "epoch": 0.6721789883268483,
+      "grad_norm": 0.4175904393196106,
+      "learning_rate": 3e-05,
+      "loss": 3.0205,
+      "step": 2073
+    },
+    {
+      "epoch": 0.6725032425421531,
+      "grad_norm": 0.32108941674232483,
+      "learning_rate": 3e-05,
+      "loss": 2.281,
+      "step": 2074
+    },
+    {
+      "epoch": 0.6728274967574579,
+      "grad_norm": 0.32783836126327515,
+      "learning_rate": 3e-05,
+      "loss": 2.4277,
+      "step": 2075
+    },
+    {
+      "epoch": 0.6731517509727627,
+      "grad_norm": 0.4507606625556946,
+      "learning_rate": 3e-05,
+      "loss": 2.6053,
+      "step": 2076
+    },
+    {
+      "epoch": 0.6734760051880675,
+      "grad_norm": 0.38450488448143005,
+      "learning_rate": 3e-05,
+      "loss": 2.483,
+      "step": 2077
+    },
+    {
+      "epoch": 0.6738002594033723,
+      "grad_norm": 0.4054701626300812,
+      "learning_rate": 3e-05,
+      "loss": 3.3597,
+      "step": 2078
+    },
+    {
+      "epoch": 0.6741245136186771,
+      "grad_norm": 0.3311620056629181,
+      "learning_rate": 3e-05,
+      "loss": 2.2422,
+      "step": 2079
+    },
+    {
+      "epoch": 0.6744487678339819,
+      "grad_norm": 0.37545108795166016,
+      "learning_rate": 3e-05,
+      "loss": 2.7883,
+      "step": 2080
+    },
+    {
+      "epoch": 0.6747730220492867,
+      "grad_norm": 0.48623767495155334,
+      "learning_rate": 3e-05,
+      "loss": 2.8242,
+      "step": 2081
+    },
+    {
+      "epoch": 0.6750972762645915,
+      "grad_norm": 0.610133171081543,
+      "learning_rate": 3e-05,
+      "loss": 3.3754,
+      "step": 2082
+    },
+    {
+      "epoch": 0.6754215304798963,
+      "grad_norm": 0.43646717071533203,
+      "learning_rate": 3e-05,
+      "loss": 2.5643,
+      "step": 2083
+    },
+    {
+      "epoch": 0.6757457846952011,
+      "grad_norm": 0.44021180272102356,
+      "learning_rate": 3e-05,
+      "loss": 2.7631,
+      "step": 2084
+    },
+    {
+      "epoch": 0.6760700389105059,
+      "grad_norm": 0.39318642020225525,
+      "learning_rate": 3e-05,
+      "loss": 2.7538,
+      "step": 2085
+    },
+    {
+      "epoch": 0.6763942931258107,
+      "grad_norm": 0.5985152721405029,
+      "learning_rate": 3e-05,
+      "loss": 3.3536,
+      "step": 2086
+    },
+    {
+      "epoch": 0.6767185473411155,
+      "grad_norm": 0.5162497162818909,
+      "learning_rate": 3e-05,
+      "loss": 2.6858,
+      "step": 2087
+    },
+    {
+      "epoch": 0.6770428015564203,
+      "grad_norm": 0.2914859652519226,
+      "learning_rate": 3e-05,
+      "loss": 2.1854,
+      "step": 2088
+    },
+    {
+      "epoch": 0.6773670557717251,
+      "grad_norm": 0.3836095631122589,
+      "learning_rate": 3e-05,
+      "loss": 2.5331,
+      "step": 2089
+    },
+    {
+      "epoch": 0.6776913099870299,
+      "grad_norm": 0.3934173285961151,
+      "learning_rate": 3e-05,
+      "loss": 2.4078,
+      "step": 2090
+    },
+    {
+      "epoch": 0.6780155642023347,
+      "grad_norm": 0.5046660304069519,
+      "learning_rate": 3e-05,
+      "loss": 2.8043,
+      "step": 2091
+    },
+    {
+      "epoch": 0.6783398184176395,
+      "grad_norm": 0.40841144323349,
+      "learning_rate": 3e-05,
+      "loss": 3.2558,
+      "step": 2092
+    },
+    {
+      "epoch": 0.6786640726329443,
+      "grad_norm": 0.3523498475551605,
+      "learning_rate": 3e-05,
+      "loss": 2.5751,
+      "step": 2093
+    },
+    {
+      "epoch": 0.6789883268482491,
+      "grad_norm": 0.4034532606601715,
+      "learning_rate": 3e-05,
+      "loss": 2.7769,
+      "step": 2094
+    },
+    {
+      "epoch": 0.6793125810635539,
+      "grad_norm": 0.45932212471961975,
+      "learning_rate": 3e-05,
+      "loss": 2.9381,
+      "step": 2095
+    },
+    {
+      "epoch": 0.6796368352788587,
+      "grad_norm": 0.28320348262786865,
+      "learning_rate": 3e-05,
+      "loss": 2.2519,
+      "step": 2096
+    },
+    {
+      "epoch": 0.6799610894941635,
+      "grad_norm": 0.4764149785041809,
+      "learning_rate": 3e-05,
+      "loss": 3.6065,
+      "step": 2097
+    },
+    {
+      "epoch": 0.6802853437094682,
+      "grad_norm": 0.4151756167411804,
+      "learning_rate": 3e-05,
+      "loss": 3.1177,
+      "step": 2098
+    },
+    {
+      "epoch": 0.680609597924773,
+      "grad_norm": 0.3596373498439789,
+      "learning_rate": 3e-05,
+      "loss": 3.0813,
+      "step": 2099
+    },
+    {
+      "epoch": 0.6809338521400778,
+      "grad_norm": 0.31587886810302734,
+      "learning_rate": 3e-05,
+      "loss": 2.2611,
+      "step": 2100
+    },
+    {
+      "epoch": 0.6812581063553826,
+      "grad_norm": 0.7640262842178345,
+      "learning_rate": 3e-05,
+      "loss": 3.7901,
+      "step": 2101
+    },
+    {
+      "epoch": 0.6815823605706874,
+      "grad_norm": 0.3791813850402832,
+      "learning_rate": 3e-05,
+      "loss": 2.8628,
+      "step": 2102
+    },
+    {
+      "epoch": 0.6819066147859922,
+      "grad_norm": 0.371419757604599,
+      "learning_rate": 3e-05,
+      "loss": 2.3305,
+      "step": 2103
+    },
+    {
+      "epoch": 0.682230869001297,
+      "grad_norm": 0.5360199809074402,
+      "learning_rate": 3e-05,
+      "loss": 3.2442,
+      "step": 2104
+    },
+    {
+      "epoch": 0.6825551232166018,
+      "grad_norm": 0.41658174991607666,
+      "learning_rate": 3e-05,
+      "loss": 3.0067,
+      "step": 2105
+    },
+    {
+      "epoch": 0.6828793774319066,
+      "grad_norm": 0.39821845293045044,
+      "learning_rate": 3e-05,
+      "loss": 2.7803,
+      "step": 2106
+    },
+    {
+      "epoch": 0.6832036316472114,
+      "grad_norm": 0.3367036283016205,
+      "learning_rate": 3e-05,
+      "loss": 2.8061,
+      "step": 2107
+    },
+    {
+      "epoch": 0.6835278858625162,
+      "grad_norm": 0.29532721638679504,
+      "learning_rate": 3e-05,
+      "loss": 2.0014,
+      "step": 2108
+    },
+    {
+      "epoch": 0.683852140077821,
+      "grad_norm": 0.3954981863498688,
+      "learning_rate": 3e-05,
+      "loss": 2.6832,
+      "step": 2109
+    },
+    {
+      "epoch": 0.6841763942931258,
+      "grad_norm": 0.2741972506046295,
+      "learning_rate": 3e-05,
+      "loss": 2.1337,
+      "step": 2110
+    },
+    {
+      "epoch": 0.6845006485084306,
+      "grad_norm": 0.43310654163360596,
+      "learning_rate": 3e-05,
+      "loss": 2.9378,
+      "step": 2111
+    },
+    {
+      "epoch": 0.6848249027237354,
+      "grad_norm": 0.4541928768157959,
+      "learning_rate": 3e-05,
+      "loss": 3.2865,
+      "step": 2112
+    },
+    {
+      "epoch": 0.6851491569390402,
+      "grad_norm": 0.5129266381263733,
+      "learning_rate": 3e-05,
+      "loss": 3.1288,
+      "step": 2113
+    },
+    {
+      "epoch": 0.685473411154345,
+      "grad_norm": 0.44395220279693604,
+      "learning_rate": 3e-05,
+      "loss": 2.8117,
+      "step": 2114
+    },
+    {
+      "epoch": 0.6857976653696498,
+      "grad_norm": 0.4852653741836548,
+      "learning_rate": 3e-05,
+      "loss": 2.9283,
+      "step": 2115
+    },
+    {
+      "epoch": 0.6861219195849546,
+      "grad_norm": 0.41069507598876953,
+      "learning_rate": 3e-05,
+      "loss": 2.5661,
+      "step": 2116
+    },
+    {
+      "epoch": 0.6864461738002594,
+      "grad_norm": 0.38821130990982056,
+      "learning_rate": 3e-05,
+      "loss": 2.5734,
+      "step": 2117
+    },
+    {
+      "epoch": 0.6867704280155642,
+      "grad_norm": 0.34163662791252136,
+      "learning_rate": 3e-05,
+      "loss": 2.4281,
+      "step": 2118
+    },
+    {
+      "epoch": 0.687094682230869,
+      "grad_norm": 0.3441922962665558,
+      "learning_rate": 3e-05,
+      "loss": 2.1771,
+      "step": 2119
+    },
+    {
+      "epoch": 0.6874189364461738,
+      "grad_norm": 0.31795835494995117,
+      "learning_rate": 3e-05,
+      "loss": 2.3901,
+      "step": 2120
+    },
+    {
+      "epoch": 0.6877431906614786,
+      "grad_norm": 0.3317286968231201,
+      "learning_rate": 3e-05,
+      "loss": 2.4665,
+      "step": 2121
+    },
+    {
+      "epoch": 0.6880674448767834,
+      "grad_norm": 0.4926995038986206,
+      "learning_rate": 3e-05,
+      "loss": 2.8115,
+      "step": 2122
+    },
+    {
+      "epoch": 0.6883916990920882,
+      "grad_norm": 0.38106173276901245,
+      "learning_rate": 3e-05,
+      "loss": 2.8997,
+      "step": 2123
+    },
+    {
+      "epoch": 0.688715953307393,
+      "grad_norm": 0.4344022870063782,
+      "learning_rate": 3e-05,
+      "loss": 3.0173,
+      "step": 2124
+    },
+    {
+      "epoch": 0.6890402075226978,
+      "grad_norm": 0.2747979462146759,
+      "learning_rate": 3e-05,
+      "loss": 2.0272,
+      "step": 2125
+    },
+    {
+      "epoch": 0.6893644617380026,
+      "grad_norm": 0.38792315125465393,
+      "learning_rate": 3e-05,
+      "loss": 2.7867,
+      "step": 2126
+    },
+    {
+      "epoch": 0.6896887159533074,
+      "grad_norm": 0.3658658564090729,
+      "learning_rate": 3e-05,
+      "loss": 2.6383,
+      "step": 2127
+    },
+    {
+      "epoch": 0.6900129701686122,
+      "grad_norm": 0.48567116260528564,
+      "learning_rate": 3e-05,
+      "loss": 2.8529,
+      "step": 2128
+    },
+    {
+      "epoch": 0.690337224383917,
+      "grad_norm": 0.34367698431015015,
+      "learning_rate": 3e-05,
+      "loss": 2.2375,
+      "step": 2129
+    },
+    {
+      "epoch": 0.6906614785992218,
+      "grad_norm": 0.28394967317581177,
+      "learning_rate": 3e-05,
+      "loss": 1.9702,
+      "step": 2130
+    },
+    {
+      "epoch": 0.6909857328145266,
+      "grad_norm": 0.5498993992805481,
+      "learning_rate": 3e-05,
+      "loss": 3.6622,
+      "step": 2131
+    },
+    {
+      "epoch": 0.6913099870298314,
+      "grad_norm": 0.41839826107025146,
+      "learning_rate": 3e-05,
+      "loss": 2.8487,
+      "step": 2132
+    },
+    {
+      "epoch": 0.6916342412451362,
+      "grad_norm": 0.28884488344192505,
+      "learning_rate": 3e-05,
+      "loss": 2.3708,
+      "step": 2133
+    },
+    {
+      "epoch": 0.691958495460441,
+      "grad_norm": 0.3998156189918518,
+      "learning_rate": 3e-05,
+      "loss": 2.4943,
+      "step": 2134
+    },
+    {
+      "epoch": 0.6922827496757458,
+      "grad_norm": 0.358763188123703,
+      "learning_rate": 3e-05,
+      "loss": 2.4489,
+      "step": 2135
+    },
+    {
+      "epoch": 0.6926070038910506,
+      "grad_norm": 0.47304004430770874,
+      "learning_rate": 3e-05,
+      "loss": 2.8686,
+      "step": 2136
+    },
+    {
+      "epoch": 0.6929312581063554,
+      "grad_norm": 0.3991437256336212,
+      "learning_rate": 3e-05,
+      "loss": 3.0019,
+      "step": 2137
+    },
+    {
+      "epoch": 0.6932555123216602,
+      "grad_norm": 0.3488306403160095,
+      "learning_rate": 3e-05,
+      "loss": 2.7727,
+      "step": 2138
+    },
+    {
+      "epoch": 0.693579766536965,
+      "grad_norm": 0.4371810853481293,
+      "learning_rate": 3e-05,
+      "loss": 2.8583,
+      "step": 2139
+    },
+    {
+      "epoch": 0.6939040207522698,
+      "grad_norm": 0.3237660825252533,
+      "learning_rate": 3e-05,
+      "loss": 2.16,
+      "step": 2140
+    },
+    {
+      "epoch": 0.6942282749675746,
+      "grad_norm": 0.46025967597961426,
+      "learning_rate": 3e-05,
+      "loss": 3.3971,
+      "step": 2141
+    },
+    {
+      "epoch": 0.6945525291828794,
+      "grad_norm": 0.35656699538230896,
+      "learning_rate": 3e-05,
+      "loss": 2.2764,
+      "step": 2142
+    },
+    {
+      "epoch": 0.6948767833981842,
+      "grad_norm": 0.5249372720718384,
+      "learning_rate": 3e-05,
+      "loss": 3.3058,
+      "step": 2143
+    },
+    {
+      "epoch": 0.695201037613489,
+      "grad_norm": 0.3431463837623596,
+      "learning_rate": 3e-05,
+      "loss": 2.4565,
+      "step": 2144
+    },
+    {
+      "epoch": 0.6955252918287937,
+      "grad_norm": 0.31208473443984985,
+      "learning_rate": 3e-05,
+      "loss": 2.2907,
+      "step": 2145
+    },
+    {
+      "epoch": 0.6958495460440985,
+      "grad_norm": 0.45563024282455444,
+      "learning_rate": 3e-05,
+      "loss": 2.7563,
+      "step": 2146
+    },
+    {
+      "epoch": 0.6961738002594033,
+      "grad_norm": 0.37553471326828003,
+      "learning_rate": 3e-05,
+      "loss": 2.3212,
+      "step": 2147
+    },
+    {
+      "epoch": 0.6964980544747081,
+      "grad_norm": 0.5518587827682495,
+      "learning_rate": 3e-05,
+      "loss": 3.3779,
+      "step": 2148
+    },
+    {
+      "epoch": 0.6968223086900129,
+      "grad_norm": 0.29581746459007263,
+      "learning_rate": 3e-05,
+      "loss": 2.1159,
+      "step": 2149
+    },
+    {
+      "epoch": 0.6971465629053177,
+      "grad_norm": 0.3965296745300293,
+      "learning_rate": 3e-05,
+      "loss": 2.9011,
+      "step": 2150
+    },
+    {
+      "epoch": 0.6974708171206225,
+      "grad_norm": 0.2598591148853302,
+      "learning_rate": 3e-05,
+      "loss": 2.0575,
+      "step": 2151
+    },
+    {
+      "epoch": 0.6977950713359273,
+      "grad_norm": 0.4065677523612976,
+      "learning_rate": 3e-05,
+      "loss": 2.8707,
+      "step": 2152
+    },
+    {
+      "epoch": 0.6981193255512321,
+      "grad_norm": 0.3236706256866455,
+      "learning_rate": 3e-05,
+      "loss": 2.4909,
+      "step": 2153
+    },
+    {
+      "epoch": 0.6984435797665369,
+      "grad_norm": 0.4086410403251648,
+      "learning_rate": 3e-05,
+      "loss": 2.8467,
+      "step": 2154
+    },
+    {
+      "epoch": 0.6987678339818417,
+      "grad_norm": 0.5302712917327881,
+      "learning_rate": 3e-05,
+      "loss": 2.8792,
+      "step": 2155
+    },
+    {
+      "epoch": 0.6990920881971465,
+      "grad_norm": 0.4546435475349426,
+      "learning_rate": 3e-05,
+      "loss": 2.8953,
+      "step": 2156
+    },
+    {
+      "epoch": 0.6994163424124513,
+      "grad_norm": 0.5234642028808594,
+      "learning_rate": 3e-05,
+      "loss": 2.7717,
+      "step": 2157
+    },
+    {
+      "epoch": 0.6997405966277561,
+      "grad_norm": 0.388166218996048,
+      "learning_rate": 3e-05,
+      "loss": 2.7164,
+      "step": 2158
+    },
+    {
+      "epoch": 0.7000648508430609,
+      "grad_norm": 0.3765428364276886,
+      "learning_rate": 3e-05,
+      "loss": 2.5951,
+      "step": 2159
+    },
+    {
+      "epoch": 0.7003891050583657,
+      "grad_norm": 0.4231514632701874,
+      "learning_rate": 3e-05,
+      "loss": 2.6728,
+      "step": 2160
+    },
+    {
+      "epoch": 0.7007133592736705,
+      "grad_norm": 0.42784833908081055,
+      "learning_rate": 3e-05,
+      "loss": 2.149,
+      "step": 2161
+    },
+    {
+      "epoch": 0.7010376134889753,
+      "grad_norm": 0.5017656683921814,
+      "learning_rate": 3e-05,
+      "loss": 2.7401,
+      "step": 2162
+    },
+    {
+      "epoch": 0.7013618677042801,
+      "grad_norm": 0.320853590965271,
+      "learning_rate": 3e-05,
+      "loss": 2.2703,
+      "step": 2163
+    },
+    {
+      "epoch": 0.7016861219195849,
+      "grad_norm": 0.6279693841934204,
+      "learning_rate": 3e-05,
+      "loss": 3.0691,
+      "step": 2164
+    },
+    {
+      "epoch": 0.7020103761348897,
+      "grad_norm": 0.4429411292076111,
+      "learning_rate": 3e-05,
+      "loss": 2.9741,
+      "step": 2165
+    },
+    {
+      "epoch": 0.7023346303501945,
+      "grad_norm": 0.3254571557044983,
+      "learning_rate": 3e-05,
+      "loss": 2.2855,
+      "step": 2166
+    },
+    {
+      "epoch": 0.7026588845654993,
+      "grad_norm": 0.4367378354072571,
+      "learning_rate": 3e-05,
+      "loss": 2.605,
+      "step": 2167
+    },
+    {
+      "epoch": 0.7029831387808041,
+      "grad_norm": 0.3365830183029175,
+      "learning_rate": 3e-05,
+      "loss": 2.308,
+      "step": 2168
+    },
+    {
+      "epoch": 0.703307392996109,
+      "grad_norm": 0.4333559274673462,
+      "learning_rate": 3e-05,
+      "loss": 2.9917,
+      "step": 2169
+    },
+    {
+      "epoch": 0.7036316472114138,
+      "grad_norm": 0.4595292806625366,
+      "learning_rate": 3e-05,
+      "loss": 2.7002,
+      "step": 2170
+    },
+    {
+      "epoch": 0.7039559014267186,
+      "grad_norm": 0.4091005325317383,
+      "learning_rate": 3e-05,
+      "loss": 2.8877,
+      "step": 2171
+    },
+    {
+      "epoch": 0.7042801556420234,
+      "grad_norm": 0.36796268820762634,
+      "learning_rate": 3e-05,
+      "loss": 2.2107,
+      "step": 2172
+    },
+    {
+      "epoch": 0.7046044098573282,
+      "grad_norm": 0.48546653985977173,
+      "learning_rate": 3e-05,
+      "loss": 3.1653,
+      "step": 2173
+    },
+    {
+      "epoch": 0.704928664072633,
+      "grad_norm": 0.45017778873443604,
+      "learning_rate": 3e-05,
+      "loss": 3.0305,
+      "step": 2174
+    },
+    {
+      "epoch": 0.7052529182879378,
+      "grad_norm": 0.4081873297691345,
+      "learning_rate": 3e-05,
+      "loss": 2.8053,
+      "step": 2175
+    },
+    {
+      "epoch": 0.7055771725032426,
+      "grad_norm": 0.3704334795475006,
+      "learning_rate": 3e-05,
+      "loss": 2.6582,
+      "step": 2176
+    },
+    {
+      "epoch": 0.7059014267185474,
+      "grad_norm": 0.4320659637451172,
+      "learning_rate": 3e-05,
+      "loss": 3.2509,
+      "step": 2177
+    },
+    {
+      "epoch": 0.7062256809338522,
+      "grad_norm": 0.5185019969940186,
+      "learning_rate": 3e-05,
+      "loss": 3.4721,
+      "step": 2178
+    },
+    {
+      "epoch": 0.706549935149157,
+      "grad_norm": 0.5227078199386597,
+      "learning_rate": 3e-05,
+      "loss": 3.5737,
+      "step": 2179
+    },
+    {
+      "epoch": 0.7068741893644618,
+      "grad_norm": 0.4452575743198395,
+      "learning_rate": 3e-05,
+      "loss": 2.9219,
+      "step": 2180
+    },
+    {
+      "epoch": 0.7071984435797666,
+      "grad_norm": 0.326530784368515,
+      "learning_rate": 3e-05,
+      "loss": 2.1224,
+      "step": 2181
+    },
+    {
+      "epoch": 0.7075226977950714,
+      "grad_norm": 0.38164928555488586,
+      "learning_rate": 3e-05,
+      "loss": 3.1738,
+      "step": 2182
+    },
+    {
+      "epoch": 0.7078469520103762,
+      "grad_norm": 0.4123162627220154,
+      "learning_rate": 3e-05,
+      "loss": 2.981,
+      "step": 2183
+    },
+    {
+      "epoch": 0.708171206225681,
+      "grad_norm": 0.40538090467453003,
+      "learning_rate": 3e-05,
+      "loss": 2.2893,
+      "step": 2184
+    },
+    {
+      "epoch": 0.7084954604409858,
+      "grad_norm": 0.5349718332290649,
+      "learning_rate": 3e-05,
+      "loss": 3.0964,
+      "step": 2185
+    },
+    {
+      "epoch": 0.7088197146562906,
+      "grad_norm": 0.31907394528388977,
+      "learning_rate": 3e-05,
+      "loss": 2.0286,
+      "step": 2186
+    },
+    {
+      "epoch": 0.7091439688715954,
+      "grad_norm": 0.4889828562736511,
+      "learning_rate": 3e-05,
+      "loss": 2.8438,
+      "step": 2187
+    },
+    {
+      "epoch": 0.7094682230869002,
+      "grad_norm": 0.43676820397377014,
+      "learning_rate": 3e-05,
+      "loss": 3.2169,
+      "step": 2188
+    },
+    {
+      "epoch": 0.709792477302205,
+      "grad_norm": 0.41915449500083923,
+      "learning_rate": 3e-05,
+      "loss": 2.6204,
+      "step": 2189
+    },
+    {
+      "epoch": 0.7101167315175098,
+      "grad_norm": 0.5238910913467407,
+      "learning_rate": 3e-05,
+      "loss": 2.8345,
+      "step": 2190
+    },
+    {
+      "epoch": 0.7104409857328146,
+      "grad_norm": 0.34570321440696716,
+      "learning_rate": 3e-05,
+      "loss": 2.4524,
+      "step": 2191
+    },
+    {
+      "epoch": 0.7107652399481194,
+      "grad_norm": 0.49384093284606934,
+      "learning_rate": 3e-05,
+      "loss": 2.6513,
+      "step": 2192
+    },
+    {
+      "epoch": 0.7110894941634242,
+      "grad_norm": 0.4730331301689148,
+      "learning_rate": 3e-05,
+      "loss": 2.8589,
+      "step": 2193
+    },
+    {
+      "epoch": 0.711413748378729,
+      "grad_norm": 0.3019007444381714,
+      "learning_rate": 3e-05,
+      "loss": 2.3156,
+      "step": 2194
+    },
+    {
+      "epoch": 0.7117380025940337,
+      "grad_norm": 0.4104849100112915,
+      "learning_rate": 3e-05,
+      "loss": 2.9581,
+      "step": 2195
+    },
+    {
+      "epoch": 0.7120622568093385,
+      "grad_norm": 0.6367778182029724,
+      "learning_rate": 3e-05,
+      "loss": 3.0819,
+      "step": 2196
+    },
+    {
+      "epoch": 0.7123865110246433,
+      "grad_norm": 0.41389626264572144,
+      "learning_rate": 3e-05,
+      "loss": 2.8292,
+      "step": 2197
+    },
+    {
+      "epoch": 0.7127107652399481,
+      "grad_norm": 0.40545669198036194,
+      "learning_rate": 3e-05,
+      "loss": 2.3455,
+      "step": 2198
+    },
+    {
+      "epoch": 0.7130350194552529,
+      "grad_norm": 0.3453202545642853,
+      "learning_rate": 3e-05,
+      "loss": 2.1682,
+      "step": 2199
+    },
+    {
+      "epoch": 0.7133592736705577,
+      "grad_norm": 0.5028184652328491,
+      "learning_rate": 3e-05,
+      "loss": 3.5263,
+      "step": 2200
+    },
+    {
+      "epoch": 0.7136835278858625,
+      "grad_norm": 0.5930689573287964,
+      "learning_rate": 3e-05,
+      "loss": 3.4326,
+      "step": 2201
+    },
+    {
+      "epoch": 0.7140077821011673,
+      "grad_norm": 0.292269766330719,
+      "learning_rate": 3e-05,
+      "loss": 2.0927,
+      "step": 2202
+    },
+    {
+      "epoch": 0.7143320363164721,
+      "grad_norm": 0.38704246282577515,
+      "learning_rate": 3e-05,
+      "loss": 2.5319,
+      "step": 2203
+    },
+    {
+      "epoch": 0.7146562905317769,
+      "grad_norm": 0.41134050488471985,
+      "learning_rate": 3e-05,
+      "loss": 2.8829,
+      "step": 2204
+    },
+    {
+      "epoch": 0.7149805447470817,
+      "grad_norm": 0.4361971318721771,
+      "learning_rate": 3e-05,
+      "loss": 2.9338,
+      "step": 2205
+    },
+    {
+      "epoch": 0.7153047989623865,
+      "grad_norm": 0.6052378416061401,
+      "learning_rate": 3e-05,
+      "loss": 2.9524,
+      "step": 2206
+    },
+    {
+      "epoch": 0.7156290531776913,
+      "grad_norm": 0.38888975977897644,
+      "learning_rate": 3e-05,
+      "loss": 2.5305,
+      "step": 2207
+    },
+    {
+      "epoch": 0.7159533073929961,
+      "grad_norm": 0.369411438703537,
+      "learning_rate": 3e-05,
+      "loss": 2.4777,
+      "step": 2208
+    },
+    {
+      "epoch": 0.7162775616083009,
+      "grad_norm": 0.4225117862224579,
+      "learning_rate": 3e-05,
+      "loss": 2.9345,
+      "step": 2209
+    },
+    {
+      "epoch": 0.7166018158236057,
+      "grad_norm": 0.4398960471153259,
+      "learning_rate": 3e-05,
+      "loss": 2.6999,
+      "step": 2210
+    },
+    {
+      "epoch": 0.7169260700389105,
+      "grad_norm": 0.7983774542808533,
+      "learning_rate": 3e-05,
+      "loss": 3.1599,
+      "step": 2211
+    },
+    {
+      "epoch": 0.7172503242542153,
+      "grad_norm": 0.5730427503585815,
+      "learning_rate": 3e-05,
+      "loss": 2.8949,
+      "step": 2212
+    },
+    {
+      "epoch": 0.7175745784695201,
+      "grad_norm": 0.5041467547416687,
+      "learning_rate": 3e-05,
+      "loss": 3.4331,
+      "step": 2213
+    },
+    {
+      "epoch": 0.7178988326848249,
+      "grad_norm": 0.3330518901348114,
+      "learning_rate": 3e-05,
+      "loss": 1.9089,
+      "step": 2214
+    },
+    {
+      "epoch": 0.7182230869001297,
+      "grad_norm": 0.5014967918395996,
+      "learning_rate": 3e-05,
+      "loss": 2.4984,
+      "step": 2215
+    },
+    {
+      "epoch": 0.7185473411154345,
+      "grad_norm": 0.4097425639629364,
+      "learning_rate": 3e-05,
+      "loss": 2.378,
+      "step": 2216
+    },
+    {
+      "epoch": 0.7188715953307393,
+      "grad_norm": 0.366003155708313,
+      "learning_rate": 3e-05,
+      "loss": 2.6421,
+      "step": 2217
+    },
+    {
+      "epoch": 0.7191958495460441,
+      "grad_norm": 0.29141566157341003,
+      "learning_rate": 3e-05,
+      "loss": 2.3524,
+      "step": 2218
+    },
+    {
+      "epoch": 0.7195201037613489,
+      "grad_norm": 0.5220842957496643,
+      "learning_rate": 3e-05,
+      "loss": 3.3061,
+      "step": 2219
+    },
+    {
+      "epoch": 0.7198443579766537,
+      "grad_norm": 0.37869974970817566,
+      "learning_rate": 3e-05,
+      "loss": 2.5203,
+      "step": 2220
+    },
+    {
+      "epoch": 0.7201686121919585,
+      "grad_norm": 0.30095043778419495,
+      "learning_rate": 3e-05,
+      "loss": 1.9822,
+      "step": 2221
+    },
+    {
+      "epoch": 0.7204928664072633,
+      "grad_norm": 0.4233868420124054,
+      "learning_rate": 3e-05,
+      "loss": 2.8602,
+      "step": 2222
+    },
+    {
+      "epoch": 0.7208171206225681,
+      "grad_norm": 0.3139123022556305,
+      "learning_rate": 3e-05,
+      "loss": 2.1258,
+      "step": 2223
+    },
+    {
+      "epoch": 0.7211413748378729,
+      "grad_norm": 0.4107092618942261,
+      "learning_rate": 3e-05,
+      "loss": 2.5078,
+      "step": 2224
+    },
+    {
+      "epoch": 0.7214656290531777,
+      "grad_norm": 0.38800305128097534,
+      "learning_rate": 3e-05,
+      "loss": 2.84,
+      "step": 2225
+    },
+    {
+      "epoch": 0.7217898832684825,
+      "grad_norm": 0.4158210754394531,
+      "learning_rate": 3e-05,
+      "loss": 2.7982,
+      "step": 2226
+    },
+    {
+      "epoch": 0.7221141374837873,
+      "grad_norm": 0.5252408981323242,
+      "learning_rate": 3e-05,
+      "loss": 2.6645,
+      "step": 2227
+    },
+    {
+      "epoch": 0.7224383916990921,
+      "grad_norm": 0.4490574598312378,
+      "learning_rate": 3e-05,
+      "loss": 3.2257,
+      "step": 2228
+    },
+    {
+      "epoch": 0.7227626459143969,
+      "grad_norm": 0.6782407760620117,
+      "learning_rate": 3e-05,
+      "loss": 3.0768,
+      "step": 2229
+    },
+    {
+      "epoch": 0.7230869001297017,
+      "grad_norm": 0.30744385719299316,
+      "learning_rate": 3e-05,
+      "loss": 2.2919,
+      "step": 2230
+    },
+    {
+      "epoch": 0.7234111543450065,
+      "grad_norm": 0.4096817672252655,
+      "learning_rate": 3e-05,
+      "loss": 2.304,
+      "step": 2231
+    },
+    {
+      "epoch": 0.7237354085603113,
+      "grad_norm": 0.2865973114967346,
+      "learning_rate": 3e-05,
+      "loss": 2.1116,
+      "step": 2232
+    },
+    {
+      "epoch": 0.7240596627756161,
+      "grad_norm": 0.4871489703655243,
+      "learning_rate": 3e-05,
+      "loss": 3.0637,
+      "step": 2233
+    },
+    {
+      "epoch": 0.7243839169909209,
+      "grad_norm": 0.4814726412296295,
+      "learning_rate": 3e-05,
+      "loss": 2.8338,
+      "step": 2234
+    },
+    {
+      "epoch": 0.7247081712062257,
+      "grad_norm": 0.3951255977153778,
+      "learning_rate": 3e-05,
+      "loss": 3.1419,
+      "step": 2235
+    },
+    {
+      "epoch": 0.7250324254215305,
+      "grad_norm": 0.36224663257598877,
+      "learning_rate": 3e-05,
+      "loss": 2.3518,
+      "step": 2236
+    },
+    {
+      "epoch": 0.7253566796368353,
+      "grad_norm": 0.3045428991317749,
+      "learning_rate": 3e-05,
+      "loss": 2.34,
+      "step": 2237
+    },
+    {
+      "epoch": 0.72568093385214,
+      "grad_norm": 0.41137826442718506,
+      "learning_rate": 3e-05,
+      "loss": 2.9127,
+      "step": 2238
+    },
+    {
+      "epoch": 0.7260051880674449,
+      "grad_norm": 0.4587876498699188,
+      "learning_rate": 3e-05,
+      "loss": 3.1336,
+      "step": 2239
+    },
+    {
+      "epoch": 0.7263294422827496,
+      "grad_norm": 0.47000613808631897,
+      "learning_rate": 3e-05,
+      "loss": 2.7276,
+      "step": 2240
+    },
+    {
+      "epoch": 0.7266536964980544,
+      "grad_norm": 0.5106474757194519,
+      "learning_rate": 3e-05,
+      "loss": 3.4562,
+      "step": 2241
+    },
+    {
+      "epoch": 0.7269779507133592,
+      "grad_norm": 0.4733826220035553,
+      "learning_rate": 3e-05,
+      "loss": 3.0017,
+      "step": 2242
+    },
+    {
+      "epoch": 0.727302204928664,
+      "grad_norm": 0.5197342038154602,
+      "learning_rate": 3e-05,
+      "loss": 3.3781,
+      "step": 2243
+    },
+    {
+      "epoch": 0.7276264591439688,
+      "grad_norm": 0.39276185631752014,
+      "learning_rate": 3e-05,
+      "loss": 2.3722,
+      "step": 2244
+    },
+    {
+      "epoch": 0.7279507133592736,
+      "grad_norm": 0.35276129841804504,
+      "learning_rate": 3e-05,
+      "loss": 2.8003,
+      "step": 2245
+    },
+    {
+      "epoch": 0.7282749675745784,
+      "grad_norm": 0.668886125087738,
+      "learning_rate": 3e-05,
+      "loss": 3.1358,
+      "step": 2246
+    },
+    {
+      "epoch": 0.7285992217898832,
+      "grad_norm": 0.3300481140613556,
+      "learning_rate": 3e-05,
+      "loss": 2.1045,
+      "step": 2247
+    },
+    {
+      "epoch": 0.728923476005188,
+      "grad_norm": 0.3842959702014923,
+      "learning_rate": 3e-05,
+      "loss": 2.5483,
+      "step": 2248
+    },
+    {
+      "epoch": 0.7292477302204928,
+      "grad_norm": 0.46831536293029785,
+      "learning_rate": 3e-05,
+      "loss": 2.652,
+      "step": 2249
+    },
+    {
+      "epoch": 0.7295719844357976,
+      "grad_norm": 0.5263088941574097,
+      "learning_rate": 3e-05,
+      "loss": 3.0283,
+      "step": 2250
+    },
+    {
+      "epoch": 0.7298962386511024,
+      "grad_norm": 0.4125734865665436,
+      "learning_rate": 3e-05,
+      "loss": 2.4589,
+      "step": 2251
+    },
+    {
+      "epoch": 0.7302204928664072,
+      "grad_norm": 0.3604307770729065,
+      "learning_rate": 3e-05,
+      "loss": 3.1339,
+      "step": 2252
+    },
+    {
+      "epoch": 0.730544747081712,
+      "grad_norm": 0.4308635890483856,
+      "learning_rate": 3e-05,
+      "loss": 2.4694,
+      "step": 2253
+    },
+    {
+      "epoch": 0.7308690012970168,
+      "grad_norm": 0.423736572265625,
+      "learning_rate": 3e-05,
+      "loss": 2.6609,
+      "step": 2254
+    },
+    {
+      "epoch": 0.7311932555123216,
+      "grad_norm": 0.3805387616157532,
+      "learning_rate": 3e-05,
+      "loss": 2.3063,
+      "step": 2255
+    },
+    {
+      "epoch": 0.7315175097276264,
+      "grad_norm": 0.335893452167511,
+      "learning_rate": 3e-05,
+      "loss": 2.5106,
+      "step": 2256
+    },
+    {
+      "epoch": 0.7318417639429312,
+      "grad_norm": 0.44538572430610657,
+      "learning_rate": 3e-05,
+      "loss": 2.5229,
+      "step": 2257
+    },
+    {
+      "epoch": 0.732166018158236,
+      "grad_norm": 0.4469459652900696,
+      "learning_rate": 3e-05,
+      "loss": 2.8471,
+      "step": 2258
+    },
+    {
+      "epoch": 0.7324902723735408,
+      "grad_norm": 0.39917227625846863,
+      "learning_rate": 3e-05,
+      "loss": 2.9706,
+      "step": 2259
+    },
+    {
+      "epoch": 0.7328145265888456,
+      "grad_norm": 0.3515602648258209,
+      "learning_rate": 3e-05,
+      "loss": 2.391,
+      "step": 2260
+    },
+    {
+      "epoch": 0.7331387808041504,
+      "grad_norm": 0.5660308599472046,
+      "learning_rate": 3e-05,
+      "loss": 3.1205,
+      "step": 2261
+    },
+    {
+      "epoch": 0.7334630350194552,
+      "grad_norm": 0.43158313632011414,
+      "learning_rate": 3e-05,
+      "loss": 3.3132,
+      "step": 2262
+    },
+    {
+      "epoch": 0.73378728923476,
+      "grad_norm": 0.36418798565864563,
+      "learning_rate": 3e-05,
+      "loss": 2.2557,
+      "step": 2263
+    },
+    {
+      "epoch": 0.7341115434500648,
+      "grad_norm": 0.5544620752334595,
+      "learning_rate": 3e-05,
+      "loss": 2.5467,
+      "step": 2264
+    },
+    {
+      "epoch": 0.7344357976653697,
+      "grad_norm": 0.3738597333431244,
+      "learning_rate": 3e-05,
+      "loss": 2.2295,
+      "step": 2265
+    },
+    {
+      "epoch": 0.7347600518806745,
+      "grad_norm": 0.4895092248916626,
+      "learning_rate": 3e-05,
+      "loss": 2.5895,
+      "step": 2266
+    },
+    {
+      "epoch": 0.7350843060959793,
+      "grad_norm": 0.33740875124931335,
+      "learning_rate": 3e-05,
+      "loss": 2.3558,
+      "step": 2267
+    },
+    {
+      "epoch": 0.7354085603112841,
+      "grad_norm": 0.4100956320762634,
+      "learning_rate": 3e-05,
+      "loss": 2.5913,
+      "step": 2268
+    },
+    {
+      "epoch": 0.7357328145265889,
+      "grad_norm": 0.35730329155921936,
+      "learning_rate": 3e-05,
+      "loss": 2.3749,
+      "step": 2269
+    },
+    {
+      "epoch": 0.7360570687418937,
+      "grad_norm": 0.4373948872089386,
+      "learning_rate": 3e-05,
+      "loss": 2.6359,
+      "step": 2270
+    },
+    {
+      "epoch": 0.7363813229571985,
+      "grad_norm": 0.5332291722297668,
+      "learning_rate": 3e-05,
+      "loss": 2.9965,
+      "step": 2271
+    },
+    {
+      "epoch": 0.7367055771725033,
+      "grad_norm": 0.9762928485870361,
+      "learning_rate": 3e-05,
+      "loss": 2.7894,
+      "step": 2272
+    },
+    {
+      "epoch": 0.7370298313878081,
+      "grad_norm": 0.38823989033699036,
+      "learning_rate": 3e-05,
+      "loss": 2.6522,
+      "step": 2273
+    },
+    {
+      "epoch": 0.7373540856031129,
+      "grad_norm": 0.33782991766929626,
+      "learning_rate": 3e-05,
+      "loss": 2.4524,
+      "step": 2274
+    },
+    {
+      "epoch": 0.7376783398184177,
+      "grad_norm": 0.35339877009391785,
+      "learning_rate": 3e-05,
+      "loss": 2.5148,
+      "step": 2275
+    },
+    {
+      "epoch": 0.7380025940337225,
+      "grad_norm": 0.4520723819732666,
+      "learning_rate": 3e-05,
+      "loss": 2.765,
+      "step": 2276
+    },
+    {
+      "epoch": 0.7383268482490273,
+      "grad_norm": 0.5133552551269531,
+      "learning_rate": 3e-05,
+      "loss": 3.1102,
+      "step": 2277
+    },
+    {
+      "epoch": 0.7386511024643321,
+      "grad_norm": 0.3917722702026367,
+      "learning_rate": 3e-05,
+      "loss": 2.7075,
+      "step": 2278
+    },
+    {
+      "epoch": 0.7389753566796369,
+      "grad_norm": 0.4251161813735962,
+      "learning_rate": 3e-05,
+      "loss": 2.5089,
+      "step": 2279
+    },
+    {
+      "epoch": 0.7392996108949417,
+      "grad_norm": 0.3842725157737732,
+      "learning_rate": 3e-05,
+      "loss": 2.4056,
+      "step": 2280
+    },
+    {
+      "epoch": 0.7396238651102465,
+      "grad_norm": 0.36579519510269165,
+      "learning_rate": 3e-05,
+      "loss": 2.3498,
+      "step": 2281
+    },
+    {
+      "epoch": 0.7399481193255513,
+      "grad_norm": 0.5414786338806152,
+      "learning_rate": 3e-05,
+      "loss": 3.1058,
+      "step": 2282
+    },
+    {
+      "epoch": 0.7402723735408561,
+      "grad_norm": 0.43491533398628235,
+      "learning_rate": 3e-05,
+      "loss": 2.5777,
+      "step": 2283
+    },
+    {
+      "epoch": 0.7405966277561609,
+      "grad_norm": 0.7089085578918457,
+      "learning_rate": 3e-05,
+      "loss": 3.2328,
+      "step": 2284
+    },
+    {
+      "epoch": 0.7409208819714657,
+      "grad_norm": 0.4430534839630127,
+      "learning_rate": 3e-05,
+      "loss": 2.8136,
+      "step": 2285
+    },
+    {
+      "epoch": 0.7412451361867705,
+      "grad_norm": 0.4015337824821472,
+      "learning_rate": 3e-05,
+      "loss": 3.0849,
+      "step": 2286
+    },
+    {
+      "epoch": 0.7415693904020753,
+      "grad_norm": 0.36706605553627014,
+      "learning_rate": 3e-05,
+      "loss": 2.7996,
+      "step": 2287
+    },
+    {
+      "epoch": 0.74189364461738,
+      "grad_norm": 0.29511839151382446,
+      "learning_rate": 3e-05,
+      "loss": 2.4341,
+      "step": 2288
+    },
+    {
+      "epoch": 0.7422178988326849,
+      "grad_norm": 0.3685775101184845,
+      "learning_rate": 3e-05,
+      "loss": 2.6448,
+      "step": 2289
+    },
+    {
+      "epoch": 0.7425421530479897,
+      "grad_norm": 0.35184699296951294,
+      "learning_rate": 3e-05,
+      "loss": 2.6694,
+      "step": 2290
+    },
+    {
+      "epoch": 0.7428664072632944,
+      "grad_norm": 0.4686950147151947,
+      "learning_rate": 3e-05,
+      "loss": 2.4713,
+      "step": 2291
+    },
+    {
+      "epoch": 0.7431906614785992,
+      "grad_norm": 0.4592655599117279,
+      "learning_rate": 3e-05,
+      "loss": 3.2976,
+      "step": 2292
+    },
+    {
+      "epoch": 0.743514915693904,
+      "grad_norm": 0.5830618143081665,
+      "learning_rate": 3e-05,
+      "loss": 3.19,
+      "step": 2293
+    },
+    {
+      "epoch": 0.7438391699092088,
+      "grad_norm": 0.3356614112854004,
+      "learning_rate": 3e-05,
+      "loss": 2.5368,
+      "step": 2294
+    },
+    {
+      "epoch": 0.7441634241245136,
+      "grad_norm": 0.3619104325771332,
+      "learning_rate": 3e-05,
+      "loss": 2.4544,
+      "step": 2295
+    },
+    {
+      "epoch": 0.7444876783398184,
+      "grad_norm": 0.34547945857048035,
+      "learning_rate": 3e-05,
+      "loss": 2.3887,
+      "step": 2296
+    },
+    {
+      "epoch": 0.7448119325551232,
+      "grad_norm": 0.2918921709060669,
+      "learning_rate": 3e-05,
+      "loss": 2.1251,
+      "step": 2297
+    },
+    {
+      "epoch": 0.745136186770428,
+      "grad_norm": 0.5134990215301514,
+      "learning_rate": 3e-05,
+      "loss": 3.1075,
+      "step": 2298
+    },
+    {
+      "epoch": 0.7454604409857328,
+      "grad_norm": 0.45499300956726074,
+      "learning_rate": 3e-05,
+      "loss": 2.9016,
+      "step": 2299
+    },
+    {
+      "epoch": 0.7457846952010376,
+      "grad_norm": 0.35816317796707153,
+      "learning_rate": 3e-05,
+      "loss": 2.5228,
+      "step": 2300
+    },
+    {
+      "epoch": 0.7461089494163424,
+      "grad_norm": 0.5518184304237366,
+      "learning_rate": 3e-05,
+      "loss": 2.889,
+      "step": 2301
+    },
+    {
+      "epoch": 0.7464332036316472,
+      "grad_norm": 0.3293609321117401,
+      "learning_rate": 3e-05,
+      "loss": 2.1232,
+      "step": 2302
+    },
+    {
+      "epoch": 0.746757457846952,
+      "grad_norm": 0.3824160397052765,
+      "learning_rate": 3e-05,
+      "loss": 2.6935,
+      "step": 2303
+    },
+    {
+      "epoch": 0.7470817120622568,
+      "grad_norm": 0.4641450345516205,
+      "learning_rate": 3e-05,
+      "loss": 3.0484,
+      "step": 2304
+    },
+    {
+      "epoch": 0.7474059662775616,
+      "grad_norm": 0.3767787516117096,
+      "learning_rate": 3e-05,
+      "loss": 2.3937,
+      "step": 2305
+    },
+    {
+      "epoch": 0.7477302204928664,
+      "grad_norm": 0.37766313552856445,
+      "learning_rate": 3e-05,
+      "loss": 2.56,
+      "step": 2306
+    },
+    {
+      "epoch": 0.7480544747081712,
+      "grad_norm": 0.414554238319397,
+      "learning_rate": 3e-05,
+      "loss": 2.7992,
+      "step": 2307
+    },
+    {
+      "epoch": 0.748378728923476,
+      "grad_norm": 0.27547192573547363,
+      "learning_rate": 3e-05,
+      "loss": 2.1439,
+      "step": 2308
+    },
+    {
+      "epoch": 0.7487029831387808,
+      "grad_norm": 0.28017109632492065,
+      "learning_rate": 3e-05,
+      "loss": 2.1976,
+      "step": 2309
+    },
+    {
+      "epoch": 0.7490272373540856,
+      "grad_norm": 0.33782684803009033,
+      "learning_rate": 3e-05,
+      "loss": 2.9187,
+      "step": 2310
+    },
+    {
+      "epoch": 0.7493514915693904,
+      "grad_norm": 0.7992986440658569,
+      "learning_rate": 3e-05,
+      "loss": 3.5877,
+      "step": 2311
+    },
+    {
+      "epoch": 0.7496757457846952,
+      "grad_norm": 0.4214220941066742,
+      "learning_rate": 3e-05,
+      "loss": 2.6212,
+      "step": 2312
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.503379225730896,
+      "learning_rate": 3e-05,
+      "loss": 2.8406,
+      "step": 2313
+    },
+    {
+      "epoch": 0.7503242542153048,
+      "grad_norm": 0.4661216139793396,
+      "learning_rate": 3e-05,
+      "loss": 3.1478,
+      "step": 2314
+    },
+    {
+      "epoch": 0.7506485084306096,
+      "grad_norm": 0.337135910987854,
+      "learning_rate": 3e-05,
+      "loss": 2.1656,
+      "step": 2315
+    },
+    {
+      "epoch": 0.7509727626459144,
+      "grad_norm": 0.3154256343841553,
+      "learning_rate": 3e-05,
+      "loss": 2.2201,
+      "step": 2316
+    },
+    {
+      "epoch": 0.7512970168612192,
+      "grad_norm": 0.6219356060028076,
+      "learning_rate": 3e-05,
+      "loss": 3.1985,
+      "step": 2317
+    },
+    {
+      "epoch": 0.751621271076524,
+      "grad_norm": 0.3656480312347412,
+      "learning_rate": 3e-05,
+      "loss": 2.4351,
+      "step": 2318
+    },
+    {
+      "epoch": 0.7519455252918288,
+      "grad_norm": 0.49076783657073975,
+      "learning_rate": 3e-05,
+      "loss": 2.7461,
+      "step": 2319
+    },
+    {
+      "epoch": 0.7522697795071336,
+      "grad_norm": 0.3122117519378662,
+      "learning_rate": 3e-05,
+      "loss": 2.113,
+      "step": 2320
+    },
+    {
+      "epoch": 0.7525940337224384,
+      "grad_norm": 0.4028626084327698,
+      "learning_rate": 3e-05,
+      "loss": 2.863,
+      "step": 2321
+    },
+    {
+      "epoch": 0.7529182879377432,
+      "grad_norm": 0.614189088344574,
+      "learning_rate": 3e-05,
+      "loss": 3.1834,
+      "step": 2322
+    },
+    {
+      "epoch": 0.753242542153048,
+      "grad_norm": 0.3237898349761963,
+      "learning_rate": 3e-05,
+      "loss": 2.4503,
+      "step": 2323
+    },
+    {
+      "epoch": 0.7535667963683528,
+      "grad_norm": 0.3026384711265564,
+      "learning_rate": 3e-05,
+      "loss": 2.2178,
+      "step": 2324
+    },
+    {
+      "epoch": 0.7538910505836576,
+      "grad_norm": 0.3652706742286682,
+      "learning_rate": 3e-05,
+      "loss": 2.4483,
+      "step": 2325
+    },
+    {
+      "epoch": 0.7542153047989624,
+      "grad_norm": 0.4900454878807068,
+      "learning_rate": 3e-05,
+      "loss": 2.7995,
+      "step": 2326
+    },
+    {
+      "epoch": 0.7545395590142672,
+      "grad_norm": 0.30699604749679565,
+      "learning_rate": 3e-05,
+      "loss": 2.0954,
+      "step": 2327
+    },
+    {
+      "epoch": 0.754863813229572,
+      "grad_norm": 0.4678918123245239,
+      "learning_rate": 3e-05,
+      "loss": 3.496,
+      "step": 2328
+    },
+    {
+      "epoch": 0.7551880674448768,
+      "grad_norm": 0.5511493682861328,
+      "learning_rate": 3e-05,
+      "loss": 3.4644,
+      "step": 2329
+    },
+    {
+      "epoch": 0.7555123216601816,
+      "grad_norm": 0.41703489422798157,
+      "learning_rate": 3e-05,
+      "loss": 3.0597,
+      "step": 2330
+    },
+    {
+      "epoch": 0.7558365758754864,
+      "grad_norm": 0.3743245601654053,
+      "learning_rate": 3e-05,
+      "loss": 2.6469,
+      "step": 2331
+    },
+    {
+      "epoch": 0.7561608300907912,
+      "grad_norm": 0.4664188623428345,
+      "learning_rate": 3e-05,
+      "loss": 3.1462,
+      "step": 2332
+    },
+    {
+      "epoch": 0.756485084306096,
+      "grad_norm": 0.3783084750175476,
+      "learning_rate": 3e-05,
+      "loss": 2.6778,
+      "step": 2333
+    },
+    {
+      "epoch": 0.7568093385214008,
+      "grad_norm": 0.5589411854743958,
+      "learning_rate": 3e-05,
+      "loss": 3.0654,
+      "step": 2334
+    },
+    {
+      "epoch": 0.7571335927367056,
+      "grad_norm": 0.450489342212677,
+      "learning_rate": 3e-05,
+      "loss": 3.2205,
+      "step": 2335
+    },
+    {
+      "epoch": 0.7574578469520103,
+      "grad_norm": 0.4207523465156555,
+      "learning_rate": 3e-05,
+      "loss": 2.9487,
+      "step": 2336
+    },
+    {
+      "epoch": 0.7577821011673151,
+      "grad_norm": 0.5124643445014954,
+      "learning_rate": 3e-05,
+      "loss": 3.5205,
+      "step": 2337
+    },
+    {
+      "epoch": 0.75810635538262,
+      "grad_norm": 0.293144553899765,
+      "learning_rate": 3e-05,
+      "loss": 2.1582,
+      "step": 2338
+    },
+    {
+      "epoch": 0.7584306095979247,
+      "grad_norm": 0.5787102580070496,
+      "learning_rate": 3e-05,
+      "loss": 2.8222,
+      "step": 2339
+    },
+    {
+      "epoch": 0.7587548638132295,
+      "grad_norm": 0.34572067856788635,
+      "learning_rate": 3e-05,
+      "loss": 2.2976,
+      "step": 2340
+    },
+    {
+      "epoch": 0.7590791180285343,
+      "grad_norm": 0.47100016474723816,
+      "learning_rate": 3e-05,
+      "loss": 3.3082,
+      "step": 2341
+    },
+    {
+      "epoch": 0.7594033722438391,
+      "grad_norm": 0.4395342767238617,
+      "learning_rate": 3e-05,
+      "loss": 3.5811,
+      "step": 2342
+    },
+    {
+      "epoch": 0.7597276264591439,
+      "grad_norm": 0.3311660587787628,
+      "learning_rate": 3e-05,
+      "loss": 2.4496,
+      "step": 2343
+    },
+    {
+      "epoch": 0.7600518806744487,
+      "grad_norm": 0.4155906140804291,
+      "learning_rate": 3e-05,
+      "loss": 3.0913,
+      "step": 2344
+    },
+    {
+      "epoch": 0.7603761348897535,
+      "grad_norm": 0.4786531925201416,
+      "learning_rate": 3e-05,
+      "loss": 3.1186,
+      "step": 2345
+    },
+    {
+      "epoch": 0.7607003891050583,
+      "grad_norm": 0.3394775092601776,
+      "learning_rate": 3e-05,
+      "loss": 2.2014,
+      "step": 2346
+    },
+    {
+      "epoch": 0.7610246433203631,
+      "grad_norm": 0.40407443046569824,
+      "learning_rate": 3e-05,
+      "loss": 2.3676,
+      "step": 2347
+    },
+    {
+      "epoch": 0.7613488975356679,
+      "grad_norm": 0.314558207988739,
+      "learning_rate": 3e-05,
+      "loss": 2.2834,
+      "step": 2348
+    },
+    {
+      "epoch": 0.7616731517509727,
+      "grad_norm": 0.3648430109024048,
+      "learning_rate": 3e-05,
+      "loss": 2.5983,
+      "step": 2349
+    },
+    {
+      "epoch": 0.7619974059662775,
+      "grad_norm": 0.36706915497779846,
+      "learning_rate": 3e-05,
+      "loss": 2.6065,
+      "step": 2350
+    },
+    {
+      "epoch": 0.7623216601815823,
+      "grad_norm": 0.6319761276245117,
+      "learning_rate": 3e-05,
+      "loss": 3.6426,
+      "step": 2351
+    },
+    {
+      "epoch": 0.7626459143968871,
+      "grad_norm": 0.36428555846214294,
+      "learning_rate": 3e-05,
+      "loss": 2.2843,
+      "step": 2352
+    },
+    {
+      "epoch": 0.7629701686121919,
+      "grad_norm": 0.3672475516796112,
+      "learning_rate": 3e-05,
+      "loss": 2.3275,
+      "step": 2353
+    },
+    {
+      "epoch": 0.7632944228274967,
+      "grad_norm": 0.4130176603794098,
+      "learning_rate": 3e-05,
+      "loss": 2.9534,
+      "step": 2354
+    },
+    {
+      "epoch": 0.7636186770428015,
+      "grad_norm": 0.30330559611320496,
+      "learning_rate": 3e-05,
+      "loss": 2.1729,
+      "step": 2355
+    },
+    {
+      "epoch": 0.7639429312581063,
+      "grad_norm": 0.3294371962547302,
+      "learning_rate": 3e-05,
+      "loss": 2.3272,
+      "step": 2356
+    },
+    {
+      "epoch": 0.7642671854734111,
+      "grad_norm": 0.66839998960495,
+      "learning_rate": 3e-05,
+      "loss": 3.8204,
+      "step": 2357
+    },
+    {
+      "epoch": 0.7645914396887159,
+      "grad_norm": 0.4124615490436554,
+      "learning_rate": 3e-05,
+      "loss": 3.0009,
+      "step": 2358
+    },
+    {
+      "epoch": 0.7649156939040207,
+      "grad_norm": 0.4964095652103424,
+      "learning_rate": 3e-05,
+      "loss": 3.0229,
+      "step": 2359
+    },
+    {
+      "epoch": 0.7652399481193255,
+      "grad_norm": 0.29815804958343506,
+      "learning_rate": 3e-05,
+      "loss": 2.1102,
+      "step": 2360
+    },
+    {
+      "epoch": 0.7655642023346303,
+      "grad_norm": 0.4074207842350006,
+      "learning_rate": 3e-05,
+      "loss": 3.199,
+      "step": 2361
+    },
+    {
+      "epoch": 0.7658884565499352,
+      "grad_norm": 0.3256722092628479,
+      "learning_rate": 3e-05,
+      "loss": 2.4568,
+      "step": 2362
+    },
+    {
+      "epoch": 0.76621271076524,
+      "grad_norm": 0.4784950315952301,
+      "learning_rate": 3e-05,
+      "loss": 2.6795,
+      "step": 2363
+    },
+    {
+      "epoch": 0.7665369649805448,
+      "grad_norm": 0.4360116422176361,
+      "learning_rate": 3e-05,
+      "loss": 2.5317,
+      "step": 2364
+    },
+    {
+      "epoch": 0.7668612191958496,
+      "grad_norm": 0.29208993911743164,
+      "learning_rate": 3e-05,
+      "loss": 2.3335,
+      "step": 2365
+    },
+    {
+      "epoch": 0.7671854734111544,
+      "grad_norm": 0.33308976888656616,
+      "learning_rate": 3e-05,
+      "loss": 2.7096,
+      "step": 2366
+    },
+    {
+      "epoch": 0.7675097276264592,
+      "grad_norm": 0.43643251061439514,
+      "learning_rate": 3e-05,
+      "loss": 2.6386,
+      "step": 2367
+    },
+    {
+      "epoch": 0.767833981841764,
+      "grad_norm": 0.33167532086372375,
+      "learning_rate": 3e-05,
+      "loss": 2.868,
+      "step": 2368
+    },
+    {
+      "epoch": 0.7681582360570688,
+      "grad_norm": 0.4683854877948761,
+      "learning_rate": 3e-05,
+      "loss": 2.998,
+      "step": 2369
+    },
+    {
+      "epoch": 0.7684824902723736,
+      "grad_norm": 0.4066368639469147,
+      "learning_rate": 3e-05,
+      "loss": 3.2835,
+      "step": 2370
+    },
+    {
+      "epoch": 0.7688067444876784,
+      "grad_norm": 0.40690481662750244,
+      "learning_rate": 3e-05,
+      "loss": 2.4395,
+      "step": 2371
+    },
+    {
+      "epoch": 0.7691309987029832,
+      "grad_norm": 0.39972618222236633,
+      "learning_rate": 3e-05,
+      "loss": 2.5705,
+      "step": 2372
+    },
+    {
+      "epoch": 0.769455252918288,
+      "grad_norm": 0.3143376410007477,
+      "learning_rate": 3e-05,
+      "loss": 2.2343,
+      "step": 2373
+    },
+    {
+      "epoch": 0.7697795071335928,
+      "grad_norm": 0.8333548307418823,
+      "learning_rate": 3e-05,
+      "loss": 3.597,
+      "step": 2374
+    },
+    {
+      "epoch": 0.7701037613488976,
+      "grad_norm": 0.39488860964775085,
+      "learning_rate": 3e-05,
+      "loss": 2.6913,
+      "step": 2375
+    },
+    {
+      "epoch": 0.7704280155642024,
+      "grad_norm": 0.42191770672798157,
+      "learning_rate": 3e-05,
+      "loss": 2.7817,
+      "step": 2376
+    },
+    {
+      "epoch": 0.7707522697795072,
+      "grad_norm": 0.4408021569252014,
+      "learning_rate": 3e-05,
+      "loss": 2.5826,
+      "step": 2377
+    },
+    {
+      "epoch": 0.771076523994812,
+      "grad_norm": 0.3749259114265442,
+      "learning_rate": 3e-05,
+      "loss": 2.6783,
+      "step": 2378
+    },
+    {
+      "epoch": 0.7714007782101168,
+      "grad_norm": 0.4815003573894501,
+      "learning_rate": 3e-05,
+      "loss": 3.0047,
+      "step": 2379
+    },
+    {
+      "epoch": 0.7717250324254216,
+      "grad_norm": 0.6340318322181702,
+      "learning_rate": 3e-05,
+      "loss": 3.4992,
+      "step": 2380
+    },
+    {
+      "epoch": 0.7720492866407264,
+      "grad_norm": 0.37394338846206665,
+      "learning_rate": 3e-05,
+      "loss": 2.8392,
+      "step": 2381
+    },
+    {
+      "epoch": 0.7723735408560312,
+      "grad_norm": 0.36805301904678345,
+      "learning_rate": 3e-05,
+      "loss": 2.3849,
+      "step": 2382
+    },
+    {
+      "epoch": 0.772697795071336,
+      "grad_norm": 0.3902498185634613,
+      "learning_rate": 3e-05,
+      "loss": 2.6987,
+      "step": 2383
+    },
+    {
+      "epoch": 0.7730220492866408,
+      "grad_norm": 0.5051146745681763,
+      "learning_rate": 3e-05,
+      "loss": 3.7633,
+      "step": 2384
+    },
+    {
+      "epoch": 0.7733463035019456,
+      "grad_norm": 0.3777894675731659,
+      "learning_rate": 3e-05,
+      "loss": 3.1966,
+      "step": 2385
+    },
+    {
+      "epoch": 0.7736705577172504,
+      "grad_norm": 0.4174041748046875,
+      "learning_rate": 3e-05,
+      "loss": 2.3575,
+      "step": 2386
+    },
+    {
+      "epoch": 0.7739948119325551,
+      "grad_norm": 0.4527999460697174,
+      "learning_rate": 3e-05,
+      "loss": 3.5398,
+      "step": 2387
+    },
+    {
+      "epoch": 0.77431906614786,
+      "grad_norm": 0.3058457374572754,
+      "learning_rate": 3e-05,
+      "loss": 2.3421,
+      "step": 2388
+    },
+    {
+      "epoch": 0.7746433203631647,
+      "grad_norm": 0.3520162105560303,
+      "learning_rate": 3e-05,
+      "loss": 2.3487,
+      "step": 2389
+    },
+    {
+      "epoch": 0.7749675745784695,
+      "grad_norm": 0.3191359341144562,
+      "learning_rate": 3e-05,
+      "loss": 2.2568,
+      "step": 2390
+    },
+    {
+      "epoch": 0.7752918287937743,
+      "grad_norm": 0.40229320526123047,
+      "learning_rate": 3e-05,
+      "loss": 2.9236,
+      "step": 2391
+    },
+    {
+      "epoch": 0.7756160830090791,
+      "grad_norm": 0.35262057185173035,
+      "learning_rate": 3e-05,
+      "loss": 2.4992,
+      "step": 2392
+    },
+    {
+      "epoch": 0.7759403372243839,
+      "grad_norm": 0.3197650909423828,
+      "learning_rate": 3e-05,
+      "loss": 2.2994,
+      "step": 2393
+    },
+    {
+      "epoch": 0.7762645914396887,
+      "grad_norm": 0.6854479312896729,
+      "learning_rate": 3e-05,
+      "loss": 3.6827,
+      "step": 2394
+    },
+    {
+      "epoch": 0.7765888456549935,
+      "grad_norm": 0.3514796495437622,
+      "learning_rate": 3e-05,
+      "loss": 2.3291,
+      "step": 2395
+    },
+    {
+      "epoch": 0.7769130998702983,
+      "grad_norm": 0.3858470916748047,
+      "learning_rate": 3e-05,
+      "loss": 2.9778,
+      "step": 2396
+    },
+    {
+      "epoch": 0.7772373540856031,
+      "grad_norm": 0.5888802409172058,
+      "learning_rate": 3e-05,
+      "loss": 3.0112,
+      "step": 2397
+    },
+    {
+      "epoch": 0.7775616083009079,
+      "grad_norm": 0.4733554422855377,
+      "learning_rate": 3e-05,
+      "loss": 3.0526,
+      "step": 2398
+    },
+    {
+      "epoch": 0.7778858625162127,
+      "grad_norm": 0.3235276937484741,
+      "learning_rate": 3e-05,
+      "loss": 2.2714,
+      "step": 2399
+    },
+    {
+      "epoch": 0.7782101167315175,
+      "grad_norm": 0.4441587030887604,
+      "learning_rate": 3e-05,
+      "loss": 2.7159,
+      "step": 2400
+    },
+    {
+      "epoch": 0.7785343709468223,
+      "grad_norm": 0.4218437075614929,
+      "learning_rate": 3e-05,
+      "loss": 2.9449,
+      "step": 2401
+    },
+    {
+      "epoch": 0.7788586251621271,
+      "grad_norm": 0.3371029198169708,
+      "learning_rate": 3e-05,
+      "loss": 2.65,
+      "step": 2402
+    },
+    {
+      "epoch": 0.7791828793774319,
+      "grad_norm": 0.3074795603752136,
+      "learning_rate": 3e-05,
+      "loss": 2.5198,
+      "step": 2403
+    },
+    {
+      "epoch": 0.7795071335927367,
+      "grad_norm": 0.42564401030540466,
+      "learning_rate": 3e-05,
+      "loss": 2.567,
+      "step": 2404
+    },
+    {
+      "epoch": 0.7798313878080415,
+      "grad_norm": 0.41960594058036804,
+      "learning_rate": 3e-05,
+      "loss": 2.5225,
+      "step": 2405
+    },
+    {
+      "epoch": 0.7801556420233463,
+      "grad_norm": 0.3062595725059509,
+      "learning_rate": 3e-05,
+      "loss": 2.1756,
+      "step": 2406
+    },
+    {
+      "epoch": 0.7804798962386511,
+      "grad_norm": 0.5379174947738647,
+      "learning_rate": 3e-05,
+      "loss": 3.4645,
+      "step": 2407
+    },
+    {
+      "epoch": 0.7808041504539559,
+      "grad_norm": 0.3551263213157654,
+      "learning_rate": 3e-05,
+      "loss": 2.5975,
+      "step": 2408
+    },
+    {
+      "epoch": 0.7811284046692607,
+      "grad_norm": 0.43215611577033997,
+      "learning_rate": 3e-05,
+      "loss": 2.8391,
+      "step": 2409
+    },
+    {
+      "epoch": 0.7814526588845655,
+      "grad_norm": 0.34690386056900024,
+      "learning_rate": 3e-05,
+      "loss": 2.4247,
+      "step": 2410
+    },
+    {
+      "epoch": 0.7817769130998703,
+      "grad_norm": 0.31602126359939575,
+      "learning_rate": 3e-05,
+      "loss": 2.332,
+      "step": 2411
+    },
+    {
+      "epoch": 0.7821011673151751,
+      "grad_norm": 0.515683114528656,
+      "learning_rate": 3e-05,
+      "loss": 2.9702,
+      "step": 2412
+    },
+    {
+      "epoch": 0.7824254215304799,
+      "grad_norm": 0.3538980185985565,
+      "learning_rate": 3e-05,
+      "loss": 2.5767,
+      "step": 2413
+    },
+    {
+      "epoch": 0.7827496757457847,
+      "grad_norm": 0.41989266872406006,
+      "learning_rate": 3e-05,
+      "loss": 2.8232,
+      "step": 2414
+    },
+    {
+      "epoch": 0.7830739299610895,
+      "grad_norm": 0.3983188271522522,
+      "learning_rate": 3e-05,
+      "loss": 2.5298,
+      "step": 2415
+    },
+    {
+      "epoch": 0.7833981841763943,
+      "grad_norm": 0.3736632466316223,
+      "learning_rate": 3e-05,
+      "loss": 2.3449,
+      "step": 2416
+    },
+    {
+      "epoch": 0.7837224383916991,
+      "grad_norm": 0.580852210521698,
+      "learning_rate": 3e-05,
+      "loss": 2.9896,
+      "step": 2417
+    },
+    {
+      "epoch": 0.7840466926070039,
+      "grad_norm": 0.6689146161079407,
+      "learning_rate": 3e-05,
+      "loss": 3.4837,
+      "step": 2418
+    },
+    {
+      "epoch": 0.7843709468223087,
+      "grad_norm": 0.5002192258834839,
+      "learning_rate": 3e-05,
+      "loss": 3.3964,
+      "step": 2419
+    },
+    {
+      "epoch": 0.7846952010376135,
+      "grad_norm": 0.30727458000183105,
+      "learning_rate": 3e-05,
+      "loss": 1.8812,
+      "step": 2420
+    },
+    {
+      "epoch": 0.7850194552529183,
+      "grad_norm": 0.5111697316169739,
+      "learning_rate": 3e-05,
+      "loss": 2.764,
+      "step": 2421
+    },
+    {
+      "epoch": 0.7853437094682231,
+      "grad_norm": 0.45970436930656433,
+      "learning_rate": 3e-05,
+      "loss": 3.0607,
+      "step": 2422
+    },
+    {
+      "epoch": 0.7856679636835279,
+      "grad_norm": 0.5948849320411682,
+      "learning_rate": 3e-05,
+      "loss": 2.8293,
+      "step": 2423
+    },
+    {
+      "epoch": 0.7859922178988327,
+      "grad_norm": 0.5860338807106018,
+      "learning_rate": 3e-05,
+      "loss": 3.1167,
+      "step": 2424
+    },
+    {
+      "epoch": 0.7863164721141375,
+      "grad_norm": 0.4114709794521332,
+      "learning_rate": 3e-05,
+      "loss": 2.6992,
+      "step": 2425
+    },
+    {
+      "epoch": 0.7866407263294423,
+      "grad_norm": 0.3666079342365265,
+      "learning_rate": 3e-05,
+      "loss": 2.4696,
+      "step": 2426
+    },
+    {
+      "epoch": 0.7869649805447471,
+      "grad_norm": 0.3620675504207611,
+      "learning_rate": 3e-05,
+      "loss": 2.2756,
+      "step": 2427
+    },
+    {
+      "epoch": 0.7872892347600519,
+      "grad_norm": 0.5178166031837463,
+      "learning_rate": 3e-05,
+      "loss": 3.3133,
+      "step": 2428
+    },
+    {
+      "epoch": 0.7876134889753567,
+      "grad_norm": 0.34999412298202515,
+      "learning_rate": 3e-05,
+      "loss": 2.2794,
+      "step": 2429
+    },
+    {
+      "epoch": 0.7879377431906615,
+      "grad_norm": 0.9158700704574585,
+      "learning_rate": 3e-05,
+      "loss": 3.7743,
+      "step": 2430
+    },
+    {
+      "epoch": 0.7882619974059663,
+      "grad_norm": 0.3351236879825592,
+      "learning_rate": 3e-05,
+      "loss": 2.4478,
+      "step": 2431
+    },
+    {
+      "epoch": 0.788586251621271,
+      "grad_norm": 0.4052659273147583,
+      "learning_rate": 3e-05,
+      "loss": 2.7933,
+      "step": 2432
+    },
+    {
+      "epoch": 0.7889105058365758,
+      "grad_norm": 0.5009724497795105,
+      "learning_rate": 3e-05,
+      "loss": 2.4725,
+      "step": 2433
+    },
+    {
+      "epoch": 0.7892347600518806,
+      "grad_norm": 0.5147562026977539,
+      "learning_rate": 3e-05,
+      "loss": 2.3575,
+      "step": 2434
+    },
+    {
+      "epoch": 0.7895590142671854,
+      "grad_norm": 0.6050236821174622,
+      "learning_rate": 3e-05,
+      "loss": 2.6647,
+      "step": 2435
+    },
+    {
+      "epoch": 0.7898832684824902,
+      "grad_norm": 0.33910393714904785,
+      "learning_rate": 3e-05,
+      "loss": 2.7343,
+      "step": 2436
+    },
+    {
+      "epoch": 0.790207522697795,
+      "grad_norm": 0.3982275724411011,
+      "learning_rate": 3e-05,
+      "loss": 2.8685,
+      "step": 2437
+    },
+    {
+      "epoch": 0.7905317769130998,
+      "grad_norm": 0.42982369661331177,
+      "learning_rate": 3e-05,
+      "loss": 2.7414,
+      "step": 2438
+    },
+    {
+      "epoch": 0.7908560311284046,
+      "grad_norm": 0.36094385385513306,
+      "learning_rate": 3e-05,
+      "loss": 2.3926,
+      "step": 2439
+    },
+    {
+      "epoch": 0.7911802853437094,
+      "grad_norm": 0.41223883628845215,
+      "learning_rate": 3e-05,
+      "loss": 2.4981,
+      "step": 2440
+    },
+    {
+      "epoch": 0.7915045395590142,
+      "grad_norm": 0.6591712832450867,
+      "learning_rate": 3e-05,
+      "loss": 3.4687,
+      "step": 2441
+    },
+    {
+      "epoch": 0.791828793774319,
+      "grad_norm": 0.607776939868927,
+      "learning_rate": 3e-05,
+      "loss": 3.1219,
+      "step": 2442
+    },
+    {
+      "epoch": 0.7921530479896238,
+      "grad_norm": 0.3989018201828003,
+      "learning_rate": 3e-05,
+      "loss": 2.6377,
+      "step": 2443
+    },
+    {
+      "epoch": 0.7924773022049286,
+      "grad_norm": 0.4042106568813324,
+      "learning_rate": 3e-05,
+      "loss": 2.6411,
+      "step": 2444
+    },
+    {
+      "epoch": 0.7928015564202334,
+      "grad_norm": 0.3195810616016388,
+      "learning_rate": 3e-05,
+      "loss": 2.3501,
+      "step": 2445
+    },
+    {
+      "epoch": 0.7931258106355382,
+      "grad_norm": 0.43243148922920227,
+      "learning_rate": 3e-05,
+      "loss": 2.4709,
+      "step": 2446
+    },
+    {
+      "epoch": 0.793450064850843,
+      "grad_norm": 0.39463886618614197,
+      "learning_rate": 3e-05,
+      "loss": 2.1617,
+      "step": 2447
+    },
+    {
+      "epoch": 0.7937743190661478,
+      "grad_norm": 0.4837309718132019,
+      "learning_rate": 3e-05,
+      "loss": 3.2268,
+      "step": 2448
+    },
+    {
+      "epoch": 0.7940985732814526,
+      "grad_norm": 0.5855808258056641,
+      "learning_rate": 3e-05,
+      "loss": 2.4861,
+      "step": 2449
+    },
+    {
+      "epoch": 0.7944228274967574,
+      "grad_norm": 0.44983506202697754,
+      "learning_rate": 3e-05,
+      "loss": 3.338,
+      "step": 2450
+    },
+    {
+      "epoch": 0.7947470817120622,
+      "grad_norm": 0.3522905707359314,
+      "learning_rate": 3e-05,
+      "loss": 2.3552,
+      "step": 2451
+    },
+    {
+      "epoch": 0.795071335927367,
+      "grad_norm": 0.31347155570983887,
+      "learning_rate": 3e-05,
+      "loss": 2.1008,
+      "step": 2452
+    },
+    {
+      "epoch": 0.7953955901426718,
+      "grad_norm": 0.32503148913383484,
+      "learning_rate": 3e-05,
+      "loss": 2.1292,
+      "step": 2453
+    },
+    {
+      "epoch": 0.7957198443579766,
+      "grad_norm": 0.37782344222068787,
+      "learning_rate": 3e-05,
+      "loss": 2.7986,
+      "step": 2454
+    },
+    {
+      "epoch": 0.7960440985732814,
+      "grad_norm": 0.4352971315383911,
+      "learning_rate": 3e-05,
+      "loss": 2.6838,
+      "step": 2455
+    },
+    {
+      "epoch": 0.7963683527885862,
+      "grad_norm": 1.1206614971160889,
+      "learning_rate": 3e-05,
+      "loss": 2.9865,
+      "step": 2456
+    },
+    {
+      "epoch": 0.796692607003891,
+      "grad_norm": 0.3629867136478424,
+      "learning_rate": 3e-05,
+      "loss": 2.852,
+      "step": 2457
+    },
+    {
+      "epoch": 0.7970168612191959,
+      "grad_norm": 0.4733630120754242,
+      "learning_rate": 3e-05,
+      "loss": 3.2809,
+      "step": 2458
+    },
+    {
+      "epoch": 0.7973411154345007,
+      "grad_norm": 0.45862874388694763,
+      "learning_rate": 3e-05,
+      "loss": 2.2956,
+      "step": 2459
+    },
+    {
+      "epoch": 0.7976653696498055,
+      "grad_norm": 0.41920119524002075,
+      "learning_rate": 3e-05,
+      "loss": 2.9199,
+      "step": 2460
+    },
+    {
+      "epoch": 0.7979896238651103,
+      "grad_norm": 0.39342162013053894,
+      "learning_rate": 3e-05,
+      "loss": 2.5266,
+      "step": 2461
+    },
+    {
+      "epoch": 0.7983138780804151,
+      "grad_norm": 0.36266854405403137,
+      "learning_rate": 3e-05,
+      "loss": 2.9153,
+      "step": 2462
+    },
+    {
+      "epoch": 0.7986381322957199,
+      "grad_norm": 0.41828829050064087,
+      "learning_rate": 3e-05,
+      "loss": 2.5687,
+      "step": 2463
+    },
+    {
+      "epoch": 0.7989623865110247,
+      "grad_norm": 0.30060237646102905,
+      "learning_rate": 3e-05,
+      "loss": 2.2698,
+      "step": 2464
+    },
+    {
+      "epoch": 0.7992866407263295,
+      "grad_norm": 0.4955594837665558,
+      "learning_rate": 3e-05,
+      "loss": 3.2504,
+      "step": 2465
+    },
+    {
+      "epoch": 0.7996108949416343,
+      "grad_norm": 0.45362624526023865,
+      "learning_rate": 3e-05,
+      "loss": 2.8649,
+      "step": 2466
+    },
+    {
+      "epoch": 0.7999351491569391,
+      "grad_norm": 0.8328549861907959,
+      "learning_rate": 3e-05,
+      "loss": 3.0347,
+      "step": 2467
+    },
+    {
+      "epoch": 0.8002594033722439,
+      "grad_norm": 0.5448450446128845,
+      "learning_rate": 3e-05,
+      "loss": 3.1436,
+      "step": 2468
+    },
+    {
+      "epoch": 0.8005836575875487,
+      "grad_norm": 0.4479685127735138,
+      "learning_rate": 3e-05,
+      "loss": 2.6674,
+      "step": 2469
+    },
+    {
+      "epoch": 0.8009079118028535,
+      "grad_norm": 0.5034958124160767,
+      "learning_rate": 3e-05,
+      "loss": 2.9556,
+      "step": 2470
+    },
+    {
+      "epoch": 0.8012321660181583,
+      "grad_norm": 0.34309765696525574,
+      "learning_rate": 3e-05,
+      "loss": 2.3069,
+      "step": 2471
+    },
+    {
+      "epoch": 0.8015564202334631,
+      "grad_norm": 0.3392779529094696,
+      "learning_rate": 3e-05,
+      "loss": 2.3435,
+      "step": 2472
+    },
+    {
+      "epoch": 0.8018806744487679,
+      "grad_norm": 0.3982827365398407,
+      "learning_rate": 3e-05,
+      "loss": 2.9475,
+      "step": 2473
+    },
+    {
+      "epoch": 0.8022049286640727,
+      "grad_norm": 0.2964060306549072,
+      "learning_rate": 3e-05,
+      "loss": 2.037,
+      "step": 2474
+    },
+    {
+      "epoch": 0.8025291828793775,
+      "grad_norm": 0.5404691100120544,
+      "learning_rate": 3e-05,
+      "loss": 2.632,
+      "step": 2475
+    },
+    {
+      "epoch": 0.8028534370946823,
+      "grad_norm": 0.9548511505126953,
+      "learning_rate": 3e-05,
+      "loss": 3.2238,
+      "step": 2476
+    },
+    {
+      "epoch": 0.8031776913099871,
+      "grad_norm": 0.3428390920162201,
+      "learning_rate": 3e-05,
+      "loss": 2.268,
+      "step": 2477
+    },
+    {
+      "epoch": 0.8035019455252919,
+      "grad_norm": 0.5679109692573547,
+      "learning_rate": 3e-05,
+      "loss": 3.2439,
+      "step": 2478
+    },
+    {
+      "epoch": 0.8038261997405967,
+      "grad_norm": 0.42290613055229187,
+      "learning_rate": 3e-05,
+      "loss": 2.4815,
+      "step": 2479
+    },
+    {
+      "epoch": 0.8041504539559015,
+      "grad_norm": 0.43277233839035034,
+      "learning_rate": 3e-05,
+      "loss": 3.064,
+      "step": 2480
+    },
+    {
+      "epoch": 0.8044747081712063,
+      "grad_norm": 0.34628555178642273,
+      "learning_rate": 3e-05,
+      "loss": 2.0626,
+      "step": 2481
+    },
+    {
+      "epoch": 0.804798962386511,
+      "grad_norm": 0.4301173985004425,
+      "learning_rate": 3e-05,
+      "loss": 2.9728,
+      "step": 2482
+    },
+    {
+      "epoch": 0.8051232166018158,
+      "grad_norm": 0.5163307785987854,
+      "learning_rate": 3e-05,
+      "loss": 3.1604,
+      "step": 2483
+    },
+    {
+      "epoch": 0.8054474708171206,
+      "grad_norm": 0.5498360991477966,
+      "learning_rate": 3e-05,
+      "loss": 3.237,
+      "step": 2484
+    },
+    {
+      "epoch": 0.8057717250324254,
+      "grad_norm": 0.624808132648468,
+      "learning_rate": 3e-05,
+      "loss": 2.9855,
+      "step": 2485
+    },
+    {
+      "epoch": 0.8060959792477302,
+      "grad_norm": 0.41645127534866333,
+      "learning_rate": 3e-05,
+      "loss": 3.1907,
+      "step": 2486
+    },
+    {
+      "epoch": 0.806420233463035,
+      "grad_norm": 0.4786670207977295,
+      "learning_rate": 3e-05,
+      "loss": 2.8285,
+      "step": 2487
+    },
+    {
+      "epoch": 0.8067444876783398,
+      "grad_norm": 0.40947258472442627,
+      "learning_rate": 3e-05,
+      "loss": 2.5437,
+      "step": 2488
+    },
+    {
+      "epoch": 0.8070687418936446,
+      "grad_norm": 0.6873460412025452,
+      "learning_rate": 3e-05,
+      "loss": 3.6767,
+      "step": 2489
+    },
+    {
+      "epoch": 0.8073929961089494,
+      "grad_norm": 0.46158403158187866,
+      "learning_rate": 3e-05,
+      "loss": 3.205,
+      "step": 2490
+    },
+    {
+      "epoch": 0.8077172503242542,
+      "grad_norm": 0.4438329339027405,
+      "learning_rate": 3e-05,
+      "loss": 2.8682,
+      "step": 2491
+    },
+    {
+      "epoch": 0.808041504539559,
+      "grad_norm": 0.4599117934703827,
+      "learning_rate": 3e-05,
+      "loss": 2.6727,
+      "step": 2492
+    },
+    {
+      "epoch": 0.8083657587548638,
+      "grad_norm": 0.3140725791454315,
+      "learning_rate": 3e-05,
+      "loss": 2.0418,
+      "step": 2493
+    },
+    {
+      "epoch": 0.8086900129701686,
+      "grad_norm": 0.42597687244415283,
+      "learning_rate": 3e-05,
+      "loss": 2.9091,
+      "step": 2494
+    },
+    {
+      "epoch": 0.8090142671854734,
+      "grad_norm": 0.32672467827796936,
+      "learning_rate": 3e-05,
+      "loss": 2.4359,
+      "step": 2495
+    },
+    {
+      "epoch": 0.8093385214007782,
+      "grad_norm": 0.35298240184783936,
+      "learning_rate": 3e-05,
+      "loss": 2.3535,
+      "step": 2496
+    },
+    {
+      "epoch": 0.809662775616083,
+      "grad_norm": 0.5052971243858337,
+      "learning_rate": 3e-05,
+      "loss": 2.984,
+      "step": 2497
+    },
+    {
+      "epoch": 0.8099870298313878,
+      "grad_norm": 0.6553210616111755,
+      "learning_rate": 3e-05,
+      "loss": 3.8033,
+      "step": 2498
+    },
+    {
+      "epoch": 0.8103112840466926,
+      "grad_norm": 0.379864364862442,
+      "learning_rate": 3e-05,
+      "loss": 2.8963,
+      "step": 2499
+    },
+    {
+      "epoch": 0.8106355382619974,
+      "grad_norm": 0.4387440085411072,
+      "learning_rate": 3e-05,
+      "loss": 3.2797,
+      "step": 2500
+    },
+    {
+      "epoch": 0.8109597924773022,
+      "grad_norm": 0.36561039090156555,
+      "learning_rate": 3e-05,
+      "loss": 2.2536,
+      "step": 2501
+    },
+    {
+      "epoch": 0.811284046692607,
+      "grad_norm": 0.36797475814819336,
+      "learning_rate": 3e-05,
+      "loss": 2.2975,
+      "step": 2502
+    },
+    {
+      "epoch": 0.8116083009079118,
+      "grad_norm": 0.3356030285358429,
+      "learning_rate": 3e-05,
+      "loss": 2.5704,
+      "step": 2503
+    },
+    {
+      "epoch": 0.8119325551232166,
+      "grad_norm": 0.4587961733341217,
+      "learning_rate": 3e-05,
+      "loss": 2.58,
+      "step": 2504
+    },
+    {
+      "epoch": 0.8122568093385214,
+      "grad_norm": 0.3308197259902954,
+      "learning_rate": 3e-05,
+      "loss": 2.4822,
+      "step": 2505
+    },
+    {
+      "epoch": 0.8125810635538262,
+      "grad_norm": 0.44469812512397766,
+      "learning_rate": 3e-05,
+      "loss": 2.6222,
+      "step": 2506
+    },
+    {
+      "epoch": 0.812905317769131,
+      "grad_norm": 1.239303708076477,
+      "learning_rate": 3e-05,
+      "loss": 4.6181,
+      "step": 2507
+    },
+    {
+      "epoch": 0.8132295719844358,
+      "grad_norm": 0.5654007792472839,
+      "learning_rate": 3e-05,
+      "loss": 2.9483,
+      "step": 2508
+    },
+    {
+      "epoch": 0.8135538261997406,
+      "grad_norm": 0.5062891244888306,
+      "learning_rate": 3e-05,
+      "loss": 2.5291,
+      "step": 2509
+    },
+    {
+      "epoch": 0.8138780804150454,
+      "grad_norm": 0.37333282828330994,
+      "learning_rate": 3e-05,
+      "loss": 2.1258,
+      "step": 2510
+    },
+    {
+      "epoch": 0.8142023346303502,
+      "grad_norm": 0.4975797235965729,
+      "learning_rate": 3e-05,
+      "loss": 2.4052,
+      "step": 2511
+    },
+    {
+      "epoch": 0.814526588845655,
+      "grad_norm": 0.41550737619400024,
+      "learning_rate": 3e-05,
+      "loss": 2.9008,
+      "step": 2512
+    },
+    {
+      "epoch": 0.8148508430609598,
+      "grad_norm": 0.42005160450935364,
+      "learning_rate": 3e-05,
+      "loss": 2.5955,
+      "step": 2513
+    },
+    {
+      "epoch": 0.8151750972762646,
+      "grad_norm": 0.325691282749176,
+      "learning_rate": 3e-05,
+      "loss": 2.3351,
+      "step": 2514
+    },
+    {
+      "epoch": 0.8154993514915694,
+      "grad_norm": 0.33476880192756653,
+      "learning_rate": 3e-05,
+      "loss": 2.215,
+      "step": 2515
+    },
+    {
+      "epoch": 0.8158236057068742,
+      "grad_norm": 0.4228813648223877,
+      "learning_rate": 3e-05,
+      "loss": 2.4178,
+      "step": 2516
+    },
+    {
+      "epoch": 0.816147859922179,
+      "grad_norm": 0.3803859353065491,
+      "learning_rate": 3e-05,
+      "loss": 2.5565,
+      "step": 2517
+    },
+    {
+      "epoch": 0.8164721141374838,
+      "grad_norm": 0.49972599744796753,
+      "learning_rate": 3e-05,
+      "loss": 3.1376,
+      "step": 2518
+    },
+    {
+      "epoch": 0.8167963683527886,
+      "grad_norm": 0.5272126197814941,
+      "learning_rate": 3e-05,
+      "loss": 3.4416,
+      "step": 2519
+    },
+    {
+      "epoch": 0.8171206225680934,
+      "grad_norm": 0.34348607063293457,
+      "learning_rate": 3e-05,
+      "loss": 2.0278,
+      "step": 2520
+    },
+    {
+      "epoch": 0.8174448767833982,
+      "grad_norm": 0.4037092924118042,
+      "learning_rate": 3e-05,
+      "loss": 2.9113,
+      "step": 2521
+    },
+    {
+      "epoch": 0.817769130998703,
+      "grad_norm": 0.5183500647544861,
+      "learning_rate": 3e-05,
+      "loss": 2.4756,
+      "step": 2522
+    },
+    {
+      "epoch": 0.8180933852140078,
+      "grad_norm": 0.38525426387786865,
+      "learning_rate": 3e-05,
+      "loss": 2.3688,
+      "step": 2523
+    },
+    {
+      "epoch": 0.8184176394293126,
+      "grad_norm": 0.5122836828231812,
+      "learning_rate": 3e-05,
+      "loss": 2.9672,
+      "step": 2524
+    },
+    {
+      "epoch": 0.8187418936446174,
+      "grad_norm": 0.49154433608055115,
+      "learning_rate": 3e-05,
+      "loss": 2.5773,
+      "step": 2525
+    },
+    {
+      "epoch": 0.8190661478599222,
+      "grad_norm": 0.47045931220054626,
+      "learning_rate": 3e-05,
+      "loss": 2.5054,
+      "step": 2526
+    },
+    {
+      "epoch": 0.819390402075227,
+      "grad_norm": 0.55659019947052,
+      "learning_rate": 3e-05,
+      "loss": 2.9826,
+      "step": 2527
+    },
+    {
+      "epoch": 0.8197146562905318,
+      "grad_norm": 0.37834882736206055,
+      "learning_rate": 3e-05,
+      "loss": 2.5049,
+      "step": 2528
+    },
+    {
+      "epoch": 0.8200389105058365,
+      "grad_norm": 0.3999679982662201,
+      "learning_rate": 3e-05,
+      "loss": 2.7577,
+      "step": 2529
+    },
+    {
+      "epoch": 0.8203631647211413,
+      "grad_norm": 0.3568398356437683,
+      "learning_rate": 3e-05,
+      "loss": 2.2883,
+      "step": 2530
+    },
+    {
+      "epoch": 0.8206874189364461,
+      "grad_norm": 0.5048381090164185,
+      "learning_rate": 3e-05,
+      "loss": 3.2605,
+      "step": 2531
+    },
+    {
+      "epoch": 0.8210116731517509,
+      "grad_norm": 0.3969660997390747,
+      "learning_rate": 3e-05,
+      "loss": 2.9495,
+      "step": 2532
+    },
+    {
+      "epoch": 0.8213359273670557,
+      "grad_norm": 0.3681194484233856,
+      "learning_rate": 3e-05,
+      "loss": 2.7069,
+      "step": 2533
+    },
+    {
+      "epoch": 0.8216601815823605,
+      "grad_norm": 0.3600887954235077,
+      "learning_rate": 3e-05,
+      "loss": 2.7059,
+      "step": 2534
+    },
+    {
+      "epoch": 0.8219844357976653,
+      "grad_norm": 0.6702931523323059,
+      "learning_rate": 3e-05,
+      "loss": 3.0399,
+      "step": 2535
+    },
+    {
+      "epoch": 0.8223086900129701,
+      "grad_norm": 0.3805461823940277,
+      "learning_rate": 3e-05,
+      "loss": 2.4817,
+      "step": 2536
+    },
+    {
+      "epoch": 0.8226329442282749,
+      "grad_norm": 0.39615169167518616,
+      "learning_rate": 3e-05,
+      "loss": 2.6965,
+      "step": 2537
+    },
+    {
+      "epoch": 0.8229571984435797,
+      "grad_norm": 0.42350131273269653,
+      "learning_rate": 3e-05,
+      "loss": 3.2785,
+      "step": 2538
+    },
+    {
+      "epoch": 0.8232814526588845,
+      "grad_norm": 0.30823394656181335,
+      "learning_rate": 3e-05,
+      "loss": 2.3712,
+      "step": 2539
+    },
+    {
+      "epoch": 0.8236057068741893,
+      "grad_norm": 0.4292287826538086,
+      "learning_rate": 3e-05,
+      "loss": 2.8165,
+      "step": 2540
+    },
+    {
+      "epoch": 0.8239299610894941,
+      "grad_norm": 0.4344503581523895,
+      "learning_rate": 3e-05,
+      "loss": 2.6711,
+      "step": 2541
+    },
+    {
+      "epoch": 0.8242542153047989,
+      "grad_norm": 0.4776137173175812,
+      "learning_rate": 3e-05,
+      "loss": 2.6339,
+      "step": 2542
+    },
+    {
+      "epoch": 0.8245784695201037,
+      "grad_norm": 0.3832608461380005,
+      "learning_rate": 3e-05,
+      "loss": 2.6367,
+      "step": 2543
+    },
+    {
+      "epoch": 0.8249027237354085,
+      "grad_norm": 0.3245062828063965,
+      "learning_rate": 3e-05,
+      "loss": 2.3325,
+      "step": 2544
+    },
+    {
+      "epoch": 0.8252269779507133,
+      "grad_norm": 0.4462580680847168,
+      "learning_rate": 3e-05,
+      "loss": 3.2501,
+      "step": 2545
+    },
+    {
+      "epoch": 0.8255512321660181,
+      "grad_norm": 0.34406760334968567,
+      "learning_rate": 3e-05,
+      "loss": 2.2555,
+      "step": 2546
+    },
+    {
+      "epoch": 0.8258754863813229,
+      "grad_norm": 0.38532233238220215,
+      "learning_rate": 3e-05,
+      "loss": 2.7812,
+      "step": 2547
+    },
+    {
+      "epoch": 0.8261997405966277,
+      "grad_norm": 0.3767668604850769,
+      "learning_rate": 3e-05,
+      "loss": 2.5655,
+      "step": 2548
+    },
+    {
+      "epoch": 0.8265239948119325,
+      "grad_norm": 0.345650315284729,
+      "learning_rate": 3e-05,
+      "loss": 2.3549,
+      "step": 2549
+    },
+    {
+      "epoch": 0.8268482490272373,
+      "grad_norm": 0.39277076721191406,
+      "learning_rate": 3e-05,
+      "loss": 3.1385,
+      "step": 2550
+    },
+    {
+      "epoch": 0.8271725032425421,
+      "grad_norm": 0.3532819449901581,
+      "learning_rate": 3e-05,
+      "loss": 2.7268,
+      "step": 2551
+    },
+    {
+      "epoch": 0.8274967574578469,
+      "grad_norm": 0.31970566511154175,
+      "learning_rate": 3e-05,
+      "loss": 2.2631,
+      "step": 2552
+    },
+    {
+      "epoch": 0.8278210116731517,
+      "grad_norm": 0.40844935178756714,
+      "learning_rate": 3e-05,
+      "loss": 2.8489,
+      "step": 2553
+    },
+    {
+      "epoch": 0.8281452658884566,
+      "grad_norm": 0.4930250942707062,
+      "learning_rate": 3e-05,
+      "loss": 3.6845,
+      "step": 2554
+    },
+    {
+      "epoch": 0.8284695201037614,
+      "grad_norm": 0.3034558892250061,
+      "learning_rate": 3e-05,
+      "loss": 2.3326,
+      "step": 2555
+    },
+    {
+      "epoch": 0.8287937743190662,
+      "grad_norm": 0.31118255853652954,
+      "learning_rate": 3e-05,
+      "loss": 2.2963,
+      "step": 2556
+    },
+    {
+      "epoch": 0.829118028534371,
+      "grad_norm": 0.35927578806877136,
+      "learning_rate": 3e-05,
+      "loss": 2.8695,
+      "step": 2557
+    },
+    {
+      "epoch": 0.8294422827496758,
+      "grad_norm": 0.3580382466316223,
+      "learning_rate": 3e-05,
+      "loss": 2.559,
+      "step": 2558
+    },
+    {
+      "epoch": 0.8297665369649806,
+      "grad_norm": 0.39714518189430237,
+      "learning_rate": 3e-05,
+      "loss": 3.2716,
+      "step": 2559
+    },
+    {
+      "epoch": 0.8300907911802854,
+      "grad_norm": 0.37651127576828003,
+      "learning_rate": 3e-05,
+      "loss": 2.722,
+      "step": 2560
+    },
+    {
+      "epoch": 0.8304150453955902,
+      "grad_norm": 0.3178909122943878,
+      "learning_rate": 3e-05,
+      "loss": 2.0989,
+      "step": 2561
+    },
+    {
+      "epoch": 0.830739299610895,
+      "grad_norm": 0.46616071462631226,
+      "learning_rate": 3e-05,
+      "loss": 2.9512,
+      "step": 2562
+    },
+    {
+      "epoch": 0.8310635538261998,
+      "grad_norm": 0.47119322419166565,
+      "learning_rate": 3e-05,
+      "loss": 2.9507,
+      "step": 2563
+    },
+    {
+      "epoch": 0.8313878080415046,
+      "grad_norm": 0.3156626224517822,
+      "learning_rate": 3e-05,
+      "loss": 2.4676,
+      "step": 2564
+    },
+    {
+      "epoch": 0.8317120622568094,
+      "grad_norm": 0.3946497440338135,
+      "learning_rate": 3e-05,
+      "loss": 2.7673,
+      "step": 2565
+    },
+    {
+      "epoch": 0.8320363164721142,
+      "grad_norm": 0.47262758016586304,
+      "learning_rate": 3e-05,
+      "loss": 3.3565,
+      "step": 2566
+    },
+    {
+      "epoch": 0.832360570687419,
+      "grad_norm": 0.3285655677318573,
+      "learning_rate": 3e-05,
+      "loss": 2.3566,
+      "step": 2567
+    },
+    {
+      "epoch": 0.8326848249027238,
+      "grad_norm": 0.4465760290622711,
+      "learning_rate": 3e-05,
+      "loss": 2.6802,
+      "step": 2568
+    },
+    {
+      "epoch": 0.8330090791180286,
+      "grad_norm": 0.518268883228302,
+      "learning_rate": 3e-05,
+      "loss": 2.9258,
+      "step": 2569
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 0.43934017419815063,
+      "learning_rate": 3e-05,
+      "loss": 3.4946,
+      "step": 2570
+    },
+    {
+      "epoch": 0.8336575875486382,
+      "grad_norm": 0.4125627279281616,
+      "learning_rate": 3e-05,
+      "loss": 3.2574,
+      "step": 2571
+    },
+    {
+      "epoch": 0.833981841763943,
+      "grad_norm": 0.3177584409713745,
+      "learning_rate": 3e-05,
+      "loss": 2.3472,
+      "step": 2572
+    },
+    {
+      "epoch": 0.8343060959792478,
+      "grad_norm": 0.35633793473243713,
+      "learning_rate": 3e-05,
+      "loss": 2.5588,
+      "step": 2573
+    },
+    {
+      "epoch": 0.8346303501945526,
+      "grad_norm": 0.4205528199672699,
+      "learning_rate": 3e-05,
+      "loss": 2.6286,
+      "step": 2574
+    },
+    {
+      "epoch": 0.8349546044098574,
+      "grad_norm": 0.31786027550697327,
+      "learning_rate": 3e-05,
+      "loss": 2.1131,
+      "step": 2575
+    },
+    {
+      "epoch": 0.8352788586251622,
+      "grad_norm": 0.36921218037605286,
+      "learning_rate": 3e-05,
+      "loss": 2.6072,
+      "step": 2576
+    },
+    {
+      "epoch": 0.835603112840467,
+      "grad_norm": 0.7755422592163086,
+      "learning_rate": 3e-05,
+      "loss": 3.0951,
+      "step": 2577
+    },
+    {
+      "epoch": 0.8359273670557718,
+      "grad_norm": 0.3597908914089203,
+      "learning_rate": 3e-05,
+      "loss": 2.1161,
+      "step": 2578
+    },
+    {
+      "epoch": 0.8362516212710766,
+      "grad_norm": 0.30889153480529785,
+      "learning_rate": 3e-05,
+      "loss": 2.1687,
+      "step": 2579
+    },
+    {
+      "epoch": 0.8365758754863813,
+      "grad_norm": 0.4455258846282959,
+      "learning_rate": 3e-05,
+      "loss": 2.8582,
+      "step": 2580
+    },
+    {
+      "epoch": 0.8369001297016861,
+      "grad_norm": 0.42885535955429077,
+      "learning_rate": 3e-05,
+      "loss": 2.8689,
+      "step": 2581
+    },
+    {
+      "epoch": 0.8372243839169909,
+      "grad_norm": 0.39168646931648254,
+      "learning_rate": 3e-05,
+      "loss": 2.6542,
+      "step": 2582
+    },
+    {
+      "epoch": 0.8375486381322957,
+      "grad_norm": 0.636063814163208,
+      "learning_rate": 3e-05,
+      "loss": 3.3539,
+      "step": 2583
+    },
+    {
+      "epoch": 0.8378728923476005,
+      "grad_norm": 0.4509631097316742,
+      "learning_rate": 3e-05,
+      "loss": 2.9477,
+      "step": 2584
+    },
+    {
+      "epoch": 0.8381971465629053,
+      "grad_norm": 0.32295799255371094,
+      "learning_rate": 3e-05,
+      "loss": 2.3418,
+      "step": 2585
+    },
+    {
+      "epoch": 0.8385214007782101,
+      "grad_norm": 0.3878966271877289,
+      "learning_rate": 3e-05,
+      "loss": 2.4673,
+      "step": 2586
+    },
+    {
+      "epoch": 0.8388456549935149,
+      "grad_norm": 0.5693126916885376,
+      "learning_rate": 3e-05,
+      "loss": 3.5359,
+      "step": 2587
+    },
+    {
+      "epoch": 0.8391699092088197,
+      "grad_norm": 0.4089338183403015,
+      "learning_rate": 3e-05,
+      "loss": 2.9013,
+      "step": 2588
+    },
+    {
+      "epoch": 0.8394941634241245,
+      "grad_norm": 0.3847983479499817,
+      "learning_rate": 3e-05,
+      "loss": 2.6613,
+      "step": 2589
+    },
+    {
+      "epoch": 0.8398184176394293,
+      "grad_norm": 0.35274919867515564,
+      "learning_rate": 3e-05,
+      "loss": 2.6001,
+      "step": 2590
+    },
+    {
+      "epoch": 0.8401426718547341,
+      "grad_norm": 0.3868503272533417,
+      "learning_rate": 3e-05,
+      "loss": 2.7571,
+      "step": 2591
+    },
+    {
+      "epoch": 0.8404669260700389,
+      "grad_norm": 0.35535064339637756,
+      "learning_rate": 3e-05,
+      "loss": 2.4846,
+      "step": 2592
+    },
+    {
+      "epoch": 0.8407911802853437,
+      "grad_norm": 0.32761335372924805,
+      "learning_rate": 3e-05,
+      "loss": 2.3381,
+      "step": 2593
+    },
+    {
+      "epoch": 0.8411154345006485,
+      "grad_norm": 0.7374361753463745,
+      "learning_rate": 3e-05,
+      "loss": 3.3606,
+      "step": 2594
+    },
+    {
+      "epoch": 0.8414396887159533,
+      "grad_norm": 0.38239794969558716,
+      "learning_rate": 3e-05,
+      "loss": 2.5066,
+      "step": 2595
+    },
+    {
+      "epoch": 0.8417639429312581,
+      "grad_norm": 0.5284444093704224,
+      "learning_rate": 3e-05,
+      "loss": 3.1201,
+      "step": 2596
+    },
+    {
+      "epoch": 0.8420881971465629,
+      "grad_norm": 0.3591344654560089,
+      "learning_rate": 3e-05,
+      "loss": 2.5326,
+      "step": 2597
+    },
+    {
+      "epoch": 0.8424124513618677,
+      "grad_norm": 0.40678292512893677,
+      "learning_rate": 3e-05,
+      "loss": 2.8796,
+      "step": 2598
+    },
+    {
+      "epoch": 0.8427367055771725,
+      "grad_norm": 0.32669878005981445,
+      "learning_rate": 3e-05,
+      "loss": 2.0289,
+      "step": 2599
+    },
+    {
+      "epoch": 0.8430609597924773,
+      "grad_norm": 0.3424832224845886,
+      "learning_rate": 3e-05,
+      "loss": 2.6956,
+      "step": 2600
+    },
+    {
+      "epoch": 0.8433852140077821,
+      "grad_norm": 0.39183852076530457,
+      "learning_rate": 3e-05,
+      "loss": 2.6057,
+      "step": 2601
+    },
+    {
+      "epoch": 0.8437094682230869,
+      "grad_norm": 0.32268592715263367,
+      "learning_rate": 3e-05,
+      "loss": 2.3223,
+      "step": 2602
+    },
+    {
+      "epoch": 0.8440337224383917,
+      "grad_norm": 0.3534855842590332,
+      "learning_rate": 3e-05,
+      "loss": 2.7076,
+      "step": 2603
+    },
+    {
+      "epoch": 0.8443579766536965,
+      "grad_norm": 0.3454987704753876,
+      "learning_rate": 3e-05,
+      "loss": 2.7013,
+      "step": 2604
+    },
+    {
+      "epoch": 0.8446822308690013,
+      "grad_norm": 0.33690574765205383,
+      "learning_rate": 3e-05,
+      "loss": 2.3578,
+      "step": 2605
+    },
+    {
+      "epoch": 0.8450064850843061,
+      "grad_norm": 0.35702723264694214,
+      "learning_rate": 3e-05,
+      "loss": 3.0745,
+      "step": 2606
+    },
+    {
+      "epoch": 0.8453307392996109,
+      "grad_norm": 0.385616660118103,
+      "learning_rate": 3e-05,
+      "loss": 3.1105,
+      "step": 2607
+    },
+    {
+      "epoch": 0.8456549935149157,
+      "grad_norm": 0.402473121881485,
+      "learning_rate": 3e-05,
+      "loss": 3.0567,
+      "step": 2608
+    },
+    {
+      "epoch": 0.8459792477302205,
+      "grad_norm": 0.40330013632774353,
+      "learning_rate": 3e-05,
+      "loss": 2.8535,
+      "step": 2609
+    },
+    {
+      "epoch": 0.8463035019455253,
+      "grad_norm": 0.4428005516529083,
+      "learning_rate": 3e-05,
+      "loss": 2.8035,
+      "step": 2610
+    },
+    {
+      "epoch": 0.8466277561608301,
+      "grad_norm": 0.4122726023197174,
+      "learning_rate": 3e-05,
+      "loss": 2.8167,
+      "step": 2611
+    },
+    {
+      "epoch": 0.8469520103761349,
+      "grad_norm": 0.443154901266098,
+      "learning_rate": 3e-05,
+      "loss": 2.2668,
+      "step": 2612
+    },
+    {
+      "epoch": 0.8472762645914397,
+      "grad_norm": 0.35798487067222595,
+      "learning_rate": 3e-05,
+      "loss": 2.6018,
+      "step": 2613
+    },
+    {
+      "epoch": 0.8476005188067445,
+      "grad_norm": 0.34277424216270447,
+      "learning_rate": 3e-05,
+      "loss": 2.4739,
+      "step": 2614
+    },
+    {
+      "epoch": 0.8479247730220493,
+      "grad_norm": 0.6179741024971008,
+      "learning_rate": 3e-05,
+      "loss": 3.2103,
+      "step": 2615
+    },
+    {
+      "epoch": 0.8482490272373541,
+      "grad_norm": 0.3809232711791992,
+      "learning_rate": 3e-05,
+      "loss": 2.5509,
+      "step": 2616
+    },
+    {
+      "epoch": 0.8485732814526589,
+      "grad_norm": 0.3152162730693817,
+      "learning_rate": 3e-05,
+      "loss": 2.1379,
+      "step": 2617
+    },
+    {
+      "epoch": 0.8488975356679637,
+      "grad_norm": 0.4666021466255188,
+      "learning_rate": 3e-05,
+      "loss": 2.6291,
+      "step": 2618
+    },
+    {
+      "epoch": 0.8492217898832685,
+      "grad_norm": 0.35232922434806824,
+      "learning_rate": 3e-05,
+      "loss": 2.5323,
+      "step": 2619
+    },
+    {
+      "epoch": 0.8495460440985733,
+      "grad_norm": 0.41973239183425903,
+      "learning_rate": 3e-05,
+      "loss": 2.6293,
+      "step": 2620
+    },
+    {
+      "epoch": 0.8498702983138781,
+      "grad_norm": 0.6264311671257019,
+      "learning_rate": 3e-05,
+      "loss": 3.004,
+      "step": 2621
+    },
+    {
+      "epoch": 0.8501945525291829,
+      "grad_norm": 0.32713234424591064,
+      "learning_rate": 3e-05,
+      "loss": 2.5416,
+      "step": 2622
+    },
+    {
+      "epoch": 0.8505188067444877,
+      "grad_norm": 0.46447426080703735,
+      "learning_rate": 3e-05,
+      "loss": 2.6433,
+      "step": 2623
+    },
+    {
+      "epoch": 0.8508430609597925,
+      "grad_norm": 0.3079228103160858,
+      "learning_rate": 3e-05,
+      "loss": 2.3274,
+      "step": 2624
+    },
+    {
+      "epoch": 0.8511673151750972,
+      "grad_norm": 0.3474709093570709,
+      "learning_rate": 3e-05,
+      "loss": 2.6277,
+      "step": 2625
+    },
+    {
+      "epoch": 0.851491569390402,
+      "grad_norm": 0.4307585656642914,
+      "learning_rate": 3e-05,
+      "loss": 2.4052,
+      "step": 2626
+    },
+    {
+      "epoch": 0.8518158236057068,
+      "grad_norm": 0.3592437207698822,
+      "learning_rate": 3e-05,
+      "loss": 2.5681,
+      "step": 2627
+    },
+    {
+      "epoch": 0.8521400778210116,
+      "grad_norm": 0.4270443618297577,
+      "learning_rate": 3e-05,
+      "loss": 3.2526,
+      "step": 2628
+    },
+    {
+      "epoch": 0.8524643320363164,
+      "grad_norm": 0.4876168370246887,
+      "learning_rate": 3e-05,
+      "loss": 3.0276,
+      "step": 2629
+    },
+    {
+      "epoch": 0.8527885862516212,
+      "grad_norm": 0.37931084632873535,
+      "learning_rate": 3e-05,
+      "loss": 2.5381,
+      "step": 2630
+    },
+    {
+      "epoch": 0.853112840466926,
+      "grad_norm": 0.2973184883594513,
+      "learning_rate": 3e-05,
+      "loss": 2.1614,
+      "step": 2631
+    },
+    {
+      "epoch": 0.8534370946822308,
+      "grad_norm": 0.29142382740974426,
+      "learning_rate": 3e-05,
+      "loss": 2.217,
+      "step": 2632
+    },
+    {
+      "epoch": 0.8537613488975356,
+      "grad_norm": 0.4031103849411011,
+      "learning_rate": 3e-05,
+      "loss": 2.4986,
+      "step": 2633
+    },
+    {
+      "epoch": 0.8540856031128404,
+      "grad_norm": 0.37009477615356445,
+      "learning_rate": 3e-05,
+      "loss": 2.1994,
+      "step": 2634
+    },
+    {
+      "epoch": 0.8544098573281452,
+      "grad_norm": 0.3675681948661804,
+      "learning_rate": 3e-05,
+      "loss": 2.6726,
+      "step": 2635
+    },
+    {
+      "epoch": 0.85473411154345,
+      "grad_norm": 0.3025639057159424,
+      "learning_rate": 3e-05,
+      "loss": 2.2346,
+      "step": 2636
+    },
+    {
+      "epoch": 0.8550583657587548,
+      "grad_norm": 0.4243810772895813,
+      "learning_rate": 3e-05,
+      "loss": 2.8755,
+      "step": 2637
+    },
+    {
+      "epoch": 0.8553826199740596,
+      "grad_norm": 0.3928433656692505,
+      "learning_rate": 3e-05,
+      "loss": 2.5225,
+      "step": 2638
+    },
+    {
+      "epoch": 0.8557068741893644,
+      "grad_norm": 0.3036589026451111,
+      "learning_rate": 3e-05,
+      "loss": 2.124,
+      "step": 2639
+    },
+    {
+      "epoch": 0.8560311284046692,
+      "grad_norm": 0.39694446325302124,
+      "learning_rate": 3e-05,
+      "loss": 3.1091,
+      "step": 2640
+    },
+    {
+      "epoch": 0.856355382619974,
+      "grad_norm": 0.45487695932388306,
+      "learning_rate": 3e-05,
+      "loss": 2.9901,
+      "step": 2641
+    },
+    {
+      "epoch": 0.8566796368352788,
+      "grad_norm": 0.28316447138786316,
+      "learning_rate": 3e-05,
+      "loss": 1.9551,
+      "step": 2642
+    },
+    {
+      "epoch": 0.8570038910505836,
+      "grad_norm": 0.42458364367485046,
+      "learning_rate": 3e-05,
+      "loss": 3.1489,
+      "step": 2643
+    },
+    {
+      "epoch": 0.8573281452658884,
+      "grad_norm": 0.513354480266571,
+      "learning_rate": 3e-05,
+      "loss": 2.7001,
+      "step": 2644
+    },
+    {
+      "epoch": 0.8576523994811932,
+      "grad_norm": 0.3946242034435272,
+      "learning_rate": 3e-05,
+      "loss": 2.6518,
+      "step": 2645
+    },
+    {
+      "epoch": 0.857976653696498,
+      "grad_norm": 0.3055664896965027,
+      "learning_rate": 3e-05,
+      "loss": 2.4231,
+      "step": 2646
+    },
+    {
+      "epoch": 0.8583009079118028,
+      "grad_norm": 0.38807713985443115,
+      "learning_rate": 3e-05,
+      "loss": 2.7147,
+      "step": 2647
+    },
+    {
+      "epoch": 0.8586251621271076,
+      "grad_norm": 0.5906252264976501,
+      "learning_rate": 3e-05,
+      "loss": 2.8098,
+      "step": 2648
+    },
+    {
+      "epoch": 0.8589494163424124,
+      "grad_norm": 0.3122623562812805,
+      "learning_rate": 3e-05,
+      "loss": 2.4199,
+      "step": 2649
+    },
+    {
+      "epoch": 0.8592736705577172,
+      "grad_norm": 0.48882147669792175,
+      "learning_rate": 3e-05,
+      "loss": 3.0228,
+      "step": 2650
+    },
+    {
+      "epoch": 0.8595979247730221,
+      "grad_norm": 0.4642896354198456,
+      "learning_rate": 3e-05,
+      "loss": 2.7884,
+      "step": 2651
+    },
+    {
+      "epoch": 0.8599221789883269,
+      "grad_norm": 0.3939104974269867,
+      "learning_rate": 3e-05,
+      "loss": 2.8917,
+      "step": 2652
+    },
+    {
+      "epoch": 0.8602464332036317,
+      "grad_norm": 0.36345723271369934,
+      "learning_rate": 3e-05,
+      "loss": 2.4608,
+      "step": 2653
+    },
+    {
+      "epoch": 0.8605706874189365,
+      "grad_norm": 0.5185490250587463,
+      "learning_rate": 3e-05,
+      "loss": 3.1858,
+      "step": 2654
+    },
+    {
+      "epoch": 0.8608949416342413,
+      "grad_norm": 0.3821839690208435,
+      "learning_rate": 3e-05,
+      "loss": 2.4788,
+      "step": 2655
+    },
+    {
+      "epoch": 0.8612191958495461,
+      "grad_norm": 0.34107857942581177,
+      "learning_rate": 3e-05,
+      "loss": 2.9056,
+      "step": 2656
+    },
+    {
+      "epoch": 0.8615434500648509,
+      "grad_norm": 0.5041527152061462,
+      "learning_rate": 3e-05,
+      "loss": 2.9933,
+      "step": 2657
+    },
+    {
+      "epoch": 0.8618677042801557,
+      "grad_norm": 0.4867437779903412,
+      "learning_rate": 3e-05,
+      "loss": 3.3626,
+      "step": 2658
+    },
+    {
+      "epoch": 0.8621919584954605,
+      "grad_norm": 0.37666672468185425,
+      "learning_rate": 3e-05,
+      "loss": 2.917,
+      "step": 2659
+    },
+    {
+      "epoch": 0.8625162127107653,
+      "grad_norm": 0.4188278913497925,
+      "learning_rate": 3e-05,
+      "loss": 3.0055,
+      "step": 2660
+    },
+    {
+      "epoch": 0.8628404669260701,
+      "grad_norm": 0.46115079522132874,
+      "learning_rate": 3e-05,
+      "loss": 2.466,
+      "step": 2661
+    },
+    {
+      "epoch": 0.8631647211413749,
+      "grad_norm": 0.4052063524723053,
+      "learning_rate": 3e-05,
+      "loss": 2.7304,
+      "step": 2662
+    },
+    {
+      "epoch": 0.8634889753566797,
+      "grad_norm": 0.4681519865989685,
+      "learning_rate": 3e-05,
+      "loss": 2.8438,
+      "step": 2663
+    },
+    {
+      "epoch": 0.8638132295719845,
+      "grad_norm": 0.459979385137558,
+      "learning_rate": 3e-05,
+      "loss": 2.6697,
+      "step": 2664
+    },
+    {
+      "epoch": 0.8641374837872893,
+      "grad_norm": 0.3445543050765991,
+      "learning_rate": 3e-05,
+      "loss": 2.3963,
+      "step": 2665
+    },
+    {
+      "epoch": 0.8644617380025941,
+      "grad_norm": 0.39826521277427673,
+      "learning_rate": 3e-05,
+      "loss": 3.0517,
+      "step": 2666
+    },
+    {
+      "epoch": 0.8647859922178989,
+      "grad_norm": 0.3604954779148102,
+      "learning_rate": 3e-05,
+      "loss": 2.6375,
+      "step": 2667
+    },
+    {
+      "epoch": 0.8651102464332037,
+      "grad_norm": 0.3214665353298187,
+      "learning_rate": 3e-05,
+      "loss": 2.3259,
+      "step": 2668
+    },
+    {
+      "epoch": 0.8654345006485085,
+      "grad_norm": 0.42449045181274414,
+      "learning_rate": 3e-05,
+      "loss": 2.9513,
+      "step": 2669
+    },
+    {
+      "epoch": 0.8657587548638133,
+      "grad_norm": 0.3370482623577118,
+      "learning_rate": 3e-05,
+      "loss": 2.4092,
+      "step": 2670
+    },
+    {
+      "epoch": 0.8660830090791181,
+      "grad_norm": 0.39176449179649353,
+      "learning_rate": 3e-05,
+      "loss": 2.5523,
+      "step": 2671
+    },
+    {
+      "epoch": 0.8664072632944229,
+      "grad_norm": 0.5043333172798157,
+      "learning_rate": 3e-05,
+      "loss": 2.9506,
+      "step": 2672
+    },
+    {
+      "epoch": 0.8667315175097277,
+      "grad_norm": 0.35785913467407227,
+      "learning_rate": 3e-05,
+      "loss": 2.5344,
+      "step": 2673
+    },
+    {
+      "epoch": 0.8670557717250325,
+      "grad_norm": 0.46933987736701965,
+      "learning_rate": 3e-05,
+      "loss": 2.8579,
+      "step": 2674
+    },
+    {
+      "epoch": 0.8673800259403373,
+      "grad_norm": 0.30082666873931885,
+      "learning_rate": 3e-05,
+      "loss": 2.2396,
+      "step": 2675
+    },
+    {
+      "epoch": 0.867704280155642,
+      "grad_norm": 0.30005112290382385,
+      "learning_rate": 3e-05,
+      "loss": 2.2228,
+      "step": 2676
+    },
+    {
+      "epoch": 0.8680285343709468,
+      "grad_norm": 0.2920472323894501,
+      "learning_rate": 3e-05,
+      "loss": 2.3787,
+      "step": 2677
+    },
+    {
+      "epoch": 0.8683527885862516,
+      "grad_norm": 0.36819520592689514,
+      "learning_rate": 3e-05,
+      "loss": 2.5358,
+      "step": 2678
+    },
+    {
+      "epoch": 0.8686770428015564,
+      "grad_norm": 0.31029748916625977,
+      "learning_rate": 3e-05,
+      "loss": 2.2721,
+      "step": 2679
+    },
+    {
+      "epoch": 0.8690012970168612,
+      "grad_norm": 0.3202894926071167,
+      "learning_rate": 3e-05,
+      "loss": 2.4968,
+      "step": 2680
+    },
+    {
+      "epoch": 0.869325551232166,
+      "grad_norm": 0.5121805667877197,
+      "learning_rate": 3e-05,
+      "loss": 2.7361,
+      "step": 2681
+    },
+    {
+      "epoch": 0.8696498054474708,
+      "grad_norm": 0.34481188654899597,
+      "learning_rate": 3e-05,
+      "loss": 2.852,
+      "step": 2682
+    },
+    {
+      "epoch": 0.8699740596627756,
+      "grad_norm": 0.4321422278881073,
+      "learning_rate": 3e-05,
+      "loss": 3.2391,
+      "step": 2683
+    },
+    {
+      "epoch": 0.8702983138780804,
+      "grad_norm": 0.3909590244293213,
+      "learning_rate": 3e-05,
+      "loss": 3.1328,
+      "step": 2684
+    },
+    {
+      "epoch": 0.8706225680933852,
+      "grad_norm": 0.4062865078449249,
+      "learning_rate": 3e-05,
+      "loss": 2.4684,
+      "step": 2685
+    },
+    {
+      "epoch": 0.87094682230869,
+      "grad_norm": 0.3624367415904999,
+      "learning_rate": 3e-05,
+      "loss": 2.2336,
+      "step": 2686
+    },
+    {
+      "epoch": 0.8712710765239948,
+      "grad_norm": 0.2776319980621338,
+      "learning_rate": 3e-05,
+      "loss": 2.0083,
+      "step": 2687
+    },
+    {
+      "epoch": 0.8715953307392996,
+      "grad_norm": 0.37192749977111816,
+      "learning_rate": 3e-05,
+      "loss": 2.855,
+      "step": 2688
+    },
+    {
+      "epoch": 0.8719195849546044,
+      "grad_norm": 0.41802042722702026,
+      "learning_rate": 3e-05,
+      "loss": 2.732,
+      "step": 2689
+    },
+    {
+      "epoch": 0.8722438391699092,
+      "grad_norm": 0.5911200642585754,
+      "learning_rate": 3e-05,
+      "loss": 3.6086,
+      "step": 2690
+    },
+    {
+      "epoch": 0.872568093385214,
+      "grad_norm": 0.8825404644012451,
+      "learning_rate": 3e-05,
+      "loss": 3.6548,
+      "step": 2691
+    },
+    {
+      "epoch": 0.8728923476005188,
+      "grad_norm": 0.43319615721702576,
+      "learning_rate": 3e-05,
+      "loss": 2.7588,
+      "step": 2692
+    },
+    {
+      "epoch": 0.8732166018158236,
+      "grad_norm": 0.31458011269569397,
+      "learning_rate": 3e-05,
+      "loss": 2.3721,
+      "step": 2693
+    },
+    {
+      "epoch": 0.8735408560311284,
+      "grad_norm": 0.4340474307537079,
+      "learning_rate": 3e-05,
+      "loss": 2.3718,
+      "step": 2694
+    },
+    {
+      "epoch": 0.8738651102464332,
+      "grad_norm": 0.4532354176044464,
+      "learning_rate": 3e-05,
+      "loss": 2.359,
+      "step": 2695
+    },
+    {
+      "epoch": 0.874189364461738,
+      "grad_norm": 0.7662169933319092,
+      "learning_rate": 3e-05,
+      "loss": 3.4521,
+      "step": 2696
+    },
+    {
+      "epoch": 0.8745136186770428,
+      "grad_norm": 0.3886234164237976,
+      "learning_rate": 3e-05,
+      "loss": 2.5929,
+      "step": 2697
+    },
+    {
+      "epoch": 0.8748378728923476,
+      "grad_norm": 0.42308375239372253,
+      "learning_rate": 3e-05,
+      "loss": 2.298,
+      "step": 2698
+    },
+    {
+      "epoch": 0.8751621271076524,
+      "grad_norm": 0.37737414240837097,
+      "learning_rate": 3e-05,
+      "loss": 2.6017,
+      "step": 2699
+    },
+    {
+      "epoch": 0.8754863813229572,
+      "grad_norm": 0.481217622756958,
+      "learning_rate": 3e-05,
+      "loss": 2.8808,
+      "step": 2700
+    },
+    {
+      "epoch": 0.875810635538262,
+      "grad_norm": 0.3273676931858063,
+      "learning_rate": 3e-05,
+      "loss": 2.1442,
+      "step": 2701
+    },
+    {
+      "epoch": 0.8761348897535668,
+      "grad_norm": 0.3413366377353668,
+      "learning_rate": 3e-05,
+      "loss": 2.6433,
+      "step": 2702
+    },
+    {
+      "epoch": 0.8764591439688716,
+      "grad_norm": 0.6859429478645325,
+      "learning_rate": 3e-05,
+      "loss": 3.0193,
+      "step": 2703
+    },
+    {
+      "epoch": 0.8767833981841764,
+      "grad_norm": 0.4487481713294983,
+      "learning_rate": 3e-05,
+      "loss": 3.342,
+      "step": 2704
+    },
+    {
+      "epoch": 0.8771076523994812,
+      "grad_norm": 0.3604808449745178,
+      "learning_rate": 3e-05,
+      "loss": 2.3343,
+      "step": 2705
+    },
+    {
+      "epoch": 0.877431906614786,
+      "grad_norm": 0.43042922019958496,
+      "learning_rate": 3e-05,
+      "loss": 2.392,
+      "step": 2706
+    },
+    {
+      "epoch": 0.8777561608300908,
+      "grad_norm": 0.4591202139854431,
+      "learning_rate": 3e-05,
+      "loss": 2.7049,
+      "step": 2707
+    },
+    {
+      "epoch": 0.8780804150453956,
+      "grad_norm": 0.3022620975971222,
+      "learning_rate": 3e-05,
+      "loss": 2.2851,
+      "step": 2708
+    },
+    {
+      "epoch": 0.8784046692607004,
+      "grad_norm": 0.33546558022499084,
+      "learning_rate": 3e-05,
+      "loss": 2.1066,
+      "step": 2709
+    },
+    {
+      "epoch": 0.8787289234760052,
+      "grad_norm": 0.5217865705490112,
+      "learning_rate": 3e-05,
+      "loss": 2.9843,
+      "step": 2710
+    },
+    {
+      "epoch": 0.87905317769131,
+      "grad_norm": 0.3478190302848816,
+      "learning_rate": 3e-05,
+      "loss": 2.6487,
+      "step": 2711
+    },
+    {
+      "epoch": 0.8793774319066148,
+      "grad_norm": 0.6613021492958069,
+      "learning_rate": 3e-05,
+      "loss": 3.5568,
+      "step": 2712
+    },
+    {
+      "epoch": 0.8797016861219196,
+      "grad_norm": 0.30653640627861023,
+      "learning_rate": 3e-05,
+      "loss": 2.3104,
+      "step": 2713
+    },
+    {
+      "epoch": 0.8800259403372244,
+      "grad_norm": 0.27938273549079895,
+      "learning_rate": 3e-05,
+      "loss": 2.1944,
+      "step": 2714
+    },
+    {
+      "epoch": 0.8803501945525292,
+      "grad_norm": 0.46174314618110657,
+      "learning_rate": 3e-05,
+      "loss": 2.8181,
+      "step": 2715
+    },
+    {
+      "epoch": 0.880674448767834,
+      "grad_norm": 0.45674929022789,
+      "learning_rate": 3e-05,
+      "loss": 3.3369,
+      "step": 2716
+    },
+    {
+      "epoch": 0.8809987029831388,
+      "grad_norm": 0.3938347399234772,
+      "learning_rate": 3e-05,
+      "loss": 3.0402,
+      "step": 2717
+    },
+    {
+      "epoch": 0.8813229571984436,
+      "grad_norm": 0.4123966097831726,
+      "learning_rate": 3e-05,
+      "loss": 3.1407,
+      "step": 2718
+    },
+    {
+      "epoch": 0.8816472114137484,
+      "grad_norm": 0.37980690598487854,
+      "learning_rate": 3e-05,
+      "loss": 2.5151,
+      "step": 2719
+    },
+    {
+      "epoch": 0.8819714656290532,
+      "grad_norm": 0.34080442786216736,
+      "learning_rate": 3e-05,
+      "loss": 2.5687,
+      "step": 2720
+    },
+    {
+      "epoch": 0.882295719844358,
+      "grad_norm": 0.5081164240837097,
+      "learning_rate": 3e-05,
+      "loss": 2.9826,
+      "step": 2721
+    },
+    {
+      "epoch": 0.8826199740596627,
+      "grad_norm": 0.4955286979675293,
+      "learning_rate": 3e-05,
+      "loss": 3.1371,
+      "step": 2722
+    },
+    {
+      "epoch": 0.8829442282749675,
+      "grad_norm": 0.3299684524536133,
+      "learning_rate": 3e-05,
+      "loss": 2.5587,
+      "step": 2723
+    },
+    {
+      "epoch": 0.8832684824902723,
+      "grad_norm": 0.4345269799232483,
+      "learning_rate": 3e-05,
+      "loss": 2.7284,
+      "step": 2724
+    },
+    {
+      "epoch": 0.8835927367055771,
+      "grad_norm": 0.4527815878391266,
+      "learning_rate": 3e-05,
+      "loss": 3.0484,
+      "step": 2725
+    },
+    {
+      "epoch": 0.8839169909208819,
+      "grad_norm": 0.447897732257843,
+      "learning_rate": 3e-05,
+      "loss": 3.3721,
+      "step": 2726
+    },
+    {
+      "epoch": 0.8842412451361867,
+      "grad_norm": 0.6444581151008606,
+      "learning_rate": 3e-05,
+      "loss": 3.6398,
+      "step": 2727
+    },
+    {
+      "epoch": 0.8845654993514915,
+      "grad_norm": 0.29388436675071716,
+      "learning_rate": 3e-05,
+      "loss": 2.2548,
+      "step": 2728
+    },
+    {
+      "epoch": 0.8848897535667963,
+      "grad_norm": 0.4167473316192627,
+      "learning_rate": 3e-05,
+      "loss": 2.7488,
+      "step": 2729
+    },
+    {
+      "epoch": 0.8852140077821011,
+      "grad_norm": 0.42017480731010437,
+      "learning_rate": 3e-05,
+      "loss": 3.0097,
+      "step": 2730
+    },
+    {
+      "epoch": 0.8855382619974059,
+      "grad_norm": 0.5737587809562683,
+      "learning_rate": 3e-05,
+      "loss": 3.2219,
+      "step": 2731
+    },
+    {
+      "epoch": 0.8858625162127107,
+      "grad_norm": 0.37995728850364685,
+      "learning_rate": 3e-05,
+      "loss": 2.8443,
+      "step": 2732
+    },
+    {
+      "epoch": 0.8861867704280155,
+      "grad_norm": 0.4261234402656555,
+      "learning_rate": 3e-05,
+      "loss": 2.7685,
+      "step": 2733
+    },
+    {
+      "epoch": 0.8865110246433203,
+      "grad_norm": 0.42956334352493286,
+      "learning_rate": 3e-05,
+      "loss": 2.6873,
+      "step": 2734
+    },
+    {
+      "epoch": 0.8868352788586251,
+      "grad_norm": 0.38059473037719727,
+      "learning_rate": 3e-05,
+      "loss": 2.5488,
+      "step": 2735
+    },
+    {
+      "epoch": 0.8871595330739299,
+      "grad_norm": 0.3475302457809448,
+      "learning_rate": 3e-05,
+      "loss": 2.2183,
+      "step": 2736
+    },
+    {
+      "epoch": 0.8874837872892347,
+      "grad_norm": 0.47955629229545593,
+      "learning_rate": 3e-05,
+      "loss": 2.8236,
+      "step": 2737
+    },
+    {
+      "epoch": 0.8878080415045395,
+      "grad_norm": 0.475238174200058,
+      "learning_rate": 3e-05,
+      "loss": 2.94,
+      "step": 2738
+    },
+    {
+      "epoch": 0.8881322957198443,
+      "grad_norm": 0.47215595841407776,
+      "learning_rate": 3e-05,
+      "loss": 2.528,
+      "step": 2739
+    },
+    {
+      "epoch": 0.8884565499351491,
+      "grad_norm": 0.3843601942062378,
+      "learning_rate": 3e-05,
+      "loss": 3.0572,
+      "step": 2740
+    },
+    {
+      "epoch": 0.8887808041504539,
+      "grad_norm": 0.39506757259368896,
+      "learning_rate": 3e-05,
+      "loss": 2.6465,
+      "step": 2741
+    },
+    {
+      "epoch": 0.8891050583657587,
+      "grad_norm": 0.3993166983127594,
+      "learning_rate": 3e-05,
+      "loss": 2.8197,
+      "step": 2742
+    },
+    {
+      "epoch": 0.8894293125810635,
+      "grad_norm": 0.3837796747684479,
+      "learning_rate": 3e-05,
+      "loss": 2.7383,
+      "step": 2743
+    },
+    {
+      "epoch": 0.8897535667963683,
+      "grad_norm": 0.3068723976612091,
+      "learning_rate": 3e-05,
+      "loss": 2.2484,
+      "step": 2744
+    },
+    {
+      "epoch": 0.8900778210116731,
+      "grad_norm": 0.4363597631454468,
+      "learning_rate": 3e-05,
+      "loss": 2.9057,
+      "step": 2745
+    },
+    {
+      "epoch": 0.8904020752269779,
+      "grad_norm": 0.3771510124206543,
+      "learning_rate": 3e-05,
+      "loss": 2.724,
+      "step": 2746
+    },
+    {
+      "epoch": 0.8907263294422828,
+      "grad_norm": 0.432451456785202,
+      "learning_rate": 3e-05,
+      "loss": 2.9468,
+      "step": 2747
+    },
+    {
+      "epoch": 0.8910505836575876,
+      "grad_norm": 0.3406469523906708,
+      "learning_rate": 3e-05,
+      "loss": 2.4589,
+      "step": 2748
+    },
+    {
+      "epoch": 0.8913748378728924,
+      "grad_norm": 0.4632258117198944,
+      "learning_rate": 3e-05,
+      "loss": 2.8529,
+      "step": 2749
+    },
+    {
+      "epoch": 0.8916990920881972,
+      "grad_norm": 0.471057653427124,
+      "learning_rate": 3e-05,
+      "loss": 2.6885,
+      "step": 2750
+    },
+    {
+      "epoch": 0.892023346303502,
+      "grad_norm": 0.44260072708129883,
+      "learning_rate": 3e-05,
+      "loss": 2.8275,
+      "step": 2751
+    },
+    {
+      "epoch": 0.8923476005188068,
+      "grad_norm": 0.5029962658882141,
+      "learning_rate": 3e-05,
+      "loss": 3.2463,
+      "step": 2752
+    },
+    {
+      "epoch": 0.8926718547341116,
+      "grad_norm": 0.30783358216285706,
+      "learning_rate": 3e-05,
+      "loss": 2.2553,
+      "step": 2753
+    },
+    {
+      "epoch": 0.8929961089494164,
+      "grad_norm": 0.5737777352333069,
+      "learning_rate": 3e-05,
+      "loss": 3.1743,
+      "step": 2754
+    },
+    {
+      "epoch": 0.8933203631647212,
+      "grad_norm": 0.49672478437423706,
+      "learning_rate": 3e-05,
+      "loss": 3.0615,
+      "step": 2755
+    },
+    {
+      "epoch": 0.893644617380026,
+      "grad_norm": 0.4389769732952118,
+      "learning_rate": 3e-05,
+      "loss": 2.6685,
+      "step": 2756
+    },
+    {
+      "epoch": 0.8939688715953308,
+      "grad_norm": 0.3022623360157013,
+      "learning_rate": 3e-05,
+      "loss": 2.1361,
+      "step": 2757
+    },
+    {
+      "epoch": 0.8942931258106356,
+      "grad_norm": 0.4630208909511566,
+      "learning_rate": 3e-05,
+      "loss": 2.9968,
+      "step": 2758
+    },
+    {
+      "epoch": 0.8946173800259404,
+      "grad_norm": 0.36338722705841064,
+      "learning_rate": 3e-05,
+      "loss": 2.3642,
+      "step": 2759
+    },
+    {
+      "epoch": 0.8949416342412452,
+      "grad_norm": 0.337150901556015,
+      "learning_rate": 3e-05,
+      "loss": 2.3365,
+      "step": 2760
+    },
+    {
+      "epoch": 0.89526588845655,
+      "grad_norm": 0.3946084678173065,
+      "learning_rate": 3e-05,
+      "loss": 2.9026,
+      "step": 2761
+    },
+    {
+      "epoch": 0.8955901426718548,
+      "grad_norm": 0.34703055024147034,
+      "learning_rate": 3e-05,
+      "loss": 2.5678,
+      "step": 2762
+    },
+    {
+      "epoch": 0.8959143968871596,
+      "grad_norm": 0.5630971789360046,
+      "learning_rate": 3e-05,
+      "loss": 2.902,
+      "step": 2763
+    },
+    {
+      "epoch": 0.8962386511024644,
+      "grad_norm": 0.3851115107536316,
+      "learning_rate": 3e-05,
+      "loss": 2.6919,
+      "step": 2764
+    },
+    {
+      "epoch": 0.8965629053177692,
+      "grad_norm": 0.47223949432373047,
+      "learning_rate": 3e-05,
+      "loss": 3.1611,
+      "step": 2765
+    },
+    {
+      "epoch": 0.896887159533074,
+      "grad_norm": 0.4977733790874481,
+      "learning_rate": 3e-05,
+      "loss": 2.7404,
+      "step": 2766
+    },
+    {
+      "epoch": 0.8972114137483788,
+      "grad_norm": 0.3668880760669708,
+      "learning_rate": 3e-05,
+      "loss": 2.5405,
+      "step": 2767
+    },
+    {
+      "epoch": 0.8975356679636836,
+      "grad_norm": 0.3690524995326996,
+      "learning_rate": 3e-05,
+      "loss": 2.3651,
+      "step": 2768
+    },
+    {
+      "epoch": 0.8978599221789884,
+      "grad_norm": 0.48774898052215576,
+      "learning_rate": 3e-05,
+      "loss": 2.8718,
+      "step": 2769
+    },
+    {
+      "epoch": 0.8981841763942932,
+      "grad_norm": 0.3156971037387848,
+      "learning_rate": 3e-05,
+      "loss": 2.3762,
+      "step": 2770
+    },
+    {
+      "epoch": 0.898508430609598,
+      "grad_norm": 0.34952786564826965,
+      "learning_rate": 3e-05,
+      "loss": 2.389,
+      "step": 2771
+    },
+    {
+      "epoch": 0.8988326848249028,
+      "grad_norm": 0.3114722669124603,
+      "learning_rate": 3e-05,
+      "loss": 1.992,
+      "step": 2772
+    },
+    {
+      "epoch": 0.8991569390402075,
+      "grad_norm": 0.45942944288253784,
+      "learning_rate": 3e-05,
+      "loss": 2.6509,
+      "step": 2773
+    },
+    {
+      "epoch": 0.8994811932555123,
+      "grad_norm": 0.45895981788635254,
+      "learning_rate": 3e-05,
+      "loss": 2.8268,
+      "step": 2774
+    },
+    {
+      "epoch": 0.8998054474708171,
+      "grad_norm": 0.2971877157688141,
+      "learning_rate": 3e-05,
+      "loss": 2.1491,
+      "step": 2775
+    },
+    {
+      "epoch": 0.9001297016861219,
+      "grad_norm": 0.4318220317363739,
+      "learning_rate": 3e-05,
+      "loss": 2.7645,
+      "step": 2776
+    },
+    {
+      "epoch": 0.9004539559014267,
+      "grad_norm": 0.7703062891960144,
+      "learning_rate": 3e-05,
+      "loss": 3.4813,
+      "step": 2777
+    },
+    {
+      "epoch": 0.9007782101167315,
+      "grad_norm": 0.4141489863395691,
+      "learning_rate": 3e-05,
+      "loss": 3.1992,
+      "step": 2778
+    },
+    {
+      "epoch": 0.9011024643320363,
+      "grad_norm": 0.5435628294944763,
+      "learning_rate": 3e-05,
+      "loss": 2.6121,
+      "step": 2779
+    },
+    {
+      "epoch": 0.9014267185473411,
+      "grad_norm": 0.42822539806365967,
+      "learning_rate": 3e-05,
+      "loss": 2.386,
+      "step": 2780
+    },
+    {
+      "epoch": 0.9017509727626459,
+      "grad_norm": 0.5057844519615173,
+      "learning_rate": 3e-05,
+      "loss": 3.3995,
+      "step": 2781
+    },
+    {
+      "epoch": 0.9020752269779507,
+      "grad_norm": 0.3723487854003906,
+      "learning_rate": 3e-05,
+      "loss": 2.8418,
+      "step": 2782
+    },
+    {
+      "epoch": 0.9023994811932555,
+      "grad_norm": 0.44552138447761536,
+      "learning_rate": 3e-05,
+      "loss": 2.6905,
+      "step": 2783
+    },
+    {
+      "epoch": 0.9027237354085603,
+      "grad_norm": 0.3392035961151123,
+      "learning_rate": 3e-05,
+      "loss": 2.8403,
+      "step": 2784
+    },
+    {
+      "epoch": 0.9030479896238651,
+      "grad_norm": 0.3491925299167633,
+      "learning_rate": 3e-05,
+      "loss": 2.8085,
+      "step": 2785
+    },
+    {
+      "epoch": 0.9033722438391699,
+      "grad_norm": 0.42016199231147766,
+      "learning_rate": 3e-05,
+      "loss": 2.7321,
+      "step": 2786
+    },
+    {
+      "epoch": 0.9036964980544747,
+      "grad_norm": 0.35728853940963745,
+      "learning_rate": 3e-05,
+      "loss": 2.2008,
+      "step": 2787
+    },
+    {
+      "epoch": 0.9040207522697795,
+      "grad_norm": 0.638726532459259,
+      "learning_rate": 3e-05,
+      "loss": 2.5416,
+      "step": 2788
+    },
+    {
+      "epoch": 0.9043450064850843,
+      "grad_norm": 0.5147861242294312,
+      "learning_rate": 3e-05,
+      "loss": 2.8886,
+      "step": 2789
+    },
+    {
+      "epoch": 0.9046692607003891,
+      "grad_norm": 0.35540589690208435,
+      "learning_rate": 3e-05,
+      "loss": 2.4398,
+      "step": 2790
+    },
+    {
+      "epoch": 0.9049935149156939,
+      "grad_norm": 0.40316134691238403,
+      "learning_rate": 3e-05,
+      "loss": 2.6153,
+      "step": 2791
+    },
+    {
+      "epoch": 0.9053177691309987,
+      "grad_norm": 0.40525537729263306,
+      "learning_rate": 3e-05,
+      "loss": 2.8621,
+      "step": 2792
+    },
+    {
+      "epoch": 0.9056420233463035,
+      "grad_norm": 0.6487452387809753,
+      "learning_rate": 3e-05,
+      "loss": 3.0157,
+      "step": 2793
+    },
+    {
+      "epoch": 0.9059662775616083,
+      "grad_norm": 0.54863041639328,
+      "learning_rate": 3e-05,
+      "loss": 2.7304,
+      "step": 2794
+    },
+    {
+      "epoch": 0.9062905317769131,
+      "grad_norm": 0.3260003626346588,
+      "learning_rate": 3e-05,
+      "loss": 2.5967,
+      "step": 2795
+    },
+    {
+      "epoch": 0.9066147859922179,
+      "grad_norm": 0.3267848491668701,
+      "learning_rate": 3e-05,
+      "loss": 2.3937,
+      "step": 2796
+    },
+    {
+      "epoch": 0.9069390402075227,
+      "grad_norm": 0.39301320910453796,
+      "learning_rate": 3e-05,
+      "loss": 2.5481,
+      "step": 2797
+    },
+    {
+      "epoch": 0.9072632944228275,
+      "grad_norm": 0.33058902621269226,
+      "learning_rate": 3e-05,
+      "loss": 2.3044,
+      "step": 2798
+    },
+    {
+      "epoch": 0.9075875486381323,
+      "grad_norm": 0.7496035099029541,
+      "learning_rate": 3e-05,
+      "loss": 3.3504,
+      "step": 2799
+    },
+    {
+      "epoch": 0.9079118028534371,
+      "grad_norm": 0.49361148476600647,
+      "learning_rate": 3e-05,
+      "loss": 3.1654,
+      "step": 2800
+    },
+    {
+      "epoch": 0.9082360570687419,
+      "grad_norm": 0.4890223741531372,
+      "learning_rate": 3e-05,
+      "loss": 2.9366,
+      "step": 2801
+    },
+    {
+      "epoch": 0.9085603112840467,
+      "grad_norm": 0.3516950309276581,
+      "learning_rate": 3e-05,
+      "loss": 2.4701,
+      "step": 2802
+    },
+    {
+      "epoch": 0.9088845654993515,
+      "grad_norm": 0.6906164884567261,
+      "learning_rate": 3e-05,
+      "loss": 2.8397,
+      "step": 2803
+    },
+    {
+      "epoch": 0.9092088197146563,
+      "grad_norm": 0.3575665056705475,
+      "learning_rate": 3e-05,
+      "loss": 1.9648,
+      "step": 2804
+    },
+    {
+      "epoch": 0.9095330739299611,
+      "grad_norm": 0.569299578666687,
+      "learning_rate": 3e-05,
+      "loss": 2.791,
+      "step": 2805
+    },
+    {
+      "epoch": 0.9098573281452659,
+      "grad_norm": 0.4068557322025299,
+      "learning_rate": 3e-05,
+      "loss": 2.7442,
+      "step": 2806
+    },
+    {
+      "epoch": 0.9101815823605707,
+      "grad_norm": 0.4188474118709564,
+      "learning_rate": 3e-05,
+      "loss": 2.5359,
+      "step": 2807
+    },
+    {
+      "epoch": 0.9105058365758755,
+      "grad_norm": 0.3562527596950531,
+      "learning_rate": 3e-05,
+      "loss": 2.5579,
+      "step": 2808
+    },
+    {
+      "epoch": 0.9108300907911803,
+      "grad_norm": 0.4983190596103668,
+      "learning_rate": 3e-05,
+      "loss": 2.9727,
+      "step": 2809
+    },
+    {
+      "epoch": 0.9111543450064851,
+      "grad_norm": 0.3710606098175049,
+      "learning_rate": 3e-05,
+      "loss": 2.5107,
+      "step": 2810
+    },
+    {
+      "epoch": 0.9114785992217899,
+      "grad_norm": 0.6386035084724426,
+      "learning_rate": 3e-05,
+      "loss": 2.6865,
+      "step": 2811
+    },
+    {
+      "epoch": 0.9118028534370947,
+      "grad_norm": 0.5449885129928589,
+      "learning_rate": 3e-05,
+      "loss": 2.639,
+      "step": 2812
+    },
+    {
+      "epoch": 0.9121271076523995,
+      "grad_norm": 0.2692824602127075,
+      "learning_rate": 3e-05,
+      "loss": 1.9886,
+      "step": 2813
+    },
+    {
+      "epoch": 0.9124513618677043,
+      "grad_norm": 0.3680189549922943,
+      "learning_rate": 3e-05,
+      "loss": 2.8158,
+      "step": 2814
+    },
+    {
+      "epoch": 0.9127756160830091,
+      "grad_norm": 0.33792880177497864,
+      "learning_rate": 3e-05,
+      "loss": 2.36,
+      "step": 2815
+    },
+    {
+      "epoch": 0.9130998702983139,
+      "grad_norm": 0.5642290711402893,
+      "learning_rate": 3e-05,
+      "loss": 2.8606,
+      "step": 2816
+    },
+    {
+      "epoch": 0.9134241245136187,
+      "grad_norm": 0.4929906725883484,
+      "learning_rate": 3e-05,
+      "loss": 2.5755,
+      "step": 2817
+    },
+    {
+      "epoch": 0.9137483787289234,
+      "grad_norm": 0.47357362508773804,
+      "learning_rate": 3e-05,
+      "loss": 2.6013,
+      "step": 2818
+    },
+    {
+      "epoch": 0.9140726329442282,
+      "grad_norm": 0.2958976626396179,
+      "learning_rate": 3e-05,
+      "loss": 2.1799,
+      "step": 2819
+    },
+    {
+      "epoch": 0.914396887159533,
+      "grad_norm": 0.3631352484226227,
+      "learning_rate": 3e-05,
+      "loss": 2.5371,
+      "step": 2820
+    },
+    {
+      "epoch": 0.9147211413748378,
+      "grad_norm": 0.30766183137893677,
+      "learning_rate": 3e-05,
+      "loss": 2.312,
+      "step": 2821
+    },
+    {
+      "epoch": 0.9150453955901426,
+      "grad_norm": 0.3280878961086273,
+      "learning_rate": 3e-05,
+      "loss": 2.2736,
+      "step": 2822
+    },
+    {
+      "epoch": 0.9153696498054474,
+      "grad_norm": 0.4098678529262543,
+      "learning_rate": 3e-05,
+      "loss": 2.5478,
+      "step": 2823
+    },
+    {
+      "epoch": 0.9156939040207522,
+      "grad_norm": 0.38793930411338806,
+      "learning_rate": 3e-05,
+      "loss": 2.4542,
+      "step": 2824
+    },
+    {
+      "epoch": 0.916018158236057,
+      "grad_norm": 0.48463496565818787,
+      "learning_rate": 3e-05,
+      "loss": 3.3822,
+      "step": 2825
+    },
+    {
+      "epoch": 0.9163424124513618,
+      "grad_norm": 0.34212133288383484,
+      "learning_rate": 3e-05,
+      "loss": 2.7721,
+      "step": 2826
+    },
+    {
+      "epoch": 0.9166666666666666,
+      "grad_norm": 0.5436279773712158,
+      "learning_rate": 3e-05,
+      "loss": 3.2855,
+      "step": 2827
+    },
+    {
+      "epoch": 0.9169909208819714,
+      "grad_norm": 0.41955289244651794,
+      "learning_rate": 3e-05,
+      "loss": 2.2284,
+      "step": 2828
+    },
+    {
+      "epoch": 0.9173151750972762,
+      "grad_norm": 0.4402017295360565,
+      "learning_rate": 3e-05,
+      "loss": 3.0775,
+      "step": 2829
+    },
+    {
+      "epoch": 0.917639429312581,
+      "grad_norm": 0.437509685754776,
+      "learning_rate": 3e-05,
+      "loss": 3.0174,
+      "step": 2830
+    },
+    {
+      "epoch": 0.9179636835278858,
+      "grad_norm": 0.5191596746444702,
+      "learning_rate": 3e-05,
+      "loss": 3.0964,
+      "step": 2831
+    },
+    {
+      "epoch": 0.9182879377431906,
+      "grad_norm": 0.42005661129951477,
+      "learning_rate": 3e-05,
+      "loss": 2.7542,
+      "step": 2832
+    },
+    {
+      "epoch": 0.9186121919584954,
+      "grad_norm": 0.43474313616752625,
+      "learning_rate": 3e-05,
+      "loss": 2.6895,
+      "step": 2833
+    },
+    {
+      "epoch": 0.9189364461738002,
+      "grad_norm": 0.37187710404396057,
+      "learning_rate": 3e-05,
+      "loss": 2.7901,
+      "step": 2834
+    },
+    {
+      "epoch": 0.919260700389105,
+      "grad_norm": 0.4248094856739044,
+      "learning_rate": 3e-05,
+      "loss": 2.8238,
+      "step": 2835
+    },
+    {
+      "epoch": 0.9195849546044098,
+      "grad_norm": 0.47297608852386475,
+      "learning_rate": 3e-05,
+      "loss": 3.0397,
+      "step": 2836
+    },
+    {
+      "epoch": 0.9199092088197146,
+      "grad_norm": 0.4211375415325165,
+      "learning_rate": 3e-05,
+      "loss": 2.7796,
+      "step": 2837
+    },
+    {
+      "epoch": 0.9202334630350194,
+      "grad_norm": 0.3145778477191925,
+      "learning_rate": 3e-05,
+      "loss": 2.4009,
+      "step": 2838
+    },
+    {
+      "epoch": 0.9205577172503242,
+      "grad_norm": 0.30159029364585876,
+      "learning_rate": 3e-05,
+      "loss": 2.0641,
+      "step": 2839
+    },
+    {
+      "epoch": 0.920881971465629,
+      "grad_norm": 0.41733670234680176,
+      "learning_rate": 3e-05,
+      "loss": 2.4793,
+      "step": 2840
+    },
+    {
+      "epoch": 0.9212062256809338,
+      "grad_norm": 0.37574756145477295,
+      "learning_rate": 3e-05,
+      "loss": 2.8451,
+      "step": 2841
+    },
+    {
+      "epoch": 0.9215304798962386,
+      "grad_norm": 0.3890656530857086,
+      "learning_rate": 3e-05,
+      "loss": 2.7522,
+      "step": 2842
+    },
+    {
+      "epoch": 0.9218547341115434,
+      "grad_norm": 0.33728641271591187,
+      "learning_rate": 3e-05,
+      "loss": 2.5546,
+      "step": 2843
+    },
+    {
+      "epoch": 0.9221789883268483,
+      "grad_norm": 0.46142762899398804,
+      "learning_rate": 3e-05,
+      "loss": 2.9692,
+      "step": 2844
+    },
+    {
+      "epoch": 0.9225032425421531,
+      "grad_norm": 0.3038177192211151,
+      "learning_rate": 3e-05,
+      "loss": 2.2815,
+      "step": 2845
+    },
+    {
+      "epoch": 0.9228274967574579,
+      "grad_norm": 0.3376043438911438,
+      "learning_rate": 3e-05,
+      "loss": 2.4997,
+      "step": 2846
+    },
+    {
+      "epoch": 0.9231517509727627,
+      "grad_norm": 0.3840211033821106,
+      "learning_rate": 3e-05,
+      "loss": 3.0438,
+      "step": 2847
+    },
+    {
+      "epoch": 0.9234760051880675,
+      "grad_norm": 0.42310595512390137,
+      "learning_rate": 3e-05,
+      "loss": 3.1037,
+      "step": 2848
+    },
+    {
+      "epoch": 0.9238002594033723,
+      "grad_norm": 0.3024848997592926,
+      "learning_rate": 3e-05,
+      "loss": 2.217,
+      "step": 2849
+    },
+    {
+      "epoch": 0.9241245136186771,
+      "grad_norm": 0.4437383711338043,
+      "learning_rate": 3e-05,
+      "loss": 2.5843,
+      "step": 2850
+    },
+    {
+      "epoch": 0.9244487678339819,
+      "grad_norm": 0.36607006192207336,
+      "learning_rate": 3e-05,
+      "loss": 2.6078,
+      "step": 2851
+    },
+    {
+      "epoch": 0.9247730220492867,
+      "grad_norm": 0.38811057806015015,
+      "learning_rate": 3e-05,
+      "loss": 2.7237,
+      "step": 2852
+    },
+    {
+      "epoch": 0.9250972762645915,
+      "grad_norm": 0.33727291226387024,
+      "learning_rate": 3e-05,
+      "loss": 2.5608,
+      "step": 2853
+    },
+    {
+      "epoch": 0.9254215304798963,
+      "grad_norm": 0.44102567434310913,
+      "learning_rate": 3e-05,
+      "loss": 2.9605,
+      "step": 2854
+    },
+    {
+      "epoch": 0.9257457846952011,
+      "grad_norm": 0.40241432189941406,
+      "learning_rate": 3e-05,
+      "loss": 2.5636,
+      "step": 2855
+    },
+    {
+      "epoch": 0.9260700389105059,
+      "grad_norm": 0.49887654185295105,
+      "learning_rate": 3e-05,
+      "loss": 3.3197,
+      "step": 2856
+    },
+    {
+      "epoch": 0.9263942931258107,
+      "grad_norm": 0.3153311014175415,
+      "learning_rate": 3e-05,
+      "loss": 2.3924,
+      "step": 2857
+    },
+    {
+      "epoch": 0.9267185473411155,
+      "grad_norm": 0.461639940738678,
+      "learning_rate": 3e-05,
+      "loss": 3.4021,
+      "step": 2858
+    },
+    {
+      "epoch": 0.9270428015564203,
+      "grad_norm": 0.38767552375793457,
+      "learning_rate": 3e-05,
+      "loss": 2.3361,
+      "step": 2859
+    },
+    {
+      "epoch": 0.9273670557717251,
+      "grad_norm": 0.5229320526123047,
+      "learning_rate": 3e-05,
+      "loss": 3.0985,
+      "step": 2860
+    },
+    {
+      "epoch": 0.9276913099870299,
+      "grad_norm": 0.3303029537200928,
+      "learning_rate": 3e-05,
+      "loss": 2.1581,
+      "step": 2861
+    },
+    {
+      "epoch": 0.9280155642023347,
+      "grad_norm": 0.3814268410205841,
+      "learning_rate": 3e-05,
+      "loss": 2.4794,
+      "step": 2862
+    },
+    {
+      "epoch": 0.9283398184176395,
+      "grad_norm": 0.32884538173675537,
+      "learning_rate": 3e-05,
+      "loss": 2.3563,
+      "step": 2863
+    },
+    {
+      "epoch": 0.9286640726329443,
+      "grad_norm": 0.36026978492736816,
+      "learning_rate": 3e-05,
+      "loss": 2.3016,
+      "step": 2864
+    },
+    {
+      "epoch": 0.9289883268482491,
+      "grad_norm": 0.3922176659107208,
+      "learning_rate": 3e-05,
+      "loss": 2.6727,
+      "step": 2865
+    },
+    {
+      "epoch": 0.9293125810635539,
+      "grad_norm": 0.33942338824272156,
+      "learning_rate": 3e-05,
+      "loss": 2.4246,
+      "step": 2866
+    },
+    {
+      "epoch": 0.9296368352788587,
+      "grad_norm": 0.601669192314148,
+      "learning_rate": 3e-05,
+      "loss": 2.9023,
+      "step": 2867
+    },
+    {
+      "epoch": 0.9299610894941635,
+      "grad_norm": 0.4000473916530609,
+      "learning_rate": 3e-05,
+      "loss": 2.9927,
+      "step": 2868
+    },
+    {
+      "epoch": 0.9302853437094682,
+      "grad_norm": 0.36253634095191956,
+      "learning_rate": 3e-05,
+      "loss": 2.5474,
+      "step": 2869
+    },
+    {
+      "epoch": 0.930609597924773,
+      "grad_norm": 0.4091234803199768,
+      "learning_rate": 3e-05,
+      "loss": 2.3792,
+      "step": 2870
+    },
+    {
+      "epoch": 0.9309338521400778,
+      "grad_norm": 0.6686627864837646,
+      "learning_rate": 3e-05,
+      "loss": 3.0272,
+      "step": 2871
+    },
+    {
+      "epoch": 0.9312581063553826,
+      "grad_norm": 0.43850553035736084,
+      "learning_rate": 3e-05,
+      "loss": 3.2522,
+      "step": 2872
+    },
+    {
+      "epoch": 0.9315823605706874,
+      "grad_norm": 0.4160626530647278,
+      "learning_rate": 3e-05,
+      "loss": 2.3593,
+      "step": 2873
+    },
+    {
+      "epoch": 0.9319066147859922,
+      "grad_norm": 0.655575692653656,
+      "learning_rate": 3e-05,
+      "loss": 3.3904,
+      "step": 2874
+    },
+    {
+      "epoch": 0.932230869001297,
+      "grad_norm": 0.35255321860313416,
+      "learning_rate": 3e-05,
+      "loss": 2.6329,
+      "step": 2875
+    },
+    {
+      "epoch": 0.9325551232166018,
+      "grad_norm": 0.5397610068321228,
+      "learning_rate": 3e-05,
+      "loss": 3.631,
+      "step": 2876
+    },
+    {
+      "epoch": 0.9328793774319066,
+      "grad_norm": 0.4061860144138336,
+      "learning_rate": 3e-05,
+      "loss": 3.2456,
+      "step": 2877
+    },
+    {
+      "epoch": 0.9332036316472114,
+      "grad_norm": 0.3845311999320984,
+      "learning_rate": 3e-05,
+      "loss": 2.8189,
+      "step": 2878
+    },
+    {
+      "epoch": 0.9335278858625162,
+      "grad_norm": 0.32939016819000244,
+      "learning_rate": 3e-05,
+      "loss": 2.4918,
+      "step": 2879
+    },
+    {
+      "epoch": 0.933852140077821,
+      "grad_norm": 0.3668307363986969,
+      "learning_rate": 3e-05,
+      "loss": 2.6354,
+      "step": 2880
+    },
+    {
+      "epoch": 0.9341763942931258,
+      "grad_norm": 0.5431780815124512,
+      "learning_rate": 3e-05,
+      "loss": 3.0529,
+      "step": 2881
+    },
+    {
+      "epoch": 0.9345006485084306,
+      "grad_norm": 0.32821381092071533,
+      "learning_rate": 3e-05,
+      "loss": 2.6774,
+      "step": 2882
+    },
+    {
+      "epoch": 0.9348249027237354,
+      "grad_norm": 0.6367476582527161,
+      "learning_rate": 3e-05,
+      "loss": 3.3287,
+      "step": 2883
+    },
+    {
+      "epoch": 0.9351491569390402,
+      "grad_norm": 0.40045371651649475,
+      "learning_rate": 3e-05,
+      "loss": 2.7858,
+      "step": 2884
+    },
+    {
+      "epoch": 0.935473411154345,
+      "grad_norm": 0.4027940630912781,
+      "learning_rate": 3e-05,
+      "loss": 2.8803,
+      "step": 2885
+    },
+    {
+      "epoch": 0.9357976653696498,
+      "grad_norm": 0.41554903984069824,
+      "learning_rate": 3e-05,
+      "loss": 2.3828,
+      "step": 2886
+    },
+    {
+      "epoch": 0.9361219195849546,
+      "grad_norm": 0.3484428822994232,
+      "learning_rate": 3e-05,
+      "loss": 2.4929,
+      "step": 2887
+    },
+    {
+      "epoch": 0.9364461738002594,
+      "grad_norm": 0.43447592854499817,
+      "learning_rate": 3e-05,
+      "loss": 3.1609,
+      "step": 2888
+    },
+    {
+      "epoch": 0.9367704280155642,
+      "grad_norm": 0.2844417989253998,
+      "learning_rate": 3e-05,
+      "loss": 2.1166,
+      "step": 2889
+    },
+    {
+      "epoch": 0.937094682230869,
+      "grad_norm": 0.3046829402446747,
+      "learning_rate": 3e-05,
+      "loss": 2.8363,
+      "step": 2890
+    },
+    {
+      "epoch": 0.9374189364461738,
+      "grad_norm": 0.4771912097930908,
+      "learning_rate": 3e-05,
+      "loss": 3.1374,
+      "step": 2891
+    },
+    {
+      "epoch": 0.9377431906614786,
+      "grad_norm": 0.4062415361404419,
+      "learning_rate": 3e-05,
+      "loss": 3.2387,
+      "step": 2892
+    },
+    {
+      "epoch": 0.9380674448767834,
+      "grad_norm": 0.39180830121040344,
+      "learning_rate": 3e-05,
+      "loss": 2.5179,
+      "step": 2893
+    },
+    {
+      "epoch": 0.9383916990920882,
+      "grad_norm": 0.4877490997314453,
+      "learning_rate": 3e-05,
+      "loss": 3.1923,
+      "step": 2894
+    },
+    {
+      "epoch": 0.938715953307393,
+      "grad_norm": 0.370892733335495,
+      "learning_rate": 3e-05,
+      "loss": 2.533,
+      "step": 2895
+    },
+    {
+      "epoch": 0.9390402075226978,
+      "grad_norm": 0.4096488654613495,
+      "learning_rate": 3e-05,
+      "loss": 3.0409,
+      "step": 2896
+    },
+    {
+      "epoch": 0.9393644617380026,
+      "grad_norm": 0.4474868178367615,
+      "learning_rate": 3e-05,
+      "loss": 3.3473,
+      "step": 2897
+    },
+    {
+      "epoch": 0.9396887159533074,
+      "grad_norm": 0.5016494989395142,
+      "learning_rate": 3e-05,
+      "loss": 3.2921,
+      "step": 2898
+    },
+    {
+      "epoch": 0.9400129701686122,
+      "grad_norm": 0.31709110736846924,
+      "learning_rate": 3e-05,
+      "loss": 2.2451,
+      "step": 2899
+    },
+    {
+      "epoch": 0.940337224383917,
+      "grad_norm": 0.3947117030620575,
+      "learning_rate": 3e-05,
+      "loss": 3.2496,
+      "step": 2900
+    },
+    {
+      "epoch": 0.9406614785992218,
+      "grad_norm": 0.607421338558197,
+      "learning_rate": 3e-05,
+      "loss": 3.1687,
+      "step": 2901
+    },
+    {
+      "epoch": 0.9409857328145266,
+      "grad_norm": 0.3988184630870819,
+      "learning_rate": 3e-05,
+      "loss": 2.8246,
+      "step": 2902
+    },
+    {
+      "epoch": 0.9413099870298314,
+      "grad_norm": 0.444690465927124,
+      "learning_rate": 3e-05,
+      "loss": 3.1113,
+      "step": 2903
+    },
+    {
+      "epoch": 0.9416342412451362,
+      "grad_norm": 0.5266732573509216,
+      "learning_rate": 3e-05,
+      "loss": 2.7119,
+      "step": 2904
+    },
+    {
+      "epoch": 0.941958495460441,
+      "grad_norm": 0.34128838777542114,
+      "learning_rate": 3e-05,
+      "loss": 2.3987,
+      "step": 2905
+    },
+    {
+      "epoch": 0.9422827496757458,
+      "grad_norm": 0.35372960567474365,
+      "learning_rate": 3e-05,
+      "loss": 2.3821,
+      "step": 2906
+    },
+    {
+      "epoch": 0.9426070038910506,
+      "grad_norm": 0.35397449135780334,
+      "learning_rate": 3e-05,
+      "loss": 2.4558,
+      "step": 2907
+    },
+    {
+      "epoch": 0.9429312581063554,
+      "grad_norm": 0.36683642864227295,
+      "learning_rate": 3e-05,
+      "loss": 2.458,
+      "step": 2908
+    },
+    {
+      "epoch": 0.9432555123216602,
+      "grad_norm": 0.4041233956813812,
+      "learning_rate": 3e-05,
+      "loss": 2.5869,
+      "step": 2909
+    },
+    {
+      "epoch": 0.943579766536965,
+      "grad_norm": 0.30623775720596313,
+      "learning_rate": 3e-05,
+      "loss": 2.3112,
+      "step": 2910
+    },
+    {
+      "epoch": 0.9439040207522698,
+      "grad_norm": 0.6615625619888306,
+      "learning_rate": 3e-05,
+      "loss": 3.3786,
+      "step": 2911
+    },
+    {
+      "epoch": 0.9442282749675746,
+      "grad_norm": 0.32024645805358887,
+      "learning_rate": 3e-05,
+      "loss": 2.456,
+      "step": 2912
+    },
+    {
+      "epoch": 0.9445525291828794,
+      "grad_norm": 0.3190930187702179,
+      "learning_rate": 3e-05,
+      "loss": 2.3919,
+      "step": 2913
+    },
+    {
+      "epoch": 0.9448767833981842,
+      "grad_norm": 0.3725919723510742,
+      "learning_rate": 3e-05,
+      "loss": 2.6623,
+      "step": 2914
+    },
+    {
+      "epoch": 0.945201037613489,
+      "grad_norm": 0.3565352261066437,
+      "learning_rate": 3e-05,
+      "loss": 2.4355,
+      "step": 2915
+    },
+    {
+      "epoch": 0.9455252918287937,
+      "grad_norm": 0.35213303565979004,
+      "learning_rate": 3e-05,
+      "loss": 2.5236,
+      "step": 2916
+    },
+    {
+      "epoch": 0.9458495460440985,
+      "grad_norm": 0.34089532494544983,
+      "learning_rate": 3e-05,
+      "loss": 2.2947,
+      "step": 2917
+    },
+    {
+      "epoch": 0.9461738002594033,
+      "grad_norm": 0.44071412086486816,
+      "learning_rate": 3e-05,
+      "loss": 2.6947,
+      "step": 2918
+    },
+    {
+      "epoch": 0.9464980544747081,
+      "grad_norm": 0.46328720450401306,
+      "learning_rate": 3e-05,
+      "loss": 2.879,
+      "step": 2919
+    },
+    {
+      "epoch": 0.9468223086900129,
+      "grad_norm": 0.3349205553531647,
+      "learning_rate": 3e-05,
+      "loss": 2.4967,
+      "step": 2920
+    },
+    {
+      "epoch": 0.9471465629053177,
+      "grad_norm": 0.7264132499694824,
+      "learning_rate": 3e-05,
+      "loss": 3.5168,
+      "step": 2921
+    },
+    {
+      "epoch": 0.9474708171206225,
+      "grad_norm": 0.5175856947898865,
+      "learning_rate": 3e-05,
+      "loss": 3.1204,
+      "step": 2922
+    },
+    {
+      "epoch": 0.9477950713359273,
+      "grad_norm": 0.4160337448120117,
+      "learning_rate": 3e-05,
+      "loss": 2.8379,
+      "step": 2923
+    },
+    {
+      "epoch": 0.9481193255512321,
+      "grad_norm": 0.3702349364757538,
+      "learning_rate": 3e-05,
+      "loss": 2.7029,
+      "step": 2924
+    },
+    {
+      "epoch": 0.9484435797665369,
+      "grad_norm": 0.4225955605506897,
+      "learning_rate": 3e-05,
+      "loss": 3.1627,
+      "step": 2925
+    },
+    {
+      "epoch": 0.9487678339818417,
+      "grad_norm": 0.4809791147708893,
+      "learning_rate": 3e-05,
+      "loss": 2.7839,
+      "step": 2926
+    },
+    {
+      "epoch": 0.9490920881971465,
+      "grad_norm": 0.4431580603122711,
+      "learning_rate": 3e-05,
+      "loss": 2.9282,
+      "step": 2927
+    },
+    {
+      "epoch": 0.9494163424124513,
+      "grad_norm": 0.3395250737667084,
+      "learning_rate": 3e-05,
+      "loss": 2.5556,
+      "step": 2928
+    },
+    {
+      "epoch": 0.9497405966277561,
+      "grad_norm": 0.276656836271286,
+      "learning_rate": 3e-05,
+      "loss": 2.1791,
+      "step": 2929
+    },
+    {
+      "epoch": 0.9500648508430609,
+      "grad_norm": 0.48410749435424805,
+      "learning_rate": 3e-05,
+      "loss": 2.7702,
+      "step": 2930
+    },
+    {
+      "epoch": 0.9503891050583657,
+      "grad_norm": 0.3885018229484558,
+      "learning_rate": 3e-05,
+      "loss": 2.617,
+      "step": 2931
+    },
+    {
+      "epoch": 0.9507133592736705,
+      "grad_norm": 0.43259403109550476,
+      "learning_rate": 3e-05,
+      "loss": 2.9791,
+      "step": 2932
+    },
+    {
+      "epoch": 0.9510376134889753,
+      "grad_norm": 0.3511403203010559,
+      "learning_rate": 3e-05,
+      "loss": 2.2164,
+      "step": 2933
+    },
+    {
+      "epoch": 0.9513618677042801,
+      "grad_norm": 0.5799418091773987,
+      "learning_rate": 3e-05,
+      "loss": 3.6058,
+      "step": 2934
+    },
+    {
+      "epoch": 0.9516861219195849,
+      "grad_norm": 0.42391878366470337,
+      "learning_rate": 3e-05,
+      "loss": 3.1324,
+      "step": 2935
+    },
+    {
+      "epoch": 0.9520103761348897,
+      "grad_norm": 0.3517759144306183,
+      "learning_rate": 3e-05,
+      "loss": 3.0195,
+      "step": 2936
+    },
+    {
+      "epoch": 0.9523346303501945,
+      "grad_norm": 0.4260661005973816,
+      "learning_rate": 3e-05,
+      "loss": 2.5405,
+      "step": 2937
+    },
+    {
+      "epoch": 0.9526588845654993,
+      "grad_norm": 0.5702102184295654,
+      "learning_rate": 3e-05,
+      "loss": 2.8695,
+      "step": 2938
+    },
+    {
+      "epoch": 0.9529831387808041,
+      "grad_norm": 0.36880719661712646,
+      "learning_rate": 3e-05,
+      "loss": 2.4278,
+      "step": 2939
+    },
+    {
+      "epoch": 0.953307392996109,
+      "grad_norm": 0.4686841368675232,
+      "learning_rate": 3e-05,
+      "loss": 2.8248,
+      "step": 2940
+    },
+    {
+      "epoch": 0.9536316472114138,
+      "grad_norm": 0.36421799659729004,
+      "learning_rate": 3e-05,
+      "loss": 2.3672,
+      "step": 2941
+    },
+    {
+      "epoch": 0.9539559014267186,
+      "grad_norm": 0.35624539852142334,
+      "learning_rate": 3e-05,
+      "loss": 2.4061,
+      "step": 2942
+    },
+    {
+      "epoch": 0.9542801556420234,
+      "grad_norm": 0.3326251804828644,
+      "learning_rate": 3e-05,
+      "loss": 2.3073,
+      "step": 2943
+    },
+    {
+      "epoch": 0.9546044098573282,
+      "grad_norm": 0.3773517906665802,
+      "learning_rate": 3e-05,
+      "loss": 2.4007,
+      "step": 2944
+    },
+    {
+      "epoch": 0.954928664072633,
+      "grad_norm": 0.4222182333469391,
+      "learning_rate": 3e-05,
+      "loss": 3.0417,
+      "step": 2945
+    },
+    {
+      "epoch": 0.9552529182879378,
+      "grad_norm": 0.4778387248516083,
+      "learning_rate": 3e-05,
+      "loss": 2.668,
+      "step": 2946
+    },
+    {
+      "epoch": 0.9555771725032426,
+      "grad_norm": 0.2907552719116211,
+      "learning_rate": 3e-05,
+      "loss": 2.3925,
+      "step": 2947
+    },
+    {
+      "epoch": 0.9559014267185474,
+      "grad_norm": 0.32670024037361145,
+      "learning_rate": 3e-05,
+      "loss": 2.3908,
+      "step": 2948
+    },
+    {
+      "epoch": 0.9562256809338522,
+      "grad_norm": 0.3718966245651245,
+      "learning_rate": 3e-05,
+      "loss": 2.6949,
+      "step": 2949
+    },
+    {
+      "epoch": 0.956549935149157,
+      "grad_norm": 0.3269632160663605,
+      "learning_rate": 3e-05,
+      "loss": 2.6948,
+      "step": 2950
+    },
+    {
+      "epoch": 0.9568741893644618,
+      "grad_norm": 0.38018998503685,
+      "learning_rate": 3e-05,
+      "loss": 2.5668,
+      "step": 2951
+    },
+    {
+      "epoch": 0.9571984435797666,
+      "grad_norm": 0.3321263790130615,
+      "learning_rate": 3e-05,
+      "loss": 2.7075,
+      "step": 2952
+    },
+    {
+      "epoch": 0.9575226977950714,
+      "grad_norm": 0.4001438319683075,
+      "learning_rate": 3e-05,
+      "loss": 2.9872,
+      "step": 2953
+    },
+    {
+      "epoch": 0.9578469520103762,
+      "grad_norm": 0.31868237257003784,
+      "learning_rate": 3e-05,
+      "loss": 2.4996,
+      "step": 2954
+    },
+    {
+      "epoch": 0.958171206225681,
+      "grad_norm": 0.6355832815170288,
+      "learning_rate": 3e-05,
+      "loss": 3.1435,
+      "step": 2955
+    },
+    {
+      "epoch": 0.9584954604409858,
+      "grad_norm": 0.7720221281051636,
+      "learning_rate": 3e-05,
+      "loss": 4.0633,
+      "step": 2956
+    },
+    {
+      "epoch": 0.9588197146562906,
+      "grad_norm": 0.31086230278015137,
+      "learning_rate": 3e-05,
+      "loss": 2.0412,
+      "step": 2957
+    },
+    {
+      "epoch": 0.9591439688715954,
+      "grad_norm": 0.42139047384262085,
+      "learning_rate": 3e-05,
+      "loss": 2.9872,
+      "step": 2958
+    },
+    {
+      "epoch": 0.9594682230869002,
+      "grad_norm": 0.40515539050102234,
+      "learning_rate": 3e-05,
+      "loss": 2.6697,
+      "step": 2959
+    },
+    {
+      "epoch": 0.959792477302205,
+      "grad_norm": 0.5258477926254272,
+      "learning_rate": 3e-05,
+      "loss": 3.1925,
+      "step": 2960
+    },
+    {
+      "epoch": 0.9601167315175098,
+      "grad_norm": 0.3948691487312317,
+      "learning_rate": 3e-05,
+      "loss": 2.6568,
+      "step": 2961
+    },
+    {
+      "epoch": 0.9604409857328146,
+      "grad_norm": 0.3589661717414856,
+      "learning_rate": 3e-05,
+      "loss": 2.5608,
+      "step": 2962
+    },
+    {
+      "epoch": 0.9607652399481194,
+      "grad_norm": 0.4283868670463562,
+      "learning_rate": 3e-05,
+      "loss": 2.7048,
+      "step": 2963
+    },
+    {
+      "epoch": 0.9610894941634242,
+      "grad_norm": 0.3407347500324249,
+      "learning_rate": 3e-05,
+      "loss": 2.4072,
+      "step": 2964
+    },
+    {
+      "epoch": 0.961413748378729,
+      "grad_norm": 0.41145041584968567,
+      "learning_rate": 3e-05,
+      "loss": 3.1117,
+      "step": 2965
+    },
+    {
+      "epoch": 0.9617380025940337,
+      "grad_norm": 0.49370816349983215,
+      "learning_rate": 3e-05,
+      "loss": 2.344,
+      "step": 2966
+    },
+    {
+      "epoch": 0.9620622568093385,
+      "grad_norm": 0.4059050381183624,
+      "learning_rate": 3e-05,
+      "loss": 2.7551,
+      "step": 2967
+    },
+    {
+      "epoch": 0.9623865110246433,
+      "grad_norm": 0.29326826333999634,
+      "learning_rate": 3e-05,
+      "loss": 2.2966,
+      "step": 2968
+    },
+    {
+      "epoch": 0.9627107652399481,
+      "grad_norm": 0.29283690452575684,
+      "learning_rate": 3e-05,
+      "loss": 2.089,
+      "step": 2969
+    },
+    {
+      "epoch": 0.9630350194552529,
+      "grad_norm": 0.35291367769241333,
+      "learning_rate": 3e-05,
+      "loss": 2.4938,
+      "step": 2970
+    },
+    {
+      "epoch": 0.9633592736705577,
+      "grad_norm": 0.33223074674606323,
+      "learning_rate": 3e-05,
+      "loss": 2.2558,
+      "step": 2971
+    },
+    {
+      "epoch": 0.9636835278858625,
+      "grad_norm": 0.35592639446258545,
+      "learning_rate": 3e-05,
+      "loss": 2.434,
+      "step": 2972
+    },
+    {
+      "epoch": 0.9640077821011673,
+      "grad_norm": 0.4056759178638458,
+      "learning_rate": 3e-05,
+      "loss": 2.6567,
+      "step": 2973
+    },
+    {
+      "epoch": 0.9643320363164721,
+      "grad_norm": 0.36522340774536133,
+      "learning_rate": 3e-05,
+      "loss": 2.9359,
+      "step": 2974
+    },
+    {
+      "epoch": 0.9646562905317769,
+      "grad_norm": 0.4097868800163269,
+      "learning_rate": 3e-05,
+      "loss": 2.8892,
+      "step": 2975
+    },
+    {
+      "epoch": 0.9649805447470817,
+      "grad_norm": 0.3220478296279907,
+      "learning_rate": 3e-05,
+      "loss": 2.2818,
+      "step": 2976
+    },
+    {
+      "epoch": 0.9653047989623865,
+      "grad_norm": 0.3704823851585388,
+      "learning_rate": 3e-05,
+      "loss": 3.421,
+      "step": 2977
+    },
+    {
+      "epoch": 0.9656290531776913,
+      "grad_norm": 0.40346559882164,
+      "learning_rate": 3e-05,
+      "loss": 2.5447,
+      "step": 2978
+    },
+    {
+      "epoch": 0.9659533073929961,
+      "grad_norm": 0.31971126794815063,
+      "learning_rate": 3e-05,
+      "loss": 2.3762,
+      "step": 2979
+    },
+    {
+      "epoch": 0.9662775616083009,
+      "grad_norm": 0.5706731677055359,
+      "learning_rate": 3e-05,
+      "loss": 3.3567,
+      "step": 2980
+    },
+    {
+      "epoch": 0.9666018158236057,
+      "grad_norm": 0.40904298424720764,
+      "learning_rate": 3e-05,
+      "loss": 2.7104,
+      "step": 2981
+    },
+    {
+      "epoch": 0.9669260700389105,
+      "grad_norm": 0.4123535454273224,
+      "learning_rate": 3e-05,
+      "loss": 2.2585,
+      "step": 2982
+    },
+    {
+      "epoch": 0.9672503242542153,
+      "grad_norm": 0.2956462800502777,
+      "learning_rate": 3e-05,
+      "loss": 2.2163,
+      "step": 2983
+    },
+    {
+      "epoch": 0.9675745784695201,
+      "grad_norm": 0.6119877099990845,
+      "learning_rate": 3e-05,
+      "loss": 3.6485,
+      "step": 2984
+    },
+    {
+      "epoch": 0.9678988326848249,
+      "grad_norm": 0.3283049166202545,
+      "learning_rate": 3e-05,
+      "loss": 2.1785,
+      "step": 2985
+    },
+    {
+      "epoch": 0.9682230869001297,
+      "grad_norm": 0.4398946762084961,
+      "learning_rate": 3e-05,
+      "loss": 3.214,
+      "step": 2986
+    },
+    {
+      "epoch": 0.9685473411154345,
+      "grad_norm": 0.359115868806839,
+      "learning_rate": 3e-05,
+      "loss": 2.8159,
+      "step": 2987
+    },
+    {
+      "epoch": 0.9688715953307393,
+      "grad_norm": 0.4592425227165222,
+      "learning_rate": 3e-05,
+      "loss": 2.7311,
+      "step": 2988
+    },
+    {
+      "epoch": 0.9691958495460441,
+      "grad_norm": 0.3868071436882019,
+      "learning_rate": 3e-05,
+      "loss": 2.6691,
+      "step": 2989
+    },
+    {
+      "epoch": 0.9695201037613489,
+      "grad_norm": 0.31837454438209534,
+      "learning_rate": 3e-05,
+      "loss": 2.2002,
+      "step": 2990
+    },
+    {
+      "epoch": 0.9698443579766537,
+      "grad_norm": 0.4580914378166199,
+      "learning_rate": 3e-05,
+      "loss": 2.8408,
+      "step": 2991
+    },
+    {
+      "epoch": 0.9701686121919585,
+      "grad_norm": 0.4604332745075226,
+      "learning_rate": 3e-05,
+      "loss": 3.2537,
+      "step": 2992
+    },
+    {
+      "epoch": 0.9704928664072633,
+      "grad_norm": 0.3047242760658264,
+      "learning_rate": 3e-05,
+      "loss": 2.0643,
+      "step": 2993
+    },
+    {
+      "epoch": 0.9708171206225681,
+      "grad_norm": 0.3484688699245453,
+      "learning_rate": 3e-05,
+      "loss": 2.5703,
+      "step": 2994
+    },
+    {
+      "epoch": 0.9711413748378729,
+      "grad_norm": 0.3195679485797882,
+      "learning_rate": 3e-05,
+      "loss": 1.9665,
+      "step": 2995
+    },
+    {
+      "epoch": 0.9714656290531777,
+      "grad_norm": 0.3346187472343445,
+      "learning_rate": 3e-05,
+      "loss": 2.8131,
+      "step": 2996
+    },
+    {
+      "epoch": 0.9717898832684825,
+      "grad_norm": 0.31272435188293457,
+      "learning_rate": 3e-05,
+      "loss": 2.5663,
+      "step": 2997
+    },
+    {
+      "epoch": 0.9721141374837873,
+      "grad_norm": 0.29805463552474976,
+      "learning_rate": 3e-05,
+      "loss": 2.3538,
+      "step": 2998
+    },
+    {
+      "epoch": 0.9724383916990921,
+      "grad_norm": 0.3429343104362488,
+      "learning_rate": 3e-05,
+      "loss": 2.2713,
+      "step": 2999
+    },
+    {
+      "epoch": 0.9727626459143969,
+      "grad_norm": 0.3730374574661255,
+      "learning_rate": 3e-05,
+      "loss": 2.768,
+      "step": 3000
+    },
+    {
+      "epoch": 0.9730869001297017,
+      "grad_norm": 0.35898303985595703,
+      "learning_rate": 3e-05,
+      "loss": 2.947,
+      "step": 3001
+    },
+    {
+      "epoch": 0.9734111543450065,
+      "grad_norm": 0.3608832061290741,
+      "learning_rate": 3e-05,
+      "loss": 2.9449,
+      "step": 3002
+    },
+    {
+      "epoch": 0.9737354085603113,
+      "grad_norm": 0.2949283719062805,
+      "learning_rate": 3e-05,
+      "loss": 2.226,
+      "step": 3003
+    },
+    {
+      "epoch": 0.9740596627756161,
+      "grad_norm": 0.2674799859523773,
+      "learning_rate": 3e-05,
+      "loss": 2.277,
+      "step": 3004
+    },
+    {
+      "epoch": 0.9743839169909209,
+      "grad_norm": 0.3855573832988739,
+      "learning_rate": 3e-05,
+      "loss": 2.6944,
+      "step": 3005
+    },
+    {
+      "epoch": 0.9747081712062257,
+      "grad_norm": 0.5746701955795288,
+      "learning_rate": 3e-05,
+      "loss": 3.6605,
+      "step": 3006
+    },
+    {
+      "epoch": 0.9750324254215305,
+      "grad_norm": 0.39888954162597656,
+      "learning_rate": 3e-05,
+      "loss": 2.7456,
+      "step": 3007
+    },
+    {
+      "epoch": 0.9753566796368353,
+      "grad_norm": 0.4286193251609802,
+      "learning_rate": 3e-05,
+      "loss": 3.2783,
+      "step": 3008
+    },
+    {
+      "epoch": 0.97568093385214,
+      "grad_norm": 0.3424946665763855,
+      "learning_rate": 3e-05,
+      "loss": 2.4886,
+      "step": 3009
+    },
+    {
+      "epoch": 0.9760051880674449,
+      "grad_norm": 0.2981758117675781,
+      "learning_rate": 3e-05,
+      "loss": 2.2779,
+      "step": 3010
+    },
+    {
+      "epoch": 0.9763294422827496,
+      "grad_norm": 0.5266610383987427,
+      "learning_rate": 3e-05,
+      "loss": 3.1423,
+      "step": 3011
+    },
+    {
+      "epoch": 0.9766536964980544,
+      "grad_norm": 0.29751449823379517,
+      "learning_rate": 3e-05,
+      "loss": 2.3128,
+      "step": 3012
+    },
+    {
+      "epoch": 0.9769779507133592,
+      "grad_norm": 0.37801334261894226,
+      "learning_rate": 3e-05,
+      "loss": 2.4549,
+      "step": 3013
+    },
+    {
+      "epoch": 0.977302204928664,
+      "grad_norm": 0.38805049657821655,
+      "learning_rate": 3e-05,
+      "loss": 2.8313,
+      "step": 3014
+    },
+    {
+      "epoch": 0.9776264591439688,
+      "grad_norm": 0.4093324840068817,
+      "learning_rate": 3e-05,
+      "loss": 3.0808,
+      "step": 3015
+    },
+    {
+      "epoch": 0.9779507133592736,
+      "grad_norm": 0.3746676445007324,
+      "learning_rate": 3e-05,
+      "loss": 2.4913,
+      "step": 3016
+    },
+    {
+      "epoch": 0.9782749675745784,
+      "grad_norm": 0.32716798782348633,
+      "learning_rate": 3e-05,
+      "loss": 2.4796,
+      "step": 3017
+    },
+    {
+      "epoch": 0.9785992217898832,
+      "grad_norm": 0.3745894730091095,
+      "learning_rate": 3e-05,
+      "loss": 2.4765,
+      "step": 3018
+    },
+    {
+      "epoch": 0.978923476005188,
+      "grad_norm": 0.3861525356769562,
+      "learning_rate": 3e-05,
+      "loss": 2.6818,
+      "step": 3019
+    },
+    {
+      "epoch": 0.9792477302204928,
+      "grad_norm": 0.46795254945755005,
+      "learning_rate": 3e-05,
+      "loss": 2.7739,
+      "step": 3020
+    },
+    {
+      "epoch": 0.9795719844357976,
+      "grad_norm": 0.2721799314022064,
+      "learning_rate": 3e-05,
+      "loss": 2.2176,
+      "step": 3021
+    },
+    {
+      "epoch": 0.9798962386511024,
+      "grad_norm": 0.35149669647216797,
+      "learning_rate": 3e-05,
+      "loss": 2.9303,
+      "step": 3022
+    },
+    {
+      "epoch": 0.9802204928664072,
+      "grad_norm": 0.31177258491516113,
+      "learning_rate": 3e-05,
+      "loss": 2.4579,
+      "step": 3023
+    },
+    {
+      "epoch": 0.980544747081712,
+      "grad_norm": 0.381080687046051,
+      "learning_rate": 3e-05,
+      "loss": 3.0968,
+      "step": 3024
+    },
+    {
+      "epoch": 0.9808690012970168,
+      "grad_norm": 0.3404100239276886,
+      "learning_rate": 3e-05,
+      "loss": 3.0314,
+      "step": 3025
+    },
+    {
+      "epoch": 0.9811932555123216,
+      "grad_norm": 0.41072750091552734,
+      "learning_rate": 3e-05,
+      "loss": 3.1186,
+      "step": 3026
+    },
+    {
+      "epoch": 0.9815175097276264,
+      "grad_norm": 0.377362996339798,
+      "learning_rate": 3e-05,
+      "loss": 2.4089,
+      "step": 3027
+    },
+    {
+      "epoch": 0.9818417639429312,
+      "grad_norm": 0.426811158657074,
+      "learning_rate": 3e-05,
+      "loss": 2.4182,
+      "step": 3028
+    },
+    {
+      "epoch": 0.982166018158236,
+      "grad_norm": 0.4739135503768921,
+      "learning_rate": 3e-05,
+      "loss": 3.1774,
+      "step": 3029
+    },
+    {
+      "epoch": 0.9824902723735408,
+      "grad_norm": 0.38212889432907104,
+      "learning_rate": 3e-05,
+      "loss": 3.1578,
+      "step": 3030
+    },
+    {
+      "epoch": 0.9828145265888456,
+      "grad_norm": 0.34025701880455017,
+      "learning_rate": 3e-05,
+      "loss": 2.5313,
+      "step": 3031
+    },
+    {
+      "epoch": 0.9831387808041504,
+      "grad_norm": 0.4236637353897095,
+      "learning_rate": 3e-05,
+      "loss": 2.5958,
+      "step": 3032
+    },
+    {
+      "epoch": 0.9834630350194552,
+      "grad_norm": 0.722714900970459,
+      "learning_rate": 3e-05,
+      "loss": 2.7469,
+      "step": 3033
+    },
+    {
+      "epoch": 0.98378728923476,
+      "grad_norm": 0.40366557240486145,
+      "learning_rate": 3e-05,
+      "loss": 2.8282,
+      "step": 3034
+    },
+    {
+      "epoch": 0.9841115434500648,
+      "grad_norm": 0.33851152658462524,
+      "learning_rate": 3e-05,
+      "loss": 2.8587,
+      "step": 3035
+    },
+    {
+      "epoch": 0.9844357976653697,
+      "grad_norm": 0.3614557087421417,
+      "learning_rate": 3e-05,
+      "loss": 2.3566,
+      "step": 3036
+    },
+    {
+      "epoch": 0.9847600518806745,
+      "grad_norm": 0.37364521622657776,
+      "learning_rate": 3e-05,
+      "loss": 2.5768,
+      "step": 3037
+    },
+    {
+      "epoch": 0.9850843060959793,
+      "grad_norm": 0.4772968590259552,
+      "learning_rate": 3e-05,
+      "loss": 2.919,
+      "step": 3038
+    },
+    {
+      "epoch": 0.9854085603112841,
+      "grad_norm": 0.2845996618270874,
+      "learning_rate": 3e-05,
+      "loss": 2.1128,
+      "step": 3039
+    },
+    {
+      "epoch": 0.9857328145265889,
+      "grad_norm": 0.4010814130306244,
+      "learning_rate": 3e-05,
+      "loss": 2.9512,
+      "step": 3040
+    },
+    {
+      "epoch": 0.9860570687418937,
+      "grad_norm": 0.4424131512641907,
+      "learning_rate": 3e-05,
+      "loss": 3.118,
+      "step": 3041
+    },
+    {
+      "epoch": 0.9863813229571985,
+      "grad_norm": 0.30934596061706543,
+      "learning_rate": 3e-05,
+      "loss": 2.1939,
+      "step": 3042
+    },
+    {
+      "epoch": 0.9867055771725033,
+      "grad_norm": 0.31177762150764465,
+      "learning_rate": 3e-05,
+      "loss": 2.3027,
+      "step": 3043
+    },
+    {
+      "epoch": 0.9870298313878081,
+      "grad_norm": 0.5044785141944885,
+      "learning_rate": 3e-05,
+      "loss": 2.9006,
+      "step": 3044
+    },
+    {
+      "epoch": 0.9873540856031129,
+      "grad_norm": 0.3778097331523895,
+      "learning_rate": 3e-05,
+      "loss": 2.5524,
+      "step": 3045
+    },
+    {
+      "epoch": 0.9876783398184177,
+      "grad_norm": 0.41896307468414307,
+      "learning_rate": 3e-05,
+      "loss": 3.0153,
+      "step": 3046
+    },
+    {
+      "epoch": 0.9880025940337225,
+      "grad_norm": 0.33241692185401917,
+      "learning_rate": 3e-05,
+      "loss": 2.5826,
+      "step": 3047
+    },
+    {
+      "epoch": 0.9883268482490273,
+      "grad_norm": 0.527088463306427,
+      "learning_rate": 3e-05,
+      "loss": 2.8947,
+      "step": 3048
+    },
+    {
+      "epoch": 0.9886511024643321,
+      "grad_norm": 0.3060944080352783,
+      "learning_rate": 3e-05,
+      "loss": 2.0872,
+      "step": 3049
+    },
+    {
+      "epoch": 0.9889753566796369,
+      "grad_norm": 0.45200052857398987,
+      "learning_rate": 3e-05,
+      "loss": 2.5816,
+      "step": 3050
+    },
+    {
+      "epoch": 0.9892996108949417,
+      "grad_norm": 0.32031768560409546,
+      "learning_rate": 3e-05,
+      "loss": 2.2273,
+      "step": 3051
+    },
+    {
+      "epoch": 0.9896238651102465,
+      "grad_norm": 0.5075026154518127,
+      "learning_rate": 3e-05,
+      "loss": 2.4648,
+      "step": 3052
+    },
+    {
+      "epoch": 0.9899481193255513,
+      "grad_norm": 0.4127306342124939,
+      "learning_rate": 3e-05,
+      "loss": 2.9744,
+      "step": 3053
+    },
+    {
+      "epoch": 0.9902723735408561,
+      "grad_norm": 0.33077454566955566,
+      "learning_rate": 3e-05,
+      "loss": 2.337,
+      "step": 3054
+    },
+    {
+      "epoch": 0.9905966277561609,
+      "grad_norm": 0.34755364060401917,
+      "learning_rate": 3e-05,
+      "loss": 2.3944,
+      "step": 3055
+    },
+    {
+      "epoch": 0.9909208819714657,
+      "grad_norm": 0.3651041090488434,
+      "learning_rate": 3e-05,
+      "loss": 2.6355,
+      "step": 3056
+    },
+    {
+      "epoch": 0.9912451361867705,
+      "grad_norm": 0.36791667342185974,
+      "learning_rate": 3e-05,
+      "loss": 2.5197,
+      "step": 3057
+    },
+    {
+      "epoch": 0.9915693904020753,
+      "grad_norm": 0.4207691550254822,
+      "learning_rate": 3e-05,
+      "loss": 2.9799,
+      "step": 3058
+    },
+    {
+      "epoch": 0.99189364461738,
+      "grad_norm": 0.32029181718826294,
+      "learning_rate": 3e-05,
+      "loss": 2.4367,
+      "step": 3059
+    },
+    {
+      "epoch": 0.9922178988326849,
+      "grad_norm": 0.3381041884422302,
+      "learning_rate": 3e-05,
+      "loss": 2.3123,
+      "step": 3060
+    },
+    {
+      "epoch": 0.9925421530479897,
+      "grad_norm": 0.3514271676540375,
+      "learning_rate": 3e-05,
+      "loss": 2.5502,
+      "step": 3061
+    },
+    {
+      "epoch": 0.9928664072632944,
+      "grad_norm": 0.5126301050186157,
+      "learning_rate": 3e-05,
+      "loss": 3.4218,
+      "step": 3062
+    },
+    {
+      "epoch": 0.9931906614785992,
+      "grad_norm": 0.4171275794506073,
+      "learning_rate": 3e-05,
+      "loss": 2.6083,
+      "step": 3063
+    },
+    {
+      "epoch": 0.993514915693904,
+      "grad_norm": 0.40363672375679016,
+      "learning_rate": 3e-05,
+      "loss": 3.1645,
+      "step": 3064
+    },
+    {
+      "epoch": 0.9938391699092088,
+      "grad_norm": 0.3384639322757721,
+      "learning_rate": 3e-05,
+      "loss": 2.6959,
+      "step": 3065
+    },
+    {
+      "epoch": 0.9941634241245136,
+      "grad_norm": 0.28195056319236755,
+      "learning_rate": 3e-05,
+      "loss": 2.173,
+      "step": 3066
+    },
+    {
+      "epoch": 0.9944876783398184,
+      "grad_norm": 0.3255248963832855,
+      "learning_rate": 3e-05,
+      "loss": 2.4768,
+      "step": 3067
+    },
+    {
+      "epoch": 0.9948119325551232,
+      "grad_norm": 0.3250022530555725,
+      "learning_rate": 3e-05,
+      "loss": 2.3102,
+      "step": 3068
+    },
+    {
+      "epoch": 0.995136186770428,
+      "grad_norm": 0.40980687737464905,
+      "learning_rate": 3e-05,
+      "loss": 3.099,
+      "step": 3069
+    },
+    {
+      "epoch": 0.9954604409857328,
+      "grad_norm": 0.5355203747749329,
+      "learning_rate": 3e-05,
+      "loss": 3.159,
+      "step": 3070
+    },
+    {
+      "epoch": 0.9957846952010376,
+      "grad_norm": 0.486395001411438,
+      "learning_rate": 3e-05,
+      "loss": 3.1025,
+      "step": 3071
+    },
+    {
+      "epoch": 0.9961089494163424,
+      "grad_norm": 0.46148258447647095,
+      "learning_rate": 3e-05,
+      "loss": 2.4056,
+      "step": 3072
+    },
+    {
+      "epoch": 0.9964332036316472,
+      "grad_norm": 0.4108019769191742,
+      "learning_rate": 3e-05,
+      "loss": 2.7107,
+      "step": 3073
+    },
+    {
+      "epoch": 0.996757457846952,
+      "grad_norm": 0.4269393980503082,
+      "learning_rate": 3e-05,
+      "loss": 2.8067,
+      "step": 3074
+    },
+    {
+      "epoch": 0.9970817120622568,
+      "grad_norm": 0.3614785671234131,
+      "learning_rate": 3e-05,
+      "loss": 3.1879,
+      "step": 3075
+    },
+    {
+      "epoch": 0.9974059662775616,
+      "grad_norm": 0.48739635944366455,
+      "learning_rate": 3e-05,
+      "loss": 3.1212,
+      "step": 3076
+    },
+    {
+      "epoch": 0.9977302204928664,
+      "grad_norm": 0.418366014957428,
+      "learning_rate": 3e-05,
+      "loss": 2.6963,
+      "step": 3077
+    },
+    {
+      "epoch": 0.9980544747081712,
+      "grad_norm": 0.4406101405620575,
+      "learning_rate": 3e-05,
+      "loss": 2.7685,
+      "step": 3078
+    },
+    {
+      "epoch": 0.998378728923476,
+      "grad_norm": 0.5660655498504639,
+      "learning_rate": 3e-05,
+      "loss": 2.8023,
+      "step": 3079
+    },
+    {
+      "epoch": 0.9987029831387808,
+      "grad_norm": 0.31638890504837036,
+      "learning_rate": 3e-05,
+      "loss": 2.1777,
+      "step": 3080
+    },
+    {
+      "epoch": 0.9990272373540856,
+      "grad_norm": 0.3518223464488983,
+      "learning_rate": 3e-05,
+      "loss": 2.429,
+      "step": 3081
+    },
+    {
+      "epoch": 0.9993514915693904,
+      "grad_norm": 0.5943860411643982,
+      "learning_rate": 3e-05,
+      "loss": 3.0217,
+      "step": 3082
+    },
+    {
+      "epoch": 0.9996757457846952,
+      "grad_norm": 0.550300657749176,
+      "learning_rate": 3e-05,
+      "loss": 3.642,
+      "step": 3083
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4627077877521515,
+      "learning_rate": 3e-05,
+      "loss": 2.9843,
+      "step": 3084
+    },
+    {
+      "epoch": 1.0,
+      "step": 3084,
+      "total_flos": 2.3769913463270277e+18,
+      "train_loss": 2.7924699883052813,
+      "train_runtime": 48184.8116,
+      "train_samples_per_second": 0.256,
+      "train_steps_per_second": 0.064
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3084,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 300,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.3769913463270277e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}