diff --git "a/checkpoint-1788/trainer_state.json" "b/checkpoint-1788/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1788/trainer_state.json"
@@ -0,0 +1,12549 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 1788,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0011185682326621924,
+      "grad_norm": 2.231250286102295,
+      "learning_rate": 5.0000000000000004e-08,
+      "loss": 1.0507,
+      "step": 1
+    },
+    {
+      "epoch": 0.0022371364653243847,
+      "grad_norm": 2.1123249530792236,
+      "learning_rate": 1.0000000000000001e-07,
+      "loss": 1.04,
+      "step": 2
+    },
+    {
+      "epoch": 0.003355704697986577,
+      "grad_norm": 2.0946707725524902,
+      "learning_rate": 1.5000000000000002e-07,
+      "loss": 1.0307,
+      "step": 3
+    },
+    {
+      "epoch": 0.0044742729306487695,
+      "grad_norm": 2.0837416648864746,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 1.0484,
+      "step": 4
+    },
+    {
+      "epoch": 0.005592841163310962,
+      "grad_norm": 1.9843275547027588,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 1.0012,
+      "step": 5
+    },
+    {
+      "epoch": 0.006711409395973154,
+      "grad_norm": 2.121988296508789,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 1.0603,
+      "step": 6
+    },
+    {
+      "epoch": 0.007829977628635347,
+      "grad_norm": 2.029029369354248,
+      "learning_rate": 3.5000000000000004e-07,
+      "loss": 1.0323,
+      "step": 7
+    },
+    {
+      "epoch": 0.008948545861297539,
+      "grad_norm": 1.9815905094146729,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 1.0273,
+      "step": 8
+    },
+    {
+      "epoch": 0.010067114093959731,
+      "grad_norm": 2.3339314460754395,
+      "learning_rate": 4.5000000000000003e-07,
+      "loss": 1.0805,
+      "step": 9
+    },
+    {
+      "epoch": 0.011185682326621925,
+      "grad_norm": 2.1078243255615234,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 1.0382,
+      "step": 10
+    },
+    {
+      "epoch": 0.012304250559284116,
+      "grad_norm": 1.8874777555465698,
+      "learning_rate": 5.5e-07,
+      "loss": 1.008,
+      "step": 11
+    },
+    {
+      "epoch": 0.013422818791946308,
+      "grad_norm": 1.9720211029052734,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 1.0065,
+      "step": 12
+    },
+    {
+      "epoch": 0.0145413870246085,
+      "grad_norm": 2.0002245903015137,
+      "learning_rate": 6.5e-07,
+      "loss": 1.0379,
+      "step": 13
+    },
+    {
+      "epoch": 0.015659955257270694,
+      "grad_norm": 1.983207106590271,
+      "learning_rate": 7.000000000000001e-07,
+      "loss": 1.0271,
+      "step": 14
+    },
+    {
+      "epoch": 0.016778523489932886,
+      "grad_norm": 1.886121153831482,
+      "learning_rate": 7.5e-07,
+      "loss": 1.0019,
+      "step": 15
+    },
+    {
+      "epoch": 0.017897091722595078,
+      "grad_norm": 1.9403958320617676,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 0.9885,
+      "step": 16
+    },
+    {
+      "epoch": 0.01901565995525727,
+      "grad_norm": 1.9739996194839478,
+      "learning_rate": 8.500000000000001e-07,
+      "loss": 0.9904,
+      "step": 17
+    },
+    {
+      "epoch": 0.020134228187919462,
+      "grad_norm": 1.7419469356536865,
+      "learning_rate": 9.000000000000001e-07,
+      "loss": 0.9709,
+      "step": 18
+    },
+    {
+      "epoch": 0.021252796420581657,
+      "grad_norm": 1.7856152057647705,
+      "learning_rate": 9.500000000000001e-07,
+      "loss": 0.9859,
+      "step": 19
+    },
+    {
+      "epoch": 0.02237136465324385,
+      "grad_norm": 1.6159933805465698,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.9895,
+      "step": 20
+    },
+    {
+      "epoch": 0.02348993288590604,
+      "grad_norm": 1.7010679244995117,
+      "learning_rate": 1.0500000000000001e-06,
+      "loss": 1.0115,
+      "step": 21
+    },
+    {
+      "epoch": 0.024608501118568233,
+      "grad_norm": 1.7860039472579956,
+      "learning_rate": 1.1e-06,
+      "loss": 0.9917,
+      "step": 22
+    },
+    {
+      "epoch": 0.025727069351230425,
+      "grad_norm": 1.3735058307647705,
+      "learning_rate": 1.1500000000000002e-06,
+      "loss": 0.9441,
+      "step": 23
+    },
+    {
+      "epoch": 0.026845637583892617,
+      "grad_norm": 1.439109206199646,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 0.9304,
+      "step": 24
+    },
+    {
+      "epoch": 0.02796420581655481,
+      "grad_norm": 1.380369782447815,
+      "learning_rate": 1.25e-06,
+      "loss": 0.9469,
+      "step": 25
+    },
+    {
+      "epoch": 0.029082774049217,
+      "grad_norm": 1.2287472486495972,
+      "learning_rate": 1.3e-06,
+      "loss": 0.8808,
+      "step": 26
+    },
+    {
+      "epoch": 0.030201342281879196,
+      "grad_norm": 1.0899194478988647,
+      "learning_rate": 1.3500000000000002e-06,
+      "loss": 0.8912,
+      "step": 27
+    },
+    {
+      "epoch": 0.03131991051454139,
+      "grad_norm": 1.0445002317428589,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 0.876,
+      "step": 28
+    },
+    {
+      "epoch": 0.03243847874720358,
+      "grad_norm": 1.0201383829116821,
+      "learning_rate": 1.45e-06,
+      "loss": 0.9003,
+      "step": 29
+    },
+    {
+      "epoch": 0.03355704697986577,
+      "grad_norm": 0.9528365731239319,
+      "learning_rate": 1.5e-06,
+      "loss": 0.8537,
+      "step": 30
+    },
+    {
+      "epoch": 0.03467561521252797,
+      "grad_norm": 0.9615768194198608,
+      "learning_rate": 1.5500000000000002e-06,
+      "loss": 0.8819,
+      "step": 31
+    },
+    {
+      "epoch": 0.035794183445190156,
+      "grad_norm": 0.9578896760940552,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 0.8859,
+      "step": 32
+    },
+    {
+      "epoch": 0.03691275167785235,
+      "grad_norm": 0.977853536605835,
+      "learning_rate": 1.6500000000000003e-06,
+      "loss": 0.8835,
+      "step": 33
+    },
+    {
+      "epoch": 0.03803131991051454,
+      "grad_norm": 0.8976068496704102,
+      "learning_rate": 1.7000000000000002e-06,
+      "loss": 0.8599,
+      "step": 34
+    },
+    {
+      "epoch": 0.039149888143176735,
+      "grad_norm": 0.8779590725898743,
+      "learning_rate": 1.75e-06,
+      "loss": 0.8708,
+      "step": 35
+    },
+    {
+      "epoch": 0.040268456375838924,
+      "grad_norm": 0.853705644607544,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 0.8465,
+      "step": 36
+    },
+    {
+      "epoch": 0.04138702460850112,
+      "grad_norm": 0.8480839729309082,
+      "learning_rate": 1.85e-06,
+      "loss": 0.8292,
+      "step": 37
+    },
+    {
+      "epoch": 0.042505592841163314,
+      "grad_norm": 0.8372538089752197,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 0.8026,
+      "step": 38
+    },
+    {
+      "epoch": 0.0436241610738255,
+      "grad_norm": 0.8592961430549622,
+      "learning_rate": 1.9500000000000004e-06,
+      "loss": 0.8153,
+      "step": 39
+    },
+    {
+      "epoch": 0.0447427293064877,
+      "grad_norm": 0.8222276568412781,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.814,
+      "step": 40
+    },
+    {
+      "epoch": 0.04586129753914989,
+      "grad_norm": 0.825672447681427,
+      "learning_rate": 2.05e-06,
+      "loss": 0.7793,
+      "step": 41
+    },
+    {
+      "epoch": 0.04697986577181208,
+      "grad_norm": 0.8016732335090637,
+      "learning_rate": 2.1000000000000002e-06,
+      "loss": 0.771,
+      "step": 42
+    },
+    {
+      "epoch": 0.04809843400447427,
+      "grad_norm": 0.7026550769805908,
+      "learning_rate": 2.15e-06,
+      "loss": 0.7664,
+      "step": 43
+    },
+    {
+      "epoch": 0.049217002237136466,
+      "grad_norm": 0.6678670644760132,
+      "learning_rate": 2.2e-06,
+      "loss": 0.7774,
+      "step": 44
+    },
+    {
+      "epoch": 0.050335570469798654,
+      "grad_norm": 0.6766750812530518,
+      "learning_rate": 2.25e-06,
+      "loss": 0.7832,
+      "step": 45
+    },
+    {
+      "epoch": 0.05145413870246085,
+      "grad_norm": 0.7094117999076843,
+      "learning_rate": 2.3000000000000004e-06,
+      "loss": 0.7861,
+      "step": 46
+    },
+    {
+      "epoch": 0.052572706935123045,
+      "grad_norm": 0.6871191263198853,
+      "learning_rate": 2.35e-06,
+      "loss": 0.7848,
+      "step": 47
+    },
+    {
+      "epoch": 0.053691275167785234,
+      "grad_norm": 0.6089867353439331,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 0.7658,
+      "step": 48
+    },
+    {
+      "epoch": 0.05480984340044743,
+      "grad_norm": 0.5112010836601257,
+      "learning_rate": 2.4500000000000003e-06,
+      "loss": 0.7921,
+      "step": 49
+    },
+    {
+      "epoch": 0.05592841163310962,
+      "grad_norm": 0.5008496046066284,
+      "learning_rate": 2.5e-06,
+      "loss": 0.7105,
+      "step": 50
+    },
+    {
+      "epoch": 0.05704697986577181,
+      "grad_norm": 0.5599631071090698,
+      "learning_rate": 2.55e-06,
+      "loss": 0.7526,
+      "step": 51
+    },
+    {
+      "epoch": 0.058165548098434,
+      "grad_norm": 0.6905913352966309,
+      "learning_rate": 2.6e-06,
+      "loss": 0.7496,
+      "step": 52
+    },
+    {
+      "epoch": 0.0592841163310962,
+      "grad_norm": 0.6198621392250061,
+      "learning_rate": 2.6500000000000005e-06,
+      "loss": 0.7297,
+      "step": 53
+    },
+    {
+      "epoch": 0.06040268456375839,
+      "grad_norm": 0.6158658862113953,
+      "learning_rate": 2.7000000000000004e-06,
+      "loss": 0.7309,
+      "step": 54
+    },
+    {
+      "epoch": 0.06152125279642058,
+      "grad_norm": 0.5798735618591309,
+      "learning_rate": 2.7500000000000004e-06,
+      "loss": 0.7102,
+      "step": 55
+    },
+    {
+      "epoch": 0.06263982102908278,
+      "grad_norm": 0.5550254583358765,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 0.7488,
+      "step": 56
+    },
+    {
+      "epoch": 0.06375838926174497,
+      "grad_norm": 0.4888734221458435,
+      "learning_rate": 2.85e-06,
+      "loss": 0.75,
+      "step": 57
+    },
+    {
+      "epoch": 0.06487695749440715,
+      "grad_norm": 0.4579496383666992,
+      "learning_rate": 2.9e-06,
+      "loss": 0.7108,
+      "step": 58
+    },
+    {
+      "epoch": 0.06599552572706935,
+      "grad_norm": 0.5775673389434814,
+      "learning_rate": 2.95e-06,
+      "loss": 0.7337,
+      "step": 59
+    },
+    {
+      "epoch": 0.06711409395973154,
+      "grad_norm": 0.5035051703453064,
+      "learning_rate": 3e-06,
+      "loss": 0.7677,
+      "step": 60
+    },
+    {
+      "epoch": 0.06823266219239374,
+      "grad_norm": 0.4771614074707031,
+      "learning_rate": 3.05e-06,
+      "loss": 0.724,
+      "step": 61
+    },
+    {
+      "epoch": 0.06935123042505593,
+      "grad_norm": 0.45495525002479553,
+      "learning_rate": 3.1000000000000004e-06,
+      "loss": 0.7393,
+      "step": 62
+    },
+    {
+      "epoch": 0.07046979865771812,
+      "grad_norm": 0.36385607719421387,
+      "learning_rate": 3.1500000000000003e-06,
+      "loss": 0.7029,
+      "step": 63
+    },
+    {
+      "epoch": 0.07158836689038031,
+      "grad_norm": 0.3554967939853668,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 0.6991,
+      "step": 64
+    },
+    {
+      "epoch": 0.07270693512304251,
+      "grad_norm": 0.36548176407814026,
+      "learning_rate": 3.2500000000000002e-06,
+      "loss": 0.7292,
+      "step": 65
+    },
+    {
+      "epoch": 0.0738255033557047,
+      "grad_norm": 0.35280168056488037,
+      "learning_rate": 3.3000000000000006e-06,
+      "loss": 0.7295,
+      "step": 66
+    },
+    {
+      "epoch": 0.07494407158836688,
+      "grad_norm": 0.3599022924900055,
+      "learning_rate": 3.3500000000000005e-06,
+      "loss": 0.6956,
+      "step": 67
+    },
+    {
+      "epoch": 0.07606263982102908,
+      "grad_norm": 0.3802206516265869,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 0.6796,
+      "step": 68
+    },
+    {
+      "epoch": 0.07718120805369127,
+      "grad_norm": 0.3787902891635895,
+      "learning_rate": 3.45e-06,
+      "loss": 0.7141,
+      "step": 69
+    },
+    {
+      "epoch": 0.07829977628635347,
+      "grad_norm": 0.374461829662323,
+      "learning_rate": 3.5e-06,
+      "loss": 0.7043,
+      "step": 70
+    },
+    {
+      "epoch": 0.07941834451901567,
+      "grad_norm": 0.34469330310821533,
+      "learning_rate": 3.5500000000000003e-06,
+      "loss": 0.7037,
+      "step": 71
+    },
+    {
+      "epoch": 0.08053691275167785,
+      "grad_norm": 0.346836119890213,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 0.7246,
+      "step": 72
+    },
+    {
+      "epoch": 0.08165548098434004,
+      "grad_norm": 0.34163376688957214,
+      "learning_rate": 3.65e-06,
+      "loss": 0.6977,
+      "step": 73
+    },
+    {
+      "epoch": 0.08277404921700224,
+      "grad_norm": 0.3481418788433075,
+      "learning_rate": 3.7e-06,
+      "loss": 0.7356,
+      "step": 74
+    },
+    {
+      "epoch": 0.08389261744966443,
+      "grad_norm": 0.3230934739112854,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.6869,
+      "step": 75
+    },
+    {
+      "epoch": 0.08501118568232663,
+      "grad_norm": 0.319917231798172,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 0.6722,
+      "step": 76
+    },
+    {
+      "epoch": 0.08612975391498881,
+      "grad_norm": 0.3535120487213135,
+      "learning_rate": 3.85e-06,
+      "loss": 0.6951,
+      "step": 77
+    },
+    {
+      "epoch": 0.087248322147651,
+      "grad_norm": 0.3229662775993347,
+      "learning_rate": 3.900000000000001e-06,
+      "loss": 0.69,
+      "step": 78
+    },
+    {
+      "epoch": 0.0883668903803132,
+      "grad_norm": 0.33365264534950256,
+      "learning_rate": 3.95e-06,
+      "loss": 0.701,
+      "step": 79
+    },
+    {
+      "epoch": 0.0894854586129754,
+      "grad_norm": 0.3302946984767914,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.6733,
+      "step": 80
+    },
+    {
+      "epoch": 0.09060402684563758,
+      "grad_norm": 0.3478582799434662,
+      "learning_rate": 4.05e-06,
+      "loss": 0.7022,
+      "step": 81
+    },
+    {
+      "epoch": 0.09172259507829977,
+      "grad_norm": 0.33355170488357544,
+      "learning_rate": 4.1e-06,
+      "loss": 0.7141,
+      "step": 82
+    },
+    {
+      "epoch": 0.09284116331096197,
+      "grad_norm": 0.3217330574989319,
+      "learning_rate": 4.15e-06,
+      "loss": 0.6799,
+      "step": 83
+    },
+    {
+      "epoch": 0.09395973154362416,
+      "grad_norm": 0.328838050365448,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 0.6943,
+      "step": 84
+    },
+    {
+      "epoch": 0.09507829977628636,
+      "grad_norm": 0.3279136121273041,
+      "learning_rate": 4.25e-06,
+      "loss": 0.6699,
+      "step": 85
+    },
+    {
+      "epoch": 0.09619686800894854,
+      "grad_norm": 0.333351194858551,
+      "learning_rate": 4.3e-06,
+      "loss": 0.712,
+      "step": 86
+    },
+    {
+      "epoch": 0.09731543624161074,
+      "grad_norm": 0.33052128553390503,
+      "learning_rate": 4.350000000000001e-06,
+      "loss": 0.7169,
+      "step": 87
+    },
+    {
+      "epoch": 0.09843400447427293,
+      "grad_norm": 0.31631597876548767,
+      "learning_rate": 4.4e-06,
+      "loss": 0.6772,
+      "step": 88
+    },
+    {
+      "epoch": 0.09955257270693513,
+      "grad_norm": 0.327311635017395,
+      "learning_rate": 4.450000000000001e-06,
+      "loss": 0.6873,
+      "step": 89
+    },
+    {
+      "epoch": 0.10067114093959731,
+      "grad_norm": 0.32048892974853516,
+      "learning_rate": 4.5e-06,
+      "loss": 0.6614,
+      "step": 90
+    },
+    {
+      "epoch": 0.1017897091722595,
+      "grad_norm": 0.32614201307296753,
+      "learning_rate": 4.5500000000000005e-06,
+      "loss": 0.7197,
+      "step": 91
+    },
+    {
+      "epoch": 0.1029082774049217,
+      "grad_norm": 0.31145355105400085,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 0.6567,
+      "step": 92
+    },
+    {
+      "epoch": 0.1040268456375839,
+      "grad_norm": 0.31379351019859314,
+      "learning_rate": 4.65e-06,
+      "loss": 0.7013,
+      "step": 93
+    },
+    {
+      "epoch": 0.10514541387024609,
+      "grad_norm": 0.32741424441337585,
+      "learning_rate": 4.7e-06,
+      "loss": 0.6737,
+      "step": 94
+    },
+    {
+      "epoch": 0.10626398210290827,
+      "grad_norm": 0.325630247592926,
+      "learning_rate": 4.75e-06,
+      "loss": 0.6673,
+      "step": 95
+    },
+    {
+      "epoch": 0.10738255033557047,
+      "grad_norm": 0.3153480291366577,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 0.6943,
+      "step": 96
+    },
+    {
+      "epoch": 0.10850111856823266,
+      "grad_norm": 0.3244793117046356,
+      "learning_rate": 4.85e-06,
+      "loss": 0.6896,
+      "step": 97
+    },
+    {
+      "epoch": 0.10961968680089486,
+      "grad_norm": 0.3078743517398834,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 0.6837,
+      "step": 98
+    },
+    {
+      "epoch": 0.11073825503355705,
+      "grad_norm": 0.3314874470233917,
+      "learning_rate": 4.95e-06,
+      "loss": 0.7416,
+      "step": 99
+    },
+    {
+      "epoch": 0.11185682326621924,
+      "grad_norm": 0.31931284070014954,
+      "learning_rate": 5e-06,
+      "loss": 0.6903,
+      "step": 100
+    },
+    {
+      "epoch": 0.11297539149888143,
+      "grad_norm": 0.3176276981830597,
+      "learning_rate": 4.999999554776598e-06,
+      "loss": 0.6761,
+      "step": 101
+    },
+    {
+      "epoch": 0.11409395973154363,
+      "grad_norm": 0.3285365700721741,
+      "learning_rate": 4.999998219106549e-06,
+      "loss": 0.6892,
+      "step": 102
+    },
+    {
+      "epoch": 0.11521252796420582,
+      "grad_norm": 0.31144094467163086,
+      "learning_rate": 4.99999599299033e-06,
+      "loss": 0.6586,
+      "step": 103
+    },
+    {
+      "epoch": 0.116331096196868,
+      "grad_norm": 0.313289076089859,
+      "learning_rate": 4.999992876428732e-06,
+      "loss": 0.708,
+      "step": 104
+    },
+    {
+      "epoch": 0.1174496644295302,
+      "grad_norm": 0.3252837061882019,
+      "learning_rate": 4.999988869422867e-06,
+      "loss": 0.7083,
+      "step": 105
+    },
+    {
+      "epoch": 0.1185682326621924,
+      "grad_norm": 0.3168275058269501,
+      "learning_rate": 4.9999839719741615e-06,
+      "loss": 0.6806,
+      "step": 106
+    },
+    {
+      "epoch": 0.11968680089485459,
+      "grad_norm": 0.31589415669441223,
+      "learning_rate": 4.9999781840843594e-06,
+      "loss": 0.6702,
+      "step": 107
+    },
+    {
+      "epoch": 0.12080536912751678,
+      "grad_norm": 0.318037748336792,
+      "learning_rate": 4.999971505755523e-06,
+      "loss": 0.6601,
+      "step": 108
+    },
+    {
+      "epoch": 0.12192393736017897,
+      "grad_norm": 0.33259475231170654,
+      "learning_rate": 4.999963936990031e-06,
+      "loss": 0.7001,
+      "step": 109
+    },
+    {
+      "epoch": 0.12304250559284116,
+      "grad_norm": 0.33322346210479736,
+      "learning_rate": 4.999955477790579e-06,
+      "loss": 0.6731,
+      "step": 110
+    },
+    {
+      "epoch": 0.12416107382550336,
+      "grad_norm": 0.31344881653785706,
+      "learning_rate": 4.999946128160179e-06,
+      "loss": 0.6667,
+      "step": 111
+    },
+    {
+      "epoch": 0.12527964205816555,
+      "grad_norm": 0.32769575715065,
+      "learning_rate": 4.999935888102162e-06,
+      "loss": 0.7123,
+      "step": 112
+    },
+    {
+      "epoch": 0.12639821029082773,
+      "grad_norm": 0.314619243144989,
+      "learning_rate": 4.9999247576201765e-06,
+      "loss": 0.683,
+      "step": 113
+    },
+    {
+      "epoch": 0.12751677852348994,
+      "grad_norm": 0.3301268219947815,
+      "learning_rate": 4.999912736718185e-06,
+      "loss": 0.6761,
+      "step": 114
+    },
+    {
+      "epoch": 0.12863534675615212,
+      "grad_norm": 0.31477460265159607,
+      "learning_rate": 4.99989982540047e-06,
+      "loss": 0.6722,
+      "step": 115
+    },
+    {
+      "epoch": 0.1297539149888143,
+      "grad_norm": 0.31430870294570923,
+      "learning_rate": 4.999886023671629e-06,
+      "loss": 0.6693,
+      "step": 116
+    },
+    {
+      "epoch": 0.13087248322147652,
+      "grad_norm": 0.31705909967422485,
+      "learning_rate": 4.999871331536581e-06,
+      "loss": 0.6567,
+      "step": 117
+    },
+    {
+      "epoch": 0.1319910514541387,
+      "grad_norm": 0.3331652879714966,
+      "learning_rate": 4.999855749000555e-06,
+      "loss": 0.6895,
+      "step": 118
+    },
+    {
+      "epoch": 0.1331096196868009,
+      "grad_norm": 0.32147714495658875,
+      "learning_rate": 4.999839276069105e-06,
+      "loss": 0.6693,
+      "step": 119
+    },
+    {
+      "epoch": 0.1342281879194631,
+      "grad_norm": 0.3312559127807617,
+      "learning_rate": 4.999821912748095e-06,
+      "loss": 0.6843,
+      "step": 120
+    },
+    {
+      "epoch": 0.13534675615212527,
+      "grad_norm": 0.34178397059440613,
+      "learning_rate": 4.999803659043712e-06,
+      "loss": 0.6774,
+      "step": 121
+    },
+    {
+      "epoch": 0.13646532438478748,
+      "grad_norm": 0.3154846727848053,
+      "learning_rate": 4.999784514962456e-06,
+      "loss": 0.6638,
+      "step": 122
+    },
+    {
+      "epoch": 0.13758389261744966,
+      "grad_norm": 0.31137940287590027,
+      "learning_rate": 4.999764480511145e-06,
+      "loss": 0.6467,
+      "step": 123
+    },
+    {
+      "epoch": 0.13870246085011187,
+      "grad_norm": 0.3188192546367645,
+      "learning_rate": 4.999743555696918e-06,
+      "loss": 0.6511,
+      "step": 124
+    },
+    {
+      "epoch": 0.13982102908277405,
+      "grad_norm": 0.30495911836624146,
+      "learning_rate": 4.999721740527225e-06,
+      "loss": 0.6637,
+      "step": 125
+    },
+    {
+      "epoch": 0.14093959731543623,
+      "grad_norm": 0.3152139186859131,
+      "learning_rate": 4.999699035009837e-06,
+      "loss": 0.6631,
+      "step": 126
+    },
+    {
+      "epoch": 0.14205816554809844,
+      "grad_norm": 0.32285481691360474,
+      "learning_rate": 4.999675439152842e-06,
+      "loss": 0.6621,
+      "step": 127
+    },
+    {
+      "epoch": 0.14317673378076062,
+      "grad_norm": 0.3176666796207428,
+      "learning_rate": 4.999650952964643e-06,
+      "loss": 0.6654,
+      "step": 128
+    },
+    {
+      "epoch": 0.14429530201342283,
+      "grad_norm": 0.314035028219223,
+      "learning_rate": 4.999625576453962e-06,
+      "loss": 0.6927,
+      "step": 129
+    },
+    {
+      "epoch": 0.14541387024608501,
+      "grad_norm": 0.3227815628051758,
+      "learning_rate": 4.999599309629839e-06,
+      "loss": 0.6865,
+      "step": 130
+    },
+    {
+      "epoch": 0.1465324384787472,
+      "grad_norm": 0.3137218952178955,
+      "learning_rate": 4.9995721525016275e-06,
+      "loss": 0.6499,
+      "step": 131
+    },
+    {
+      "epoch": 0.1476510067114094,
+      "grad_norm": 0.32401353120803833,
+      "learning_rate": 4.999544105079001e-06,
+      "loss": 0.64,
+      "step": 132
+    },
+    {
+      "epoch": 0.1487695749440716,
+      "grad_norm": 0.3110584020614624,
+      "learning_rate": 4.99951516737195e-06,
+      "loss": 0.6747,
+      "step": 133
+    },
+    {
+      "epoch": 0.14988814317673377,
+      "grad_norm": 0.3246876895427704,
+      "learning_rate": 4.999485339390781e-06,
+      "loss": 0.6943,
+      "step": 134
+    },
+    {
+      "epoch": 0.15100671140939598,
+      "grad_norm": 0.3346574008464813,
+      "learning_rate": 4.999454621146117e-06,
+      "loss": 0.6675,
+      "step": 135
+    },
+    {
+      "epoch": 0.15212527964205816,
+      "grad_norm": 0.3305971920490265,
+      "learning_rate": 4.999423012648902e-06,
+      "loss": 0.7065,
+      "step": 136
+    },
+    {
+      "epoch": 0.15324384787472037,
+      "grad_norm": 0.31732234358787537,
+      "learning_rate": 4.9993905139103924e-06,
+      "loss": 0.7038,
+      "step": 137
+    },
+    {
+      "epoch": 0.15436241610738255,
+      "grad_norm": 0.3233291208744049,
+      "learning_rate": 4.999357124942163e-06,
+      "loss": 0.6856,
+      "step": 138
+    },
+    {
+      "epoch": 0.15548098434004473,
+      "grad_norm": 0.31733304262161255,
+      "learning_rate": 4.999322845756107e-06,
+      "loss": 0.702,
+      "step": 139
+    },
+    {
+      "epoch": 0.15659955257270694,
+      "grad_norm": 0.33124351501464844,
+      "learning_rate": 4.9992876763644346e-06,
+      "loss": 0.6616,
+      "step": 140
+    },
+    {
+      "epoch": 0.15771812080536912,
+      "grad_norm": 0.3264501094818115,
+      "learning_rate": 4.999251616779671e-06,
+      "loss": 0.6773,
+      "step": 141
+    },
+    {
+      "epoch": 0.15883668903803133,
+      "grad_norm": 0.34606418013572693,
+      "learning_rate": 4.999214667014662e-06,
+      "loss": 0.6765,
+      "step": 142
+    },
+    {
+      "epoch": 0.1599552572706935,
+      "grad_norm": 0.3292436897754669,
+      "learning_rate": 4.999176827082566e-06,
+      "loss": 0.6692,
+      "step": 143
+    },
+    {
+      "epoch": 0.1610738255033557,
+      "grad_norm": 0.31322377920150757,
+      "learning_rate": 4.9991380969968615e-06,
+      "loss": 0.6811,
+      "step": 144
+    },
+    {
+      "epoch": 0.1621923937360179,
+      "grad_norm": 0.32053160667419434,
+      "learning_rate": 4.999098476771344e-06,
+      "loss": 0.6544,
+      "step": 145
+    },
+    {
+      "epoch": 0.16331096196868009,
+      "grad_norm": 0.34363314509391785,
+      "learning_rate": 4.9990579664201244e-06,
+      "loss": 0.6839,
+      "step": 146
+    },
+    {
+      "epoch": 0.1644295302013423,
+      "grad_norm": 0.3260481357574463,
+      "learning_rate": 4.999016565957633e-06,
+      "loss": 0.7048,
+      "step": 147
+    },
+    {
+      "epoch": 0.16554809843400448,
+      "grad_norm": 0.3410928547382355,
+      "learning_rate": 4.998974275398614e-06,
+      "loss": 0.6846,
+      "step": 148
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 0.33661726117134094,
+      "learning_rate": 4.998931094758132e-06,
+      "loss": 0.6468,
+      "step": 149
+    },
+    {
+      "epoch": 0.16778523489932887,
+      "grad_norm": 0.32493966817855835,
+      "learning_rate": 4.998887024051565e-06,
+      "loss": 0.6741,
+      "step": 150
+    },
+    {
+      "epoch": 0.16890380313199105,
+      "grad_norm": 0.3396671414375305,
+      "learning_rate": 4.998842063294613e-06,
+      "loss": 0.6703,
+      "step": 151
+    },
+    {
+      "epoch": 0.17002237136465326,
+      "grad_norm": 0.32851850986480713,
+      "learning_rate": 4.998796212503287e-06,
+      "loss": 0.6589,
+      "step": 152
+    },
+    {
+      "epoch": 0.17114093959731544,
+      "grad_norm": 0.33433711528778076,
+      "learning_rate": 4.99874947169392e-06,
+      "loss": 0.6594,
+      "step": 153
+    },
+    {
+      "epoch": 0.17225950782997762,
+      "grad_norm": 0.3388006091117859,
+      "learning_rate": 4.99870184088316e-06,
+      "loss": 0.6604,
+      "step": 154
+    },
+    {
+      "epoch": 0.17337807606263983,
+      "grad_norm": 0.31802693009376526,
+      "learning_rate": 4.998653320087971e-06,
+      "loss": 0.6467,
+      "step": 155
+    },
+    {
+      "epoch": 0.174496644295302,
+      "grad_norm": 0.33016613125801086,
+      "learning_rate": 4.998603909325636e-06,
+      "loss": 0.6599,
+      "step": 156
+    },
+    {
+      "epoch": 0.1756152125279642,
+      "grad_norm": 0.32546237111091614,
+      "learning_rate": 4.998553608613755e-06,
+      "loss": 0.6519,
+      "step": 157
+    },
+    {
+      "epoch": 0.1767337807606264,
+      "grad_norm": 0.3362942337989807,
+      "learning_rate": 4.998502417970242e-06,
+      "loss": 0.671,
+      "step": 158
+    },
+    {
+      "epoch": 0.17785234899328858,
+      "grad_norm": 0.33070167899131775,
+      "learning_rate": 4.998450337413331e-06,
+      "loss": 0.6624,
+      "step": 159
+    },
+    {
+      "epoch": 0.1789709172259508,
+      "grad_norm": 0.32430973649024963,
+      "learning_rate": 4.998397366961571e-06,
+      "loss": 0.6263,
+      "step": 160
+    },
+    {
+      "epoch": 0.18008948545861297,
+      "grad_norm": 0.32481464743614197,
+      "learning_rate": 4.998343506633831e-06,
+      "loss": 0.6683,
+      "step": 161
+    },
+    {
+      "epoch": 0.18120805369127516,
+      "grad_norm": 0.33035483956336975,
+      "learning_rate": 4.998288756449292e-06,
+      "loss": 0.6816,
+      "step": 162
+    },
+    {
+      "epoch": 0.18232662192393737,
+      "grad_norm": 0.33188679814338684,
+      "learning_rate": 4.998233116427458e-06,
+      "loss": 0.6693,
+      "step": 163
+    },
+    {
+      "epoch": 0.18344519015659955,
+      "grad_norm": 0.33667680621147156,
+      "learning_rate": 4.998176586588145e-06,
+      "loss": 0.6619,
+      "step": 164
+    },
+    {
+      "epoch": 0.18456375838926176,
+      "grad_norm": 0.33836281299591064,
+      "learning_rate": 4.998119166951488e-06,
+      "loss": 0.6697,
+      "step": 165
+    },
+    {
+      "epoch": 0.18568232662192394,
+      "grad_norm": 0.31710004806518555,
+      "learning_rate": 4.998060857537938e-06,
+      "loss": 0.6386,
+      "step": 166
+    },
+    {
+      "epoch": 0.18680089485458612,
+      "grad_norm": 0.3220674395561218,
+      "learning_rate": 4.9980016583682655e-06,
+      "loss": 0.6477,
+      "step": 167
+    },
+    {
+      "epoch": 0.18791946308724833,
+      "grad_norm": 0.31624945998191833,
+      "learning_rate": 4.997941569463554e-06,
+      "loss": 0.6771,
+      "step": 168
+    },
+    {
+      "epoch": 0.1890380313199105,
+      "grad_norm": 0.33520039916038513,
+      "learning_rate": 4.997880590845208e-06,
+      "loss": 0.6777,
+      "step": 169
+    },
+    {
+      "epoch": 0.19015659955257272,
+      "grad_norm": 0.33738774061203003,
+      "learning_rate": 4.997818722534944e-06,
+      "loss": 0.6603,
+      "step": 170
+    },
+    {
+      "epoch": 0.1912751677852349,
+      "grad_norm": 0.33408045768737793,
+      "learning_rate": 4.9977559645548e-06,
+      "loss": 0.6581,
+      "step": 171
+    },
+    {
+      "epoch": 0.19239373601789708,
+      "grad_norm": 0.3269501328468323,
+      "learning_rate": 4.997692316927129e-06,
+      "loss": 0.6623,
+      "step": 172
+    },
+    {
+      "epoch": 0.1935123042505593,
+      "grad_norm": 0.33073386549949646,
+      "learning_rate": 4.997627779674601e-06,
+      "loss": 0.6465,
+      "step": 173
+    },
+    {
+      "epoch": 0.19463087248322147,
+      "grad_norm": 0.33027294278144836,
+      "learning_rate": 4.997562352820201e-06,
+      "loss": 0.6795,
+      "step": 174
+    },
+    {
+      "epoch": 0.19574944071588368,
+      "grad_norm": 0.3329165577888489,
+      "learning_rate": 4.997496036387235e-06,
+      "loss": 0.6717,
+      "step": 175
+    },
+    {
+      "epoch": 0.19686800894854586,
+      "grad_norm": 0.3321872353553772,
+      "learning_rate": 4.997428830399322e-06,
+      "loss": 0.6415,
+      "step": 176
+    },
+    {
+      "epoch": 0.19798657718120805,
+      "grad_norm": 0.3222750723361969,
+      "learning_rate": 4.997360734880401e-06,
+      "loss": 0.657,
+      "step": 177
+    },
+    {
+      "epoch": 0.19910514541387025,
+      "grad_norm": 0.33294835686683655,
+      "learning_rate": 4.997291749854725e-06,
+      "loss": 0.6931,
+      "step": 178
+    },
+    {
+      "epoch": 0.20022371364653244,
+      "grad_norm": 0.3413322865962982,
+      "learning_rate": 4.997221875346863e-06,
+      "loss": 0.6761,
+      "step": 179
+    },
+    {
+      "epoch": 0.20134228187919462,
+      "grad_norm": 0.3300095796585083,
+      "learning_rate": 4.997151111381707e-06,
+      "loss": 0.6626,
+      "step": 180
+    },
+    {
+      "epoch": 0.20246085011185683,
+      "grad_norm": 0.337289035320282,
+      "learning_rate": 4.997079457984459e-06,
+      "loss": 0.6861,
+      "step": 181
+    },
+    {
+      "epoch": 0.203579418344519,
+      "grad_norm": 0.3266119658946991,
+      "learning_rate": 4.997006915180642e-06,
+      "loss": 0.6687,
+      "step": 182
+    },
+    {
+      "epoch": 0.20469798657718122,
+      "grad_norm": 0.33044853806495667,
+      "learning_rate": 4.996933482996092e-06,
+      "loss": 0.6637,
+      "step": 183
+    },
+    {
+      "epoch": 0.2058165548098434,
+      "grad_norm": 0.33716171979904175,
+      "learning_rate": 4.996859161456965e-06,
+      "loss": 0.6644,
+      "step": 184
+    },
+    {
+      "epoch": 0.20693512304250558,
+      "grad_norm": 0.32554203271865845,
+      "learning_rate": 4.996783950589733e-06,
+      "loss": 0.6524,
+      "step": 185
+    },
+    {
+      "epoch": 0.2080536912751678,
+      "grad_norm": 0.3271404504776001,
+      "learning_rate": 4.996707850421184e-06,
+      "loss": 0.6581,
+      "step": 186
+    },
+    {
+      "epoch": 0.20917225950782997,
+      "grad_norm": 0.34464138746261597,
+      "learning_rate": 4.996630860978424e-06,
+      "loss": 0.6768,
+      "step": 187
+    },
+    {
+      "epoch": 0.21029082774049218,
+      "grad_norm": 0.3408767282962799,
+      "learning_rate": 4.996552982288875e-06,
+      "loss": 0.6556,
+      "step": 188
+    },
+    {
+      "epoch": 0.21140939597315436,
+      "grad_norm": 0.3375307023525238,
+      "learning_rate": 4.996474214380276e-06,
+      "loss": 0.6819,
+      "step": 189
+    },
+    {
+      "epoch": 0.21252796420581654,
+      "grad_norm": 0.3313542902469635,
+      "learning_rate": 4.99639455728068e-06,
+      "loss": 0.6483,
+      "step": 190
+    },
+    {
+      "epoch": 0.21364653243847875,
+      "grad_norm": 0.3327822685241699,
+      "learning_rate": 4.996314011018462e-06,
+      "loss": 0.6669,
+      "step": 191
+    },
+    {
+      "epoch": 0.21476510067114093,
+      "grad_norm": 0.33021339774131775,
+      "learning_rate": 4.99623257562231e-06,
+      "loss": 0.6734,
+      "step": 192
+    },
+    {
+      "epoch": 0.21588366890380314,
+      "grad_norm": 0.32687169313430786,
+      "learning_rate": 4.996150251121229e-06,
+      "loss": 0.6387,
+      "step": 193
+    },
+    {
+      "epoch": 0.21700223713646533,
+      "grad_norm": 0.3394392728805542,
+      "learning_rate": 4.996067037544542e-06,
+      "loss": 0.6623,
+      "step": 194
+    },
+    {
+      "epoch": 0.2181208053691275,
+      "grad_norm": 0.33284223079681396,
+      "learning_rate": 4.995982934921887e-06,
+      "loss": 0.6405,
+      "step": 195
+    },
+    {
+      "epoch": 0.21923937360178972,
+      "grad_norm": 0.344235360622406,
+      "learning_rate": 4.995897943283221e-06,
+      "loss": 0.6741,
+      "step": 196
+    },
+    {
+      "epoch": 0.2203579418344519,
+      "grad_norm": 0.33437255024909973,
+      "learning_rate": 4.995812062658815e-06,
+      "loss": 0.6718,
+      "step": 197
+    },
+    {
+      "epoch": 0.2214765100671141,
+      "grad_norm": 0.3216111361980438,
+      "learning_rate": 4.995725293079257e-06,
+      "loss": 0.6709,
+      "step": 198
+    },
+    {
+      "epoch": 0.2225950782997763,
+      "grad_norm": 0.3448997139930725,
+      "learning_rate": 4.9956376345754556e-06,
+      "loss": 0.6458,
+      "step": 199
+    },
+    {
+      "epoch": 0.22371364653243847,
+      "grad_norm": 0.34354478120803833,
+      "learning_rate": 4.99554908717863e-06,
+      "loss": 0.6615,
+      "step": 200
+    },
+    {
+      "epoch": 0.22483221476510068,
+      "grad_norm": 0.3417740762233734,
+      "learning_rate": 4.99545965092032e-06,
+      "loss": 0.6847,
+      "step": 201
+    },
+    {
+      "epoch": 0.22595078299776286,
+      "grad_norm": 0.3366676867008209,
+      "learning_rate": 4.99536932583238e-06,
+      "loss": 0.649,
+      "step": 202
+    },
+    {
+      "epoch": 0.22706935123042504,
+      "grad_norm": 0.3610089421272278,
+      "learning_rate": 4.995278111946983e-06,
+      "loss": 0.6616,
+      "step": 203
+    },
+    {
+      "epoch": 0.22818791946308725,
+      "grad_norm": 0.3359774053096771,
+      "learning_rate": 4.995186009296618e-06,
+      "loss": 0.6519,
+      "step": 204
+    },
+    {
+      "epoch": 0.22930648769574943,
+      "grad_norm": 0.34075963497161865,
+      "learning_rate": 4.9950930179140885e-06,
+      "loss": 0.6762,
+      "step": 205
+    },
+    {
+      "epoch": 0.23042505592841164,
+      "grad_norm": 0.32416507601737976,
+      "learning_rate": 4.994999137832517e-06,
+      "loss": 0.6499,
+      "step": 206
+    },
+    {
+      "epoch": 0.23154362416107382,
+      "grad_norm": 0.32749176025390625,
+      "learning_rate": 4.99490436908534e-06,
+      "loss": 0.645,
+      "step": 207
+    },
+    {
+      "epoch": 0.232662192393736,
+      "grad_norm": 0.3349708318710327,
+      "learning_rate": 4.994808711706314e-06,
+      "loss": 0.6676,
+      "step": 208
+    },
+    {
+      "epoch": 0.23378076062639822,
+      "grad_norm": 0.3491227328777313,
+      "learning_rate": 4.9947121657295094e-06,
+      "loss": 0.6287,
+      "step": 209
+    },
+    {
+      "epoch": 0.2348993288590604,
+      "grad_norm": 0.3400874733924866,
+      "learning_rate": 4.994614731189314e-06,
+      "loss": 0.6473,
+      "step": 210
+    },
+    {
+      "epoch": 0.2360178970917226,
+      "grad_norm": 0.3388952612876892,
+      "learning_rate": 4.994516408120432e-06,
+      "loss": 0.6821,
+      "step": 211
+    },
+    {
+      "epoch": 0.2371364653243848,
+      "grad_norm": 0.33224812150001526,
+      "learning_rate": 4.994417196557884e-06,
+      "loss": 0.649,
+      "step": 212
+    },
+    {
+      "epoch": 0.23825503355704697,
+      "grad_norm": 0.3307199478149414,
+      "learning_rate": 4.994317096537006e-06,
+      "loss": 0.6581,
+      "step": 213
+    },
+    {
+      "epoch": 0.23937360178970918,
+      "grad_norm": 0.3505164682865143,
+      "learning_rate": 4.994216108093452e-06,
+      "loss": 0.6498,
+      "step": 214
+    },
+    {
+      "epoch": 0.24049217002237136,
+      "grad_norm": 0.3284938335418701,
+      "learning_rate": 4.994114231263193e-06,
+      "loss": 0.6503,
+      "step": 215
+    },
+    {
+      "epoch": 0.24161073825503357,
+      "grad_norm": 0.3475011885166168,
+      "learning_rate": 4.994011466082514e-06,
+      "loss": 0.6724,
+      "step": 216
+    },
+    {
+      "epoch": 0.24272930648769575,
+      "grad_norm": 0.34667858481407166,
+      "learning_rate": 4.993907812588019e-06,
+      "loss": 0.6373,
+      "step": 217
+    },
+    {
+      "epoch": 0.24384787472035793,
+      "grad_norm": 0.34276899695396423,
+      "learning_rate": 4.993803270816627e-06,
+      "loss": 0.6513,
+      "step": 218
+    },
+    {
+      "epoch": 0.24496644295302014,
+      "grad_norm": 0.34308409690856934,
+      "learning_rate": 4.993697840805572e-06,
+      "loss": 0.6596,
+      "step": 219
+    },
+    {
+      "epoch": 0.24608501118568232,
+      "grad_norm": 0.33108261227607727,
+      "learning_rate": 4.9935915225924075e-06,
+      "loss": 0.6623,
+      "step": 220
+    },
+    {
+      "epoch": 0.24720357941834453,
+      "grad_norm": 0.3529888093471527,
+      "learning_rate": 4.9934843162150015e-06,
+      "loss": 0.658,
+      "step": 221
+    },
+    {
+      "epoch": 0.2483221476510067,
+      "grad_norm": 0.3457166850566864,
+      "learning_rate": 4.993376221711538e-06,
+      "loss": 0.6342,
+      "step": 222
+    },
+    {
+      "epoch": 0.2494407158836689,
+      "grad_norm": 0.35108813643455505,
+      "learning_rate": 4.993267239120519e-06,
+      "loss": 0.6325,
+      "step": 223
+    },
+    {
+      "epoch": 0.2505592841163311,
+      "grad_norm": 0.3448682129383087,
+      "learning_rate": 4.993157368480761e-06,
+      "loss": 0.6746,
+      "step": 224
+    },
+    {
+      "epoch": 0.2516778523489933,
+      "grad_norm": 0.34094589948654175,
+      "learning_rate": 4.993046609831397e-06,
+      "loss": 0.6313,
+      "step": 225
+    },
+    {
+      "epoch": 0.25279642058165547,
+      "grad_norm": 0.33934327960014343,
+      "learning_rate": 4.9929349632118785e-06,
+      "loss": 0.6371,
+      "step": 226
+    },
+    {
+      "epoch": 0.2539149888143177,
+      "grad_norm": 0.3517382740974426,
+      "learning_rate": 4.99282242866197e-06,
+      "loss": 0.6411,
+      "step": 227
+    },
+    {
+      "epoch": 0.2550335570469799,
+      "grad_norm": 0.34098172187805176,
+      "learning_rate": 4.992709006221755e-06,
+      "loss": 0.6648,
+      "step": 228
+    },
+    {
+      "epoch": 0.25615212527964204,
+      "grad_norm": 0.3397183120250702,
+      "learning_rate": 4.992594695931632e-06,
+      "loss": 0.6038,
+      "step": 229
+    },
+    {
+      "epoch": 0.25727069351230425,
+      "grad_norm": 0.35994404554367065,
+      "learning_rate": 4.992479497832316e-06,
+      "loss": 0.6832,
+      "step": 230
+    },
+    {
+      "epoch": 0.25838926174496646,
+      "grad_norm": 0.3505656123161316,
+      "learning_rate": 4.992363411964838e-06,
+      "loss": 0.682,
+      "step": 231
+    },
+    {
+      "epoch": 0.2595078299776286,
+      "grad_norm": 0.343159556388855,
+      "learning_rate": 4.992246438370545e-06,
+      "loss": 0.6597,
+      "step": 232
+    },
+    {
+      "epoch": 0.2606263982102908,
+      "grad_norm": 0.36491280794143677,
+      "learning_rate": 4.9921285770911e-06,
+      "loss": 0.6422,
+      "step": 233
+    },
+    {
+      "epoch": 0.26174496644295303,
+      "grad_norm": 0.3656606078147888,
+      "learning_rate": 4.992009828168484e-06,
+      "loss": 0.6988,
+      "step": 234
+    },
+    {
+      "epoch": 0.26286353467561524,
+      "grad_norm": 0.3348519206047058,
+      "learning_rate": 4.991890191644993e-06,
+      "loss": 0.6281,
+      "step": 235
+    },
+    {
+      "epoch": 0.2639821029082774,
+      "grad_norm": 0.3407367467880249,
+      "learning_rate": 4.991769667563237e-06,
+      "loss": 0.6487,
+      "step": 236
+    },
+    {
+      "epoch": 0.2651006711409396,
+      "grad_norm": 0.3644556999206543,
+      "learning_rate": 4.991648255966145e-06,
+      "loss": 0.6443,
+      "step": 237
+    },
+    {
+      "epoch": 0.2662192393736018,
+      "grad_norm": 0.3346659243106842,
+      "learning_rate": 4.991525956896962e-06,
+      "loss": 0.632,
+      "step": 238
+    },
+    {
+      "epoch": 0.26733780760626397,
+      "grad_norm": 0.36535120010375977,
+      "learning_rate": 4.991402770399249e-06,
+      "loss": 0.6347,
+      "step": 239
+    },
+    {
+      "epoch": 0.2684563758389262,
+      "grad_norm": 0.34253549575805664,
+      "learning_rate": 4.991278696516879e-06,
+      "loss": 0.6946,
+      "step": 240
+    },
+    {
+      "epoch": 0.2695749440715884,
+      "grad_norm": 0.35125476121902466,
+      "learning_rate": 4.9911537352940485e-06,
+      "loss": 0.6669,
+      "step": 241
+    },
+    {
+      "epoch": 0.27069351230425054,
+      "grad_norm": 0.35836276412010193,
+      "learning_rate": 4.991027886775264e-06,
+      "loss": 0.6534,
+      "step": 242
+    },
+    {
+      "epoch": 0.27181208053691275,
+      "grad_norm": 0.34344252943992615,
+      "learning_rate": 4.990901151005349e-06,
+      "loss": 0.6595,
+      "step": 243
+    },
+    {
+      "epoch": 0.27293064876957496,
+      "grad_norm": 0.35524797439575195,
+      "learning_rate": 4.9907735280294465e-06,
+      "loss": 0.6612,
+      "step": 244
+    },
+    {
+      "epoch": 0.2740492170022371,
+      "grad_norm": 0.3483973741531372,
+      "learning_rate": 4.990645017893013e-06,
+      "loss": 0.6694,
+      "step": 245
+    },
+    {
+      "epoch": 0.2751677852348993,
+      "grad_norm": 0.34605199098587036,
+      "learning_rate": 4.990515620641819e-06,
+      "loss": 0.6453,
+      "step": 246
+    },
+    {
+      "epoch": 0.27628635346756153,
+      "grad_norm": 0.3441944122314453,
+      "learning_rate": 4.990385336321954e-06,
+      "loss": 0.6356,
+      "step": 247
+    },
+    {
+      "epoch": 0.27740492170022374,
+      "grad_norm": 0.36233291029930115,
+      "learning_rate": 4.990254164979823e-06,
+      "loss": 0.673,
+      "step": 248
+    },
+    {
+      "epoch": 0.2785234899328859,
+      "grad_norm": 0.3624320328235626,
+      "learning_rate": 4.990122106662145e-06,
+      "loss": 0.6459,
+      "step": 249
+    },
+    {
+      "epoch": 0.2796420581655481,
+      "grad_norm": 0.3635922372341156,
+      "learning_rate": 4.989989161415959e-06,
+      "loss": 0.6552,
+      "step": 250
+    },
+    {
+      "epoch": 0.2807606263982103,
+      "grad_norm": 0.3370678424835205,
+      "learning_rate": 4.989855329288615e-06,
+      "loss": 0.6098,
+      "step": 251
+    },
+    {
+      "epoch": 0.28187919463087246,
+      "grad_norm": 0.35488393902778625,
+      "learning_rate": 4.989720610327782e-06,
+      "loss": 0.6554,
+      "step": 252
+    },
+    {
+      "epoch": 0.2829977628635347,
+      "grad_norm": 0.3495752513408661,
+      "learning_rate": 4.989585004581444e-06,
+      "loss": 0.6339,
+      "step": 253
+    },
+    {
+      "epoch": 0.2841163310961969,
+      "grad_norm": 0.3451905846595764,
+      "learning_rate": 4.989448512097901e-06,
+      "loss": 0.6954,
+      "step": 254
+    },
+    {
+      "epoch": 0.28523489932885904,
+      "grad_norm": 0.3532280921936035,
+      "learning_rate": 4.989311132925768e-06,
+      "loss": 0.6198,
+      "step": 255
+    },
+    {
+      "epoch": 0.28635346756152125,
+      "grad_norm": 0.3570882976055145,
+      "learning_rate": 4.989172867113976e-06,
+      "loss": 0.6492,
+      "step": 256
+    },
+    {
+      "epoch": 0.28747203579418346,
+      "grad_norm": 0.33201339840888977,
+      "learning_rate": 4.9890337147117755e-06,
+      "loss": 0.6324,
+      "step": 257
+    },
+    {
+      "epoch": 0.28859060402684567,
+      "grad_norm": 0.3354071080684662,
+      "learning_rate": 4.988893675768726e-06,
+      "loss": 0.628,
+      "step": 258
+    },
+    {
+      "epoch": 0.2897091722595078,
+      "grad_norm": 0.341133713722229,
+      "learning_rate": 4.988752750334708e-06,
+      "loss": 0.6316,
+      "step": 259
+    },
+    {
+      "epoch": 0.29082774049217003,
+      "grad_norm": 0.33613571524620056,
+      "learning_rate": 4.9886109384599165e-06,
+      "loss": 0.6401,
+      "step": 260
+    },
+    {
+      "epoch": 0.29194630872483224,
+      "grad_norm": 0.3657302260398865,
+      "learning_rate": 4.988468240194861e-06,
+      "loss": 0.6743,
+      "step": 261
+    },
+    {
+      "epoch": 0.2930648769574944,
+      "grad_norm": 0.3349529504776001,
+      "learning_rate": 4.988324655590369e-06,
+      "loss": 0.6121,
+      "step": 262
+    },
+    {
+      "epoch": 0.2941834451901566,
+      "grad_norm": 0.34364601969718933,
+      "learning_rate": 4.98818018469758e-06,
+      "loss": 0.6427,
+      "step": 263
+    },
+    {
+      "epoch": 0.2953020134228188,
+      "grad_norm": 0.35895341634750366,
+      "learning_rate": 4.988034827567953e-06,
+      "loss": 0.6913,
+      "step": 264
+    },
+    {
+      "epoch": 0.29642058165548096,
+      "grad_norm": 0.3571792244911194,
+      "learning_rate": 4.987888584253262e-06,
+      "loss": 0.6286,
+      "step": 265
+    },
+    {
+      "epoch": 0.2975391498881432,
+      "grad_norm": 0.3684253394603729,
+      "learning_rate": 4.987741454805594e-06,
+      "loss": 0.6365,
+      "step": 266
+    },
+    {
+      "epoch": 0.2986577181208054,
+      "grad_norm": 0.3521265983581543,
+      "learning_rate": 4.987593439277353e-06,
+      "loss": 0.6172,
+      "step": 267
+    },
+    {
+      "epoch": 0.29977628635346754,
+      "grad_norm": 0.34943872690200806,
+      "learning_rate": 4.98744453772126e-06,
+      "loss": 0.6382,
+      "step": 268
+    },
+    {
+      "epoch": 0.30089485458612975,
+      "grad_norm": 0.35075268149375916,
+      "learning_rate": 4.9872947501903515e-06,
+      "loss": 0.6497,
+      "step": 269
+    },
+    {
+      "epoch": 0.30201342281879195,
+      "grad_norm": 0.3558429479598999,
+      "learning_rate": 4.987144076737978e-06,
+      "loss": 0.6561,
+      "step": 270
+    },
+    {
+      "epoch": 0.30313199105145416,
+      "grad_norm": 0.35898861289024353,
+      "learning_rate": 4.986992517417805e-06,
+      "loss": 0.6613,
+      "step": 271
+    },
+    {
+      "epoch": 0.3042505592841163,
+      "grad_norm": 0.35135966539382935,
+      "learning_rate": 4.986840072283815e-06,
+      "loss": 0.6507,
+      "step": 272
+    },
+    {
+      "epoch": 0.3053691275167785,
+      "grad_norm": 0.3530326783657074,
+      "learning_rate": 4.986686741390308e-06,
+      "loss": 0.6459,
+      "step": 273
+    },
+    {
+      "epoch": 0.30648769574944074,
+      "grad_norm": 0.34631597995758057,
+      "learning_rate": 4.986532524791894e-06,
+      "loss": 0.6074,
+      "step": 274
+    },
+    {
+      "epoch": 0.3076062639821029,
+      "grad_norm": 0.37036123871803284,
+      "learning_rate": 4.986377422543503e-06,
+      "loss": 0.6416,
+      "step": 275
+    },
+    {
+      "epoch": 0.3087248322147651,
+      "grad_norm": 0.3576701283454895,
+      "learning_rate": 4.98622143470038e-06,
+      "loss": 0.5939,
+      "step": 276
+    },
+    {
+      "epoch": 0.3098434004474273,
+      "grad_norm": 0.35285863280296326,
+      "learning_rate": 4.986064561318083e-06,
+      "loss": 0.6405,
+      "step": 277
+    },
+    {
+      "epoch": 0.31096196868008946,
+      "grad_norm": 0.35576099157333374,
+      "learning_rate": 4.985906802452488e-06,
+      "loss": 0.6348,
+      "step": 278
+    },
+    {
+      "epoch": 0.31208053691275167,
+      "grad_norm": 0.35739952325820923,
+      "learning_rate": 4.985748158159785e-06,
+      "loss": 0.65,
+      "step": 279
+    },
+    {
+      "epoch": 0.3131991051454139,
+      "grad_norm": 0.35885411500930786,
+      "learning_rate": 4.985588628496481e-06,
+      "loss": 0.6575,
+      "step": 280
+    },
+    {
+      "epoch": 0.3143176733780761,
+      "grad_norm": 0.36760038137435913,
+      "learning_rate": 4.985428213519396e-06,
+      "loss": 0.6606,
+      "step": 281
+    },
+    {
+      "epoch": 0.31543624161073824,
+      "grad_norm": 0.35371875762939453,
+      "learning_rate": 4.9852669132856645e-06,
+      "loss": 0.6495,
+      "step": 282
+    },
+    {
+      "epoch": 0.31655480984340045,
+      "grad_norm": 0.3489198684692383,
+      "learning_rate": 4.985104727852741e-06,
+      "loss": 0.6402,
+      "step": 283
+    },
+    {
+      "epoch": 0.31767337807606266,
+      "grad_norm": 0.3595716953277588,
+      "learning_rate": 4.984941657278392e-06,
+      "loss": 0.6495,
+      "step": 284
+    },
+    {
+      "epoch": 0.3187919463087248,
+      "grad_norm": 0.3551527261734009,
+      "learning_rate": 4.984777701620698e-06,
+      "loss": 0.6555,
+      "step": 285
+    },
+    {
+      "epoch": 0.319910514541387,
+      "grad_norm": 0.35428524017333984,
+      "learning_rate": 4.984612860938059e-06,
+      "loss": 0.6435,
+      "step": 286
+    },
+    {
+      "epoch": 0.32102908277404923,
+      "grad_norm": 0.36495980620384216,
+      "learning_rate": 4.984447135289185e-06,
+      "loss": 0.6375,
+      "step": 287
+    },
+    {
+      "epoch": 0.3221476510067114,
+      "grad_norm": 0.36956459283828735,
+      "learning_rate": 4.984280524733107e-06,
+      "loss": 0.654,
+      "step": 288
+    },
+    {
+      "epoch": 0.3232662192393736,
+      "grad_norm": 0.34770330786705017,
+      "learning_rate": 4.984113029329166e-06,
+      "loss": 0.6313,
+      "step": 289
+    },
+    {
+      "epoch": 0.3243847874720358,
+      "grad_norm": 0.3676060140132904,
+      "learning_rate": 4.9839446491370215e-06,
+      "loss": 0.6697,
+      "step": 290
+    },
+    {
+      "epoch": 0.32550335570469796,
+      "grad_norm": 0.3490992486476898,
+      "learning_rate": 4.983775384216646e-06,
+      "loss": 0.6343,
+      "step": 291
+    },
+    {
+      "epoch": 0.32662192393736017,
+      "grad_norm": 0.36720773577690125,
+      "learning_rate": 4.983605234628328e-06,
+      "loss": 0.6609,
+      "step": 292
+    },
+    {
+      "epoch": 0.3277404921700224,
+      "grad_norm": 0.3619595766067505,
+      "learning_rate": 4.983434200432672e-06,
+      "loss": 0.6635,
+      "step": 293
+    },
+    {
+      "epoch": 0.3288590604026846,
+      "grad_norm": 0.35261741280555725,
+      "learning_rate": 4.983262281690596e-06,
+      "loss": 0.6273,
+      "step": 294
+    },
+    {
+      "epoch": 0.32997762863534674,
+      "grad_norm": 0.34182801842689514,
+      "learning_rate": 4.983089478463335e-06,
+      "loss": 0.6271,
+      "step": 295
+    },
+    {
+      "epoch": 0.33109619686800895,
+      "grad_norm": 0.3623373806476593,
+      "learning_rate": 4.982915790812436e-06,
+      "loss": 0.6491,
+      "step": 296
+    },
+    {
+      "epoch": 0.33221476510067116,
+      "grad_norm": 0.3656613826751709,
+      "learning_rate": 4.982741218799763e-06,
+      "loss": 0.6672,
+      "step": 297
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.3533206880092621,
+      "learning_rate": 4.982565762487498e-06,
+      "loss": 0.6257,
+      "step": 298
+    },
+    {
+      "epoch": 0.3344519015659955,
+      "grad_norm": 0.35345301032066345,
+      "learning_rate": 4.982389421938131e-06,
+      "loss": 0.6486,
+      "step": 299
+    },
+    {
+      "epoch": 0.33557046979865773,
+      "grad_norm": 0.36566492915153503,
+      "learning_rate": 4.982212197214472e-06,
+      "loss": 0.6411,
+      "step": 300
+    },
+    {
+      "epoch": 0.3366890380313199,
+      "grad_norm": 0.3681536018848419,
+      "learning_rate": 4.982034088379646e-06,
+      "loss": 0.6208,
+      "step": 301
+    },
+    {
+      "epoch": 0.3378076062639821,
+      "grad_norm": 0.3649962246417999,
+      "learning_rate": 4.98185509549709e-06,
+      "loss": 0.6557,
+      "step": 302
+    },
+    {
+      "epoch": 0.3389261744966443,
+      "grad_norm": 0.350563108921051,
+      "learning_rate": 4.981675218630557e-06,
+      "loss": 0.6361,
+      "step": 303
+    },
+    {
+      "epoch": 0.3400447427293065,
+      "grad_norm": 0.35432034730911255,
+      "learning_rate": 4.981494457844117e-06,
+      "loss": 0.6294,
+      "step": 304
+    },
+    {
+      "epoch": 0.34116331096196867,
+      "grad_norm": 0.3590239882469177,
+      "learning_rate": 4.981312813202153e-06,
+      "loss": 0.6318,
+      "step": 305
+    },
+    {
+      "epoch": 0.3422818791946309,
+      "grad_norm": 0.3664059042930603,
+      "learning_rate": 4.981130284769361e-06,
+      "loss": 0.648,
+      "step": 306
+    },
+    {
+      "epoch": 0.3434004474272931,
+      "grad_norm": 0.35878944396972656,
+      "learning_rate": 4.9809468726107555e-06,
+      "loss": 0.619,
+      "step": 307
+    },
+    {
+      "epoch": 0.34451901565995524,
+      "grad_norm": 0.35763007402420044,
+      "learning_rate": 4.980762576791664e-06,
+      "loss": 0.6567,
+      "step": 308
+    },
+    {
+      "epoch": 0.34563758389261745,
+      "grad_norm": 0.35878288745880127,
+      "learning_rate": 4.980577397377728e-06,
+      "loss": 0.6421,
+      "step": 309
+    },
+    {
+      "epoch": 0.34675615212527966,
+      "grad_norm": 0.3597519099712372,
+      "learning_rate": 4.980391334434906e-06,
+      "loss": 0.6352,
+      "step": 310
+    },
+    {
+      "epoch": 0.3478747203579418,
+      "grad_norm": 0.3498169481754303,
+      "learning_rate": 4.980204388029466e-06,
+      "loss": 0.626,
+      "step": 311
+    },
+    {
+      "epoch": 0.348993288590604,
+      "grad_norm": 0.36353105306625366,
+      "learning_rate": 4.980016558227998e-06,
+      "loss": 0.6513,
+      "step": 312
+    },
+    {
+      "epoch": 0.35011185682326623,
+      "grad_norm": 0.355794221162796,
+      "learning_rate": 4.979827845097402e-06,
+      "loss": 0.6385,
+      "step": 313
+    },
+    {
+      "epoch": 0.3512304250559284,
+      "grad_norm": 0.3594406545162201,
+      "learning_rate": 4.979638248704894e-06,
+      "loss": 0.6134,
+      "step": 314
+    },
+    {
+      "epoch": 0.3523489932885906,
+      "grad_norm": 0.3639025092124939,
+      "learning_rate": 4.979447769118002e-06,
+      "loss": 0.6386,
+      "step": 315
+    },
+    {
+      "epoch": 0.3534675615212528,
+      "grad_norm": 0.3681359887123108,
+      "learning_rate": 4.979256406404574e-06,
+      "loss": 0.6213,
+      "step": 316
+    },
+    {
+      "epoch": 0.354586129753915,
+      "grad_norm": 0.35021546483039856,
+      "learning_rate": 4.979064160632766e-06,
+      "loss": 0.5933,
+      "step": 317
+    },
+    {
+      "epoch": 0.35570469798657717,
+      "grad_norm": 0.35783469676971436,
+      "learning_rate": 4.978871031871054e-06,
+      "loss": 0.6054,
+      "step": 318
+    },
+    {
+      "epoch": 0.3568232662192394,
+      "grad_norm": 0.3768575191497803,
+      "learning_rate": 4.978677020188226e-06,
+      "loss": 0.651,
+      "step": 319
+    },
+    {
+      "epoch": 0.3579418344519016,
+      "grad_norm": 0.3841581642627716,
+      "learning_rate": 4.978482125653385e-06,
+      "loss": 0.6447,
+      "step": 320
+    },
+    {
+      "epoch": 0.35906040268456374,
+      "grad_norm": 0.3749678134918213,
+      "learning_rate": 4.978286348335949e-06,
+      "loss": 0.6403,
+      "step": 321
+    },
+    {
+      "epoch": 0.36017897091722595,
+      "grad_norm": 0.35757121443748474,
+      "learning_rate": 4.978089688305647e-06,
+      "loss": 0.6297,
+      "step": 322
+    },
+    {
+      "epoch": 0.36129753914988816,
+      "grad_norm": 0.36303797364234924,
+      "learning_rate": 4.977892145632528e-06,
+      "loss": 0.6438,
+      "step": 323
+    },
+    {
+      "epoch": 0.3624161073825503,
+      "grad_norm": 0.3670295476913452,
+      "learning_rate": 4.977693720386951e-06,
+      "loss": 0.6055,
+      "step": 324
+    },
+    {
+      "epoch": 0.3635346756152125,
+      "grad_norm": 0.3521486818790436,
+      "learning_rate": 4.977494412639591e-06,
+      "loss": 0.6072,
+      "step": 325
+    },
+    {
+      "epoch": 0.36465324384787473,
+      "grad_norm": 0.35688498616218567,
+      "learning_rate": 4.9772942224614375e-06,
+      "loss": 0.6252,
+      "step": 326
+    },
+    {
+      "epoch": 0.36577181208053694,
+      "grad_norm": 0.3542022109031677,
+      "learning_rate": 4.9770931499237925e-06,
+      "loss": 0.6407,
+      "step": 327
+    },
+    {
+      "epoch": 0.3668903803131991,
+      "grad_norm": 0.3773319125175476,
+      "learning_rate": 4.976891195098277e-06,
+      "loss": 0.6524,
+      "step": 328
+    },
+    {
+      "epoch": 0.3680089485458613,
+      "grad_norm": 0.3697628080844879,
+      "learning_rate": 4.97668835805682e-06,
+      "loss": 0.6503,
+      "step": 329
+    },
+    {
+      "epoch": 0.3691275167785235,
+      "grad_norm": 0.36458712816238403,
+      "learning_rate": 4.976484638871669e-06,
+      "loss": 0.6722,
+      "step": 330
+    },
+    {
+      "epoch": 0.37024608501118567,
+      "grad_norm": 0.3523670434951782,
+      "learning_rate": 4.976280037615385e-06,
+      "loss": 0.6273,
+      "step": 331
+    },
+    {
+      "epoch": 0.3713646532438479,
+      "grad_norm": 0.35473543405532837,
+      "learning_rate": 4.9760745543608414e-06,
+      "loss": 0.6243,
+      "step": 332
+    },
+    {
+      "epoch": 0.3724832214765101,
+      "grad_norm": 0.36719077825546265,
+      "learning_rate": 4.9758681891812276e-06,
+      "loss": 0.6476,
+      "step": 333
+    },
+    {
+      "epoch": 0.37360178970917224,
+      "grad_norm": 0.3712293207645416,
+      "learning_rate": 4.9756609421500464e-06,
+      "loss": 0.6475,
+      "step": 334
+    },
+    {
+      "epoch": 0.37472035794183445,
+      "grad_norm": 0.36749109625816345,
+      "learning_rate": 4.9754528133411144e-06,
+      "loss": 0.6428,
+      "step": 335
+    },
+    {
+      "epoch": 0.37583892617449666,
+      "grad_norm": 0.35749316215515137,
+      "learning_rate": 4.975243802828563e-06,
+      "loss": 0.6123,
+      "step": 336
+    },
+    {
+      "epoch": 0.3769574944071588,
+      "grad_norm": 0.37466734647750854,
+      "learning_rate": 4.975033910686837e-06,
+      "loss": 0.6393,
+      "step": 337
+    },
+    {
+      "epoch": 0.378076062639821,
+      "grad_norm": 0.36750441789627075,
+      "learning_rate": 4.974823136990697e-06,
+      "loss": 0.6405,
+      "step": 338
+    },
+    {
+      "epoch": 0.37919463087248323,
+      "grad_norm": 0.3831922113895416,
+      "learning_rate": 4.9746114818152135e-06,
+      "loss": 0.6633,
+      "step": 339
+    },
+    {
+      "epoch": 0.38031319910514544,
+      "grad_norm": 0.35679274797439575,
+      "learning_rate": 4.974398945235776e-06,
+      "loss": 0.6431,
+      "step": 340
+    },
+    {
+      "epoch": 0.3814317673378076,
+      "grad_norm": 0.36524152755737305,
+      "learning_rate": 4.974185527328084e-06,
+      "loss": 0.6419,
+      "step": 341
+    },
+    {
+      "epoch": 0.3825503355704698,
+      "grad_norm": 0.3668903410434723,
+      "learning_rate": 4.9739712281681525e-06,
+      "loss": 0.6418,
+      "step": 342
+    },
+    {
+      "epoch": 0.383668903803132,
+      "grad_norm": 0.37841862440109253,
+      "learning_rate": 4.973756047832312e-06,
+      "loss": 0.6585,
+      "step": 343
+    },
+    {
+      "epoch": 0.38478747203579416,
+      "grad_norm": 0.37758868932724,
+      "learning_rate": 4.9735399863972024e-06,
+      "loss": 0.6493,
+      "step": 344
+    },
+    {
+      "epoch": 0.3859060402684564,
+      "grad_norm": 0.3663494288921356,
+      "learning_rate": 4.973323043939783e-06,
+      "loss": 0.6728,
+      "step": 345
+    },
+    {
+      "epoch": 0.3870246085011186,
+      "grad_norm": 0.3930216431617737,
+      "learning_rate": 4.973105220537322e-06,
+      "loss": 0.6608,
+      "step": 346
+    },
+    {
+      "epoch": 0.38814317673378074,
+      "grad_norm": 0.390828400850296,
+      "learning_rate": 4.972886516267404e-06,
+      "loss": 0.6497,
+      "step": 347
+    },
+    {
+      "epoch": 0.38926174496644295,
+      "grad_norm": 0.37600764632225037,
+      "learning_rate": 4.972666931207927e-06,
+      "loss": 0.6426,
+      "step": 348
+    },
+    {
+      "epoch": 0.39038031319910516,
+      "grad_norm": 0.36520275473594666,
+      "learning_rate": 4.972446465437103e-06,
+      "loss": 0.645,
+      "step": 349
+    },
+    {
+      "epoch": 0.39149888143176736,
+      "grad_norm": 0.3984422981739044,
+      "learning_rate": 4.972225119033457e-06,
+      "loss": 0.6368,
+      "step": 350
+    },
+    {
+      "epoch": 0.3926174496644295,
+      "grad_norm": 0.3725559711456299,
+      "learning_rate": 4.972002892075827e-06,
+      "loss": 0.625,
+      "step": 351
+    },
+    {
+      "epoch": 0.39373601789709173,
+      "grad_norm": 0.3889387547969818,
+      "learning_rate": 4.9717797846433655e-06,
+      "loss": 0.6258,
+      "step": 352
+    },
+    {
+      "epoch": 0.39485458612975394,
+      "grad_norm": 0.37537676095962524,
+      "learning_rate": 4.97155579681554e-06,
+      "loss": 0.64,
+      "step": 353
+    },
+    {
+      "epoch": 0.3959731543624161,
+      "grad_norm": 0.3795606791973114,
+      "learning_rate": 4.97133092867213e-06,
+      "loss": 0.6849,
+      "step": 354
+    },
+    {
+      "epoch": 0.3970917225950783,
+      "grad_norm": 0.38519197702407837,
+      "learning_rate": 4.971105180293228e-06,
+      "loss": 0.6493,
+      "step": 355
+    },
+    {
+      "epoch": 0.3982102908277405,
+      "grad_norm": 0.3789883553981781,
+      "learning_rate": 4.97087855175924e-06,
+      "loss": 0.6281,
+      "step": 356
+    },
+    {
+      "epoch": 0.39932885906040266,
+      "grad_norm": 0.3687867820262909,
+      "learning_rate": 4.970651043150887e-06,
+      "loss": 0.6278,
+      "step": 357
+    },
+    {
+      "epoch": 0.4004474272930649,
+      "grad_norm": 0.365581214427948,
+      "learning_rate": 4.970422654549204e-06,
+      "loss": 0.647,
+      "step": 358
+    },
+    {
+      "epoch": 0.4015659955257271,
+      "grad_norm": 0.37871256470680237,
+      "learning_rate": 4.970193386035537e-06,
+      "loss": 0.6349,
+      "step": 359
+    },
+    {
+      "epoch": 0.40268456375838924,
+      "grad_norm": 0.37639445066452026,
+      "learning_rate": 4.969963237691547e-06,
+      "loss": 0.6544,
+      "step": 360
+    },
+    {
+      "epoch": 0.40380313199105144,
+      "grad_norm": 0.38033226132392883,
+      "learning_rate": 4.9697322095992075e-06,
+      "loss": 0.6216,
+      "step": 361
+    },
+    {
+      "epoch": 0.40492170022371365,
+      "grad_norm": 0.3785533010959625,
+      "learning_rate": 4.969500301840805e-06,
+      "loss": 0.6379,
+      "step": 362
+    },
+    {
+      "epoch": 0.40604026845637586,
+      "grad_norm": 0.36831173300743103,
+      "learning_rate": 4.969267514498942e-06,
+      "loss": 0.6305,
+      "step": 363
+    },
+    {
+      "epoch": 0.407158836689038,
+      "grad_norm": 0.3860856592655182,
+      "learning_rate": 4.969033847656531e-06,
+      "loss": 0.6428,
+      "step": 364
+    },
+    {
+      "epoch": 0.4082774049217002,
+      "grad_norm": 0.37992534041404724,
+      "learning_rate": 4.9687993013968e-06,
+      "loss": 0.629,
+      "step": 365
+    },
+    {
+      "epoch": 0.40939597315436244,
+      "grad_norm": 0.37179872393608093,
+      "learning_rate": 4.9685638758032885e-06,
+      "loss": 0.6146,
+      "step": 366
+    },
+    {
+      "epoch": 0.4105145413870246,
+      "grad_norm": 0.3762771487236023,
+      "learning_rate": 4.96832757095985e-06,
+      "loss": 0.6294,
+      "step": 367
+    },
+    {
+      "epoch": 0.4116331096196868,
+      "grad_norm": 0.37078356742858887,
+      "learning_rate": 4.968090386950653e-06,
+      "loss": 0.6438,
+      "step": 368
+    },
+    {
+      "epoch": 0.412751677852349,
+      "grad_norm": 0.3619535565376282,
+      "learning_rate": 4.967852323860176e-06,
+      "loss": 0.6229,
+      "step": 369
+    },
+    {
+      "epoch": 0.41387024608501116,
+      "grad_norm": 0.3610592782497406,
+      "learning_rate": 4.967613381773211e-06,
+      "loss": 0.6332,
+      "step": 370
+    },
+    {
+      "epoch": 0.41498881431767337,
+      "grad_norm": 0.36372244358062744,
+      "learning_rate": 4.9673735607748665e-06,
+      "loss": 0.6379,
+      "step": 371
+    },
+    {
+      "epoch": 0.4161073825503356,
+      "grad_norm": 0.3713506758213043,
+      "learning_rate": 4.96713286095056e-06,
+      "loss": 0.6051,
+      "step": 372
+    },
+    {
+      "epoch": 0.4172259507829978,
+      "grad_norm": 0.37290191650390625,
+      "learning_rate": 4.9668912823860244e-06,
+      "loss": 0.6431,
+      "step": 373
+    },
+    {
+      "epoch": 0.41834451901565994,
+      "grad_norm": 0.3736407458782196,
+      "learning_rate": 4.966648825167305e-06,
+      "loss": 0.6296,
+      "step": 374
+    },
+    {
+      "epoch": 0.41946308724832215,
+      "grad_norm": 0.38261404633522034,
+      "learning_rate": 4.9664054893807586e-06,
+      "loss": 0.6559,
+      "step": 375
+    },
+    {
+      "epoch": 0.42058165548098436,
+      "grad_norm": 0.36865612864494324,
+      "learning_rate": 4.966161275113057e-06,
+      "loss": 0.6372,
+      "step": 376
+    },
+    {
+      "epoch": 0.4217002237136465,
+      "grad_norm": 0.3745094835758209,
+      "learning_rate": 4.965916182451185e-06,
+      "loss": 0.6526,
+      "step": 377
+    },
+    {
+      "epoch": 0.4228187919463087,
+      "grad_norm": 0.3758225440979004,
+      "learning_rate": 4.965670211482437e-06,
+      "loss": 0.6423,
+      "step": 378
+    },
+    {
+      "epoch": 0.42393736017897093,
+      "grad_norm": 0.37716934084892273,
+      "learning_rate": 4.965423362294426e-06,
+      "loss": 0.6431,
+      "step": 379
+    },
+    {
+      "epoch": 0.4250559284116331,
+      "grad_norm": 0.3757461905479431,
+      "learning_rate": 4.965175634975072e-06,
+      "loss": 0.6335,
+      "step": 380
+    },
+    {
+      "epoch": 0.4261744966442953,
+      "grad_norm": 0.3701077401638031,
+      "learning_rate": 4.964927029612611e-06,
+      "loss": 0.6182,
+      "step": 381
+    },
+    {
+      "epoch": 0.4272930648769575,
+      "grad_norm": 0.38263121247291565,
+      "learning_rate": 4.96467754629559e-06,
+      "loss": 0.6371,
+      "step": 382
+    },
+    {
+      "epoch": 0.42841163310961966,
+      "grad_norm": 0.3740926682949066,
+      "learning_rate": 4.9644271851128715e-06,
+      "loss": 0.6272,
+      "step": 383
+    },
+    {
+      "epoch": 0.42953020134228187,
+      "grad_norm": 0.39056089520454407,
+      "learning_rate": 4.964175946153627e-06,
+      "loss": 0.624,
+      "step": 384
+    },
+    {
+      "epoch": 0.4306487695749441,
+      "grad_norm": 0.3867873549461365,
+      "learning_rate": 4.963923829507343e-06,
+      "loss": 0.6714,
+      "step": 385
+    },
+    {
+      "epoch": 0.4317673378076063,
+      "grad_norm": 0.3808860182762146,
+      "learning_rate": 4.963670835263819e-06,
+      "loss": 0.6412,
+      "step": 386
+    },
+    {
+      "epoch": 0.43288590604026844,
+      "grad_norm": 0.3839844763278961,
+      "learning_rate": 4.963416963513166e-06,
+      "loss": 0.6288,
+      "step": 387
+    },
+    {
+      "epoch": 0.43400447427293065,
+      "grad_norm": 0.37187114357948303,
+      "learning_rate": 4.963162214345806e-06,
+      "loss": 0.6307,
+      "step": 388
+    },
+    {
+      "epoch": 0.43512304250559286,
+      "grad_norm": 0.36723873019218445,
+      "learning_rate": 4.962906587852477e-06,
+      "loss": 0.6285,
+      "step": 389
+    },
+    {
+      "epoch": 0.436241610738255,
+      "grad_norm": 0.3754160404205322,
+      "learning_rate": 4.962650084124226e-06,
+      "loss": 0.6227,
+      "step": 390
+    },
+    {
+      "epoch": 0.4373601789709172,
+      "grad_norm": 0.374239981174469,
+      "learning_rate": 4.962392703252417e-06,
+      "loss": 0.6612,
+      "step": 391
+    },
+    {
+      "epoch": 0.43847874720357943,
+      "grad_norm": 0.3758632242679596,
+      "learning_rate": 4.9621344453287214e-06,
+      "loss": 0.6408,
+      "step": 392
+    },
+    {
+      "epoch": 0.4395973154362416,
+      "grad_norm": 0.3839190602302551,
+      "learning_rate": 4.9618753104451254e-06,
+      "loss": 0.6524,
+      "step": 393
+    },
+    {
+      "epoch": 0.4407158836689038,
+      "grad_norm": 0.36766374111175537,
+      "learning_rate": 4.961615298693928e-06,
+      "loss": 0.6232,
+      "step": 394
+    },
+    {
+      "epoch": 0.441834451901566,
+      "grad_norm": 0.3692423701286316,
+      "learning_rate": 4.961354410167739e-06,
+      "loss": 0.6436,
+      "step": 395
+    },
+    {
+      "epoch": 0.4429530201342282,
+      "grad_norm": 0.3720521926879883,
+      "learning_rate": 4.961092644959482e-06,
+      "loss": 0.6346,
+      "step": 396
+    },
+    {
+      "epoch": 0.44407158836689037,
+      "grad_norm": 0.373910129070282,
+      "learning_rate": 4.960830003162392e-06,
+      "loss": 0.6211,
+      "step": 397
+    },
+    {
+      "epoch": 0.4451901565995526,
+      "grad_norm": 0.37455853819847107,
+      "learning_rate": 4.960566484870017e-06,
+      "loss": 0.6366,
+      "step": 398
+    },
+    {
+      "epoch": 0.4463087248322148,
+      "grad_norm": 0.387390673160553,
+      "learning_rate": 4.960302090176215e-06,
+      "loss": 0.6543,
+      "step": 399
+    },
+    {
+      "epoch": 0.44742729306487694,
+      "grad_norm": 0.3862502872943878,
+      "learning_rate": 4.960036819175159e-06,
+      "loss": 0.6351,
+      "step": 400
+    },
+    {
+      "epoch": 0.44854586129753915,
+      "grad_norm": 0.38686901330947876,
+      "learning_rate": 4.959770671961334e-06,
+      "loss": 0.6247,
+      "step": 401
+    },
+    {
+      "epoch": 0.44966442953020136,
+      "grad_norm": 0.38111770153045654,
+      "learning_rate": 4.959503648629534e-06,
+      "loss": 0.6624,
+      "step": 402
+    },
+    {
+      "epoch": 0.4507829977628635,
+      "grad_norm": 0.3962753713130951,
+      "learning_rate": 4.959235749274866e-06,
+      "loss": 0.6224,
+      "step": 403
+    },
+    {
+      "epoch": 0.4519015659955257,
+      "grad_norm": 0.36403393745422363,
+      "learning_rate": 4.958966973992754e-06,
+      "loss": 0.6215,
+      "step": 404
+    },
+    {
+      "epoch": 0.45302013422818793,
+      "grad_norm": 0.3858584463596344,
+      "learning_rate": 4.958697322878926e-06,
+      "loss": 0.6473,
+      "step": 405
+    },
+    {
+      "epoch": 0.4541387024608501,
+      "grad_norm": 0.39325979351997375,
+      "learning_rate": 4.958426796029429e-06,
+      "loss": 0.6664,
+      "step": 406
+    },
+    {
+      "epoch": 0.4552572706935123,
+      "grad_norm": 0.37423112988471985,
+      "learning_rate": 4.958155393540618e-06,
+      "loss": 0.6416,
+      "step": 407
+    },
+    {
+      "epoch": 0.4563758389261745,
+      "grad_norm": 0.3979191482067108,
+      "learning_rate": 4.9578831155091585e-06,
+      "loss": 0.6493,
+      "step": 408
+    },
+    {
+      "epoch": 0.4574944071588367,
+      "grad_norm": 0.375473290681839,
+      "learning_rate": 4.957609962032034e-06,
+      "loss": 0.6246,
+      "step": 409
+    },
+    {
+      "epoch": 0.45861297539149887,
+      "grad_norm": 0.37951260805130005,
+      "learning_rate": 4.957335933206533e-06,
+      "loss": 0.6374,
+      "step": 410
+    },
+    {
+      "epoch": 0.4597315436241611,
+      "grad_norm": 0.384162575006485,
+      "learning_rate": 4.9570610291302605e-06,
+      "loss": 0.6411,
+      "step": 411
+    },
+    {
+      "epoch": 0.4608501118568233,
+      "grad_norm": 0.37713801860809326,
+      "learning_rate": 4.95678524990113e-06,
+      "loss": 0.6384,
+      "step": 412
+    },
+    {
+      "epoch": 0.46196868008948544,
+      "grad_norm": 0.3779420554637909,
+      "learning_rate": 4.95650859561737e-06,
+      "loss": 0.6238,
+      "step": 413
+    },
+    {
+      "epoch": 0.46308724832214765,
+      "grad_norm": 0.3826324939727783,
+      "learning_rate": 4.956231066377517e-06,
+      "loss": 0.6373,
+      "step": 414
+    },
+    {
+      "epoch": 0.46420581655480986,
+      "grad_norm": 0.3693124055862427,
+      "learning_rate": 4.955952662280422e-06,
+      "loss": 0.6264,
+      "step": 415
+    },
+    {
+      "epoch": 0.465324384787472,
+      "grad_norm": 0.3891177177429199,
+      "learning_rate": 4.9556733834252465e-06,
+      "loss": 0.6755,
+      "step": 416
+    },
+    {
+      "epoch": 0.4664429530201342,
+      "grad_norm": 0.3732079863548279,
+      "learning_rate": 4.955393229911465e-06,
+      "loss": 0.6163,
+      "step": 417
+    },
+    {
+      "epoch": 0.46756152125279643,
+      "grad_norm": 0.39267081022262573,
+      "learning_rate": 4.955112201838859e-06,
+      "loss": 0.653,
+      "step": 418
+    },
+    {
+      "epoch": 0.46868008948545864,
+      "grad_norm": 0.37127041816711426,
+      "learning_rate": 4.9548302993075275e-06,
+      "loss": 0.6024,
+      "step": 419
+    },
+    {
+      "epoch": 0.4697986577181208,
+      "grad_norm": 0.38274380564689636,
+      "learning_rate": 4.954547522417878e-06,
+      "loss": 0.6103,
+      "step": 420
+    },
+    {
+      "epoch": 0.470917225950783,
+      "grad_norm": 0.39440205693244934,
+      "learning_rate": 4.954263871270627e-06,
+      "loss": 0.6388,
+      "step": 421
+    },
+    {
+      "epoch": 0.4720357941834452,
+      "grad_norm": 0.38207298517227173,
+      "learning_rate": 4.953979345966808e-06,
+      "loss": 0.6157,
+      "step": 422
+    },
+    {
+      "epoch": 0.47315436241610737,
+      "grad_norm": 0.37390536069869995,
+      "learning_rate": 4.953693946607762e-06,
+      "loss": 0.612,
+      "step": 423
+    },
+    {
+      "epoch": 0.4742729306487696,
+      "grad_norm": 0.3679952621459961,
+      "learning_rate": 4.953407673295141e-06,
+      "loss": 0.5962,
+      "step": 424
+    },
+    {
+      "epoch": 0.4753914988814318,
+      "grad_norm": 0.36741313338279724,
+      "learning_rate": 4.953120526130911e-06,
+      "loss": 0.5802,
+      "step": 425
+    },
+    {
+      "epoch": 0.47651006711409394,
+      "grad_norm": 0.40101951360702515,
+      "learning_rate": 4.952832505217347e-06,
+      "loss": 0.631,
+      "step": 426
+    },
+    {
+      "epoch": 0.47762863534675615,
+      "grad_norm": 0.37646785378456116,
+      "learning_rate": 4.952543610657036e-06,
+      "loss": 0.6192,
+      "step": 427
+    },
+    {
+      "epoch": 0.47874720357941836,
+      "grad_norm": 0.3909439444541931,
+      "learning_rate": 4.952253842552876e-06,
+      "loss": 0.6288,
+      "step": 428
+    },
+    {
+      "epoch": 0.4798657718120805,
+      "grad_norm": 0.379685640335083,
+      "learning_rate": 4.9519632010080765e-06,
+      "loss": 0.6296,
+      "step": 429
+    },
+    {
+      "epoch": 0.4809843400447427,
+      "grad_norm": 0.3872782588005066,
+      "learning_rate": 4.9516716861261575e-06,
+      "loss": 0.6307,
+      "step": 430
+    },
+    {
+      "epoch": 0.48210290827740493,
+      "grad_norm": 0.4066009223461151,
+      "learning_rate": 4.951379298010951e-06,
+      "loss": 0.6454,
+      "step": 431
+    },
+    {
+      "epoch": 0.48322147651006714,
+      "grad_norm": 0.38412487506866455,
+      "learning_rate": 4.951086036766599e-06,
+      "loss": 0.6254,
+      "step": 432
+    },
+    {
+      "epoch": 0.4843400447427293,
+      "grad_norm": 0.37819865345954895,
+      "learning_rate": 4.9507919024975545e-06,
+      "loss": 0.629,
+      "step": 433
+    },
+    {
+      "epoch": 0.4854586129753915,
+      "grad_norm": 0.38674691319465637,
+      "learning_rate": 4.950496895308582e-06,
+      "loss": 0.6357,
+      "step": 434
+    },
+    {
+      "epoch": 0.4865771812080537,
+      "grad_norm": 0.39304593205451965,
+      "learning_rate": 4.950201015304758e-06,
+      "loss": 0.6475,
+      "step": 435
+    },
+    {
+      "epoch": 0.48769574944071586,
+      "grad_norm": 0.381124347448349,
+      "learning_rate": 4.949904262591467e-06,
+      "loss": 0.6523,
+      "step": 436
+    },
+    {
+      "epoch": 0.4888143176733781,
+      "grad_norm": 0.4084749221801758,
+      "learning_rate": 4.949606637274408e-06,
+      "loss": 0.6773,
+      "step": 437
+    },
+    {
+      "epoch": 0.4899328859060403,
+      "grad_norm": 0.3967250883579254,
+      "learning_rate": 4.949308139459586e-06,
+      "loss": 0.6263,
+      "step": 438
+    },
+    {
+      "epoch": 0.49105145413870244,
+      "grad_norm": 0.39761948585510254,
+      "learning_rate": 4.949008769253322e-06,
+      "loss": 0.6273,
+      "step": 439
+    },
+    {
+      "epoch": 0.49217002237136465,
+      "grad_norm": 0.3865715265274048,
+      "learning_rate": 4.948708526762244e-06,
+      "loss": 0.6464,
+      "step": 440
+    },
+    {
+      "epoch": 0.49328859060402686,
+      "grad_norm": 0.3970697820186615,
+      "learning_rate": 4.948407412093292e-06,
+      "loss": 0.6229,
+      "step": 441
+    },
+    {
+      "epoch": 0.49440715883668906,
+      "grad_norm": 0.3817065954208374,
+      "learning_rate": 4.948105425353718e-06,
+      "loss": 0.6375,
+      "step": 442
+    },
+    {
+      "epoch": 0.4955257270693512,
+      "grad_norm": 0.3877985179424286,
+      "learning_rate": 4.947802566651082e-06,
+      "loss": 0.6389,
+      "step": 443
+    },
+    {
+      "epoch": 0.4966442953020134,
+      "grad_norm": 0.40800127387046814,
+      "learning_rate": 4.947498836093257e-06,
+      "loss": 0.6627,
+      "step": 444
+    },
+    {
+      "epoch": 0.49776286353467564,
+      "grad_norm": 0.40732380747795105,
+      "learning_rate": 4.947194233788423e-06,
+      "loss": 0.6156,
+      "step": 445
+    },
+    {
+      "epoch": 0.4988814317673378,
+      "grad_norm": 0.3948177695274353,
+      "learning_rate": 4.946888759845074e-06,
+      "loss": 0.6481,
+      "step": 446
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.38517189025878906,
+      "learning_rate": 4.9465824143720145e-06,
+      "loss": 0.6224,
+      "step": 447
+    },
+    {
+      "epoch": 0.5011185682326622,
+      "grad_norm": 0.3713424503803253,
+      "learning_rate": 4.946275197478358e-06,
+      "loss": 0.626,
+      "step": 448
+    },
+    {
+      "epoch": 0.5022371364653244,
+      "grad_norm": 0.4172223210334778,
+      "learning_rate": 4.945967109273527e-06,
+      "loss": 0.6405,
+      "step": 449
+    },
+    {
+      "epoch": 0.5033557046979866,
+      "grad_norm": 0.4550599157810211,
+      "learning_rate": 4.945658149867257e-06,
+      "loss": 0.6103,
+      "step": 450
+    },
+    {
+      "epoch": 0.5044742729306487,
+      "grad_norm": 0.3938581347465515,
+      "learning_rate": 4.945348319369593e-06,
+      "loss": 0.6304,
+      "step": 451
+    },
+    {
+      "epoch": 0.5055928411633109,
+      "grad_norm": 0.3923262059688568,
+      "learning_rate": 4.94503761789089e-06,
+      "loss": 0.6603,
+      "step": 452
+    },
+    {
+      "epoch": 0.5067114093959731,
+      "grad_norm": 0.3978983163833618,
+      "learning_rate": 4.944726045541814e-06,
+      "loss": 0.6445,
+      "step": 453
+    },
+    {
+      "epoch": 0.5078299776286354,
+      "grad_norm": 0.4101882576942444,
+      "learning_rate": 4.9444136024333374e-06,
+      "loss": 0.6223,
+      "step": 454
+    },
+    {
+      "epoch": 0.5089485458612976,
+      "grad_norm": 0.4056575298309326,
+      "learning_rate": 4.944100288676749e-06,
+      "loss": 0.6343,
+      "step": 455
+    },
+    {
+      "epoch": 0.5100671140939598,
+      "grad_norm": 0.39720436930656433,
+      "learning_rate": 4.943786104383644e-06,
+      "loss": 0.6246,
+      "step": 456
+    },
+    {
+      "epoch": 0.5111856823266219,
+      "grad_norm": 0.3909725248813629,
+      "learning_rate": 4.943471049665925e-06,
+      "loss": 0.6339,
+      "step": 457
+    },
+    {
+      "epoch": 0.5123042505592841,
+      "grad_norm": 0.3773731291294098,
+      "learning_rate": 4.943155124635812e-06,
+      "loss": 0.6215,
+      "step": 458
+    },
+    {
+      "epoch": 0.5134228187919463,
+      "grad_norm": 0.4020001292228699,
+      "learning_rate": 4.9428383294058295e-06,
+      "loss": 0.6269,
+      "step": 459
+    },
+    {
+      "epoch": 0.5145413870246085,
+      "grad_norm": 0.3916706144809723,
+      "learning_rate": 4.942520664088812e-06,
+      "loss": 0.6233,
+      "step": 460
+    },
+    {
+      "epoch": 0.5156599552572707,
+      "grad_norm": 0.38717713952064514,
+      "learning_rate": 4.9422021287979076e-06,
+      "loss": 0.6216,
+      "step": 461
+    },
+    {
+      "epoch": 0.5167785234899329,
+      "grad_norm": 0.38645485043525696,
+      "learning_rate": 4.941882723646568e-06,
+      "loss": 0.6092,
+      "step": 462
+    },
+    {
+      "epoch": 0.5178970917225951,
+      "grad_norm": 0.38496407866477966,
+      "learning_rate": 4.9415624487485615e-06,
+      "loss": 0.6368,
+      "step": 463
+    },
+    {
+      "epoch": 0.5190156599552572,
+      "grad_norm": 0.3946744501590729,
+      "learning_rate": 4.941241304217962e-06,
+      "loss": 0.6525,
+      "step": 464
+    },
+    {
+      "epoch": 0.5201342281879194,
+      "grad_norm": 0.3994438648223877,
+      "learning_rate": 4.940919290169155e-06,
+      "loss": 0.6314,
+      "step": 465
+    },
+    {
+      "epoch": 0.5212527964205816,
+      "grad_norm": 0.3929794728755951,
+      "learning_rate": 4.940596406716834e-06,
+      "loss": 0.6148,
+      "step": 466
+    },
+    {
+      "epoch": 0.5223713646532439,
+      "grad_norm": 0.42620542645454407,
+      "learning_rate": 4.940272653976005e-06,
+      "loss": 0.6468,
+      "step": 467
+    },
+    {
+      "epoch": 0.5234899328859061,
+      "grad_norm": 0.4014374613761902,
+      "learning_rate": 4.9399480320619805e-06,
+      "loss": 0.6451,
+      "step": 468
+    },
+    {
+      "epoch": 0.5246085011185683,
+      "grad_norm": 0.39342424273490906,
+      "learning_rate": 4.939622541090384e-06,
+      "loss": 0.6696,
+      "step": 469
+    },
+    {
+      "epoch": 0.5257270693512305,
+      "grad_norm": 0.3870956301689148,
+      "learning_rate": 4.939296181177149e-06,
+      "loss": 0.6451,
+      "step": 470
+    },
+    {
+      "epoch": 0.5268456375838926,
+      "grad_norm": 0.39973214268684387,
+      "learning_rate": 4.938968952438518e-06,
+      "loss": 0.6254,
+      "step": 471
+    },
+    {
+      "epoch": 0.5279642058165548,
+      "grad_norm": 0.3956799805164337,
+      "learning_rate": 4.938640854991041e-06,
+      "loss": 0.6169,
+      "step": 472
+    },
+    {
+      "epoch": 0.529082774049217,
+      "grad_norm": 0.38881829380989075,
+      "learning_rate": 4.938311888951583e-06,
+      "loss": 0.5989,
+      "step": 473
+    },
+    {
+      "epoch": 0.5302013422818792,
+      "grad_norm": 0.392107218503952,
+      "learning_rate": 4.93798205443731e-06,
+      "loss": 0.6284,
+      "step": 474
+    },
+    {
+      "epoch": 0.5313199105145414,
+      "grad_norm": 0.4042797088623047,
+      "learning_rate": 4.937651351565707e-06,
+      "loss": 0.6235,
+      "step": 475
+    },
+    {
+      "epoch": 0.5324384787472036,
+      "grad_norm": 0.380206435918808,
+      "learning_rate": 4.937319780454559e-06,
+      "loss": 0.5894,
+      "step": 476
+    },
+    {
+      "epoch": 0.5335570469798657,
+      "grad_norm": 0.3989536166191101,
+      "learning_rate": 4.936987341221968e-06,
+      "loss": 0.6522,
+      "step": 477
+    },
+    {
+      "epoch": 0.5346756152125279,
+      "grad_norm": 0.38699498772621155,
+      "learning_rate": 4.9366540339863395e-06,
+      "loss": 0.6202,
+      "step": 478
+    },
+    {
+      "epoch": 0.5357941834451901,
+      "grad_norm": 0.4171985387802124,
+      "learning_rate": 4.936319858866391e-06,
+      "loss": 0.624,
+      "step": 479
+    },
+    {
+      "epoch": 0.5369127516778524,
+      "grad_norm": 0.3932148218154907,
+      "learning_rate": 4.93598481598115e-06,
+      "loss": 0.6215,
+      "step": 480
+    },
+    {
+      "epoch": 0.5380313199105146,
+      "grad_norm": 0.3934101462364197,
+      "learning_rate": 4.935648905449949e-06,
+      "loss": 0.6402,
+      "step": 481
+    },
+    {
+      "epoch": 0.5391498881431768,
+      "grad_norm": 0.3917444348335266,
+      "learning_rate": 4.935312127392434e-06,
+      "loss": 0.641,
+      "step": 482
+    },
+    {
+      "epoch": 0.540268456375839,
+      "grad_norm": 0.4036387503147125,
+      "learning_rate": 4.9349744819285584e-06,
+      "loss": 0.6405,
+      "step": 483
+    },
+    {
+      "epoch": 0.5413870246085011,
+      "grad_norm": 0.38611260056495667,
+      "learning_rate": 4.934635969178584e-06,
+      "loss": 0.6231,
+      "step": 484
+    },
+    {
+      "epoch": 0.5425055928411633,
+      "grad_norm": 0.39185649156570435,
+      "learning_rate": 4.9342965892630805e-06,
+      "loss": 0.6214,
+      "step": 485
+    },
+    {
+      "epoch": 0.5436241610738255,
+      "grad_norm": 0.3736090362071991,
+      "learning_rate": 4.933956342302929e-06,
+      "loss": 0.6053,
+      "step": 486
+    },
+    {
+      "epoch": 0.5447427293064877,
+      "grad_norm": 0.39649662375450134,
+      "learning_rate": 4.93361522841932e-06,
+      "loss": 0.6408,
+      "step": 487
+    },
+    {
+      "epoch": 0.5458612975391499,
+      "grad_norm": 0.3990592658519745,
+      "learning_rate": 4.933273247733746e-06,
+      "loss": 0.6081,
+      "step": 488
+    },
+    {
+      "epoch": 0.5469798657718121,
+      "grad_norm": 0.39177680015563965,
+      "learning_rate": 4.932930400368019e-06,
+      "loss": 0.6114,
+      "step": 489
+    },
+    {
+      "epoch": 0.5480984340044742,
+      "grad_norm": 0.3953116536140442,
+      "learning_rate": 4.9325866864442495e-06,
+      "loss": 0.6339,
+      "step": 490
+    },
+    {
+      "epoch": 0.5492170022371364,
+      "grad_norm": 0.38563409447669983,
+      "learning_rate": 4.932242106084864e-06,
+      "loss": 0.6331,
+      "step": 491
+    },
+    {
+      "epoch": 0.5503355704697986,
+      "grad_norm": 0.40618443489074707,
+      "learning_rate": 4.931896659412593e-06,
+      "loss": 0.6441,
+      "step": 492
+    },
+    {
+      "epoch": 0.5514541387024608,
+      "grad_norm": 0.4008066654205322,
+      "learning_rate": 4.931550346550479e-06,
+      "loss": 0.6243,
+      "step": 493
+    },
+    {
+      "epoch": 0.5525727069351231,
+      "grad_norm": 0.39776620268821716,
+      "learning_rate": 4.931203167621868e-06,
+      "loss": 0.6152,
+      "step": 494
+    },
+    {
+      "epoch": 0.5536912751677853,
+      "grad_norm": 0.38687410950660706,
+      "learning_rate": 4.930855122750421e-06,
+      "loss": 0.5969,
+      "step": 495
+    },
+    {
+      "epoch": 0.5548098434004475,
+      "grad_norm": 0.3877246081829071,
+      "learning_rate": 4.9305062120601035e-06,
+      "loss": 0.6016,
+      "step": 496
+    },
+    {
+      "epoch": 0.5559284116331096,
+      "grad_norm": 0.40948086977005005,
+      "learning_rate": 4.930156435675189e-06,
+      "loss": 0.6168,
+      "step": 497
+    },
+    {
+      "epoch": 0.5570469798657718,
+      "grad_norm": 0.404021292924881,
+      "learning_rate": 4.929805793720262e-06,
+      "loss": 0.6092,
+      "step": 498
+    },
+    {
+      "epoch": 0.558165548098434,
+      "grad_norm": 0.39509114623069763,
+      "learning_rate": 4.929454286320211e-06,
+      "loss": 0.6346,
+      "step": 499
+    },
+    {
+      "epoch": 0.5592841163310962,
+      "grad_norm": 0.39687812328338623,
+      "learning_rate": 4.9291019136002385e-06,
+      "loss": 0.639,
+      "step": 500
+    },
+    {
+      "epoch": 0.5604026845637584,
+      "grad_norm": 0.39210405945777893,
+      "learning_rate": 4.92874867568585e-06,
+      "loss": 0.6083,
+      "step": 501
+    },
+    {
+      "epoch": 0.5615212527964206,
+      "grad_norm": 0.4022452235221863,
+      "learning_rate": 4.928394572702862e-06,
+      "loss": 0.6252,
+      "step": 502
+    },
+    {
+      "epoch": 0.5626398210290827,
+      "grad_norm": 0.40317103266716003,
+      "learning_rate": 4.928039604777399e-06,
+      "loss": 0.614,
+      "step": 503
+    },
+    {
+      "epoch": 0.5637583892617449,
+      "grad_norm": 0.4097250998020172,
+      "learning_rate": 4.9276837720358924e-06,
+      "loss": 0.6218,
+      "step": 504
+    },
+    {
+      "epoch": 0.5648769574944071,
+      "grad_norm": 0.3927348554134369,
+      "learning_rate": 4.927327074605083e-06,
+      "loss": 0.6079,
+      "step": 505
+    },
+    {
+      "epoch": 0.5659955257270693,
+      "grad_norm": 0.3961605131626129,
+      "learning_rate": 4.9269695126120185e-06,
+      "loss": 0.612,
+      "step": 506
+    },
+    {
+      "epoch": 0.5671140939597316,
+      "grad_norm": 0.3945924639701843,
+      "learning_rate": 4.926611086184054e-06,
+      "loss": 0.6268,
+      "step": 507
+    },
+    {
+      "epoch": 0.5682326621923938,
+      "grad_norm": 0.39072591066360474,
+      "learning_rate": 4.926251795448854e-06,
+      "loss": 0.6176,
+      "step": 508
+    },
+    {
+      "epoch": 0.569351230425056,
+      "grad_norm": 0.39760643243789673,
+      "learning_rate": 4.9258916405343904e-06,
+      "loss": 0.6437,
+      "step": 509
+    },
+    {
+      "epoch": 0.5704697986577181,
+      "grad_norm": 0.39866000413894653,
+      "learning_rate": 4.925530621568942e-06,
+      "loss": 0.6383,
+      "step": 510
+    },
+    {
+      "epoch": 0.5715883668903803,
+      "grad_norm": 0.3932257294654846,
+      "learning_rate": 4.925168738681097e-06,
+      "loss": 0.6156,
+      "step": 511
+    },
+    {
+      "epoch": 0.5727069351230425,
+      "grad_norm": 0.39929500222206116,
+      "learning_rate": 4.924805991999751e-06,
+      "loss": 0.6069,
+      "step": 512
+    },
+    {
+      "epoch": 0.5738255033557047,
+      "grad_norm": 0.41192054748535156,
+      "learning_rate": 4.924442381654105e-06,
+      "loss": 0.6451,
+      "step": 513
+    },
+    {
+      "epoch": 0.5749440715883669,
+      "grad_norm": 0.41147273778915405,
+      "learning_rate": 4.92407790777367e-06,
+      "loss": 0.647,
+      "step": 514
+    },
+    {
+      "epoch": 0.5760626398210291,
+      "grad_norm": 0.4128178358078003,
+      "learning_rate": 4.923712570488264e-06,
+      "loss": 0.5909,
+      "step": 515
+    },
+    {
+      "epoch": 0.5771812080536913,
+      "grad_norm": 0.4081036150455475,
+      "learning_rate": 4.923346369928012e-06,
+      "loss": 0.6248,
+      "step": 516
+    },
+    {
+      "epoch": 0.5782997762863534,
+      "grad_norm": 0.3965778350830078,
+      "learning_rate": 4.922979306223347e-06,
+      "loss": 0.6019,
+      "step": 517
+    },
+    {
+      "epoch": 0.5794183445190156,
+      "grad_norm": 0.3979526162147522,
+      "learning_rate": 4.922611379505009e-06,
+      "loss": 0.6368,
+      "step": 518
+    },
+    {
+      "epoch": 0.5805369127516778,
+      "grad_norm": 0.38306665420532227,
+      "learning_rate": 4.922242589904046e-06,
+      "loss": 0.62,
+      "step": 519
+    },
+    {
+      "epoch": 0.5816554809843401,
+      "grad_norm": 0.3833399713039398,
+      "learning_rate": 4.921872937551814e-06,
+      "loss": 0.6064,
+      "step": 520
+    },
+    {
+      "epoch": 0.5827740492170023,
+      "grad_norm": 0.39361608028411865,
+      "learning_rate": 4.921502422579973e-06,
+      "loss": 0.6236,
+      "step": 521
+    },
+    {
+      "epoch": 0.5838926174496645,
+      "grad_norm": 0.39250272512435913,
+      "learning_rate": 4.921131045120494e-06,
+      "loss": 0.624,
+      "step": 522
+    },
+    {
+      "epoch": 0.5850111856823266,
+      "grad_norm": 0.40747684240341187,
+      "learning_rate": 4.920758805305654e-06,
+      "loss": 0.6096,
+      "step": 523
+    },
+    {
+      "epoch": 0.5861297539149888,
+      "grad_norm": 0.39987003803253174,
+      "learning_rate": 4.920385703268037e-06,
+      "loss": 0.6282,
+      "step": 524
+    },
+    {
+      "epoch": 0.587248322147651,
+      "grad_norm": 0.39122274518013,
+      "learning_rate": 4.920011739140532e-06,
+      "loss": 0.6479,
+      "step": 525
+    },
+    {
+      "epoch": 0.5883668903803132,
+      "grad_norm": 0.39809542894363403,
+      "learning_rate": 4.919636913056339e-06,
+      "loss": 0.6213,
+      "step": 526
+    },
+    {
+      "epoch": 0.5894854586129754,
+      "grad_norm": 0.39921343326568604,
+      "learning_rate": 4.919261225148963e-06,
+      "loss": 0.6118,
+      "step": 527
+    },
+    {
+      "epoch": 0.5906040268456376,
+      "grad_norm": 0.4086368680000305,
+      "learning_rate": 4.9188846755522155e-06,
+      "loss": 0.6214,
+      "step": 528
+    },
+    {
+      "epoch": 0.5917225950782998,
+      "grad_norm": 0.4066048264503479,
+      "learning_rate": 4.918507264400216e-06,
+      "loss": 0.6316,
+      "step": 529
+    },
+    {
+      "epoch": 0.5928411633109619,
+      "grad_norm": 0.41961807012557983,
+      "learning_rate": 4.91812899182739e-06,
+      "loss": 0.5948,
+      "step": 530
+    },
+    {
+      "epoch": 0.5939597315436241,
+      "grad_norm": 0.39992618560791016,
+      "learning_rate": 4.917749857968469e-06,
+      "loss": 0.6113,
+      "step": 531
+    },
+    {
+      "epoch": 0.5950782997762863,
+      "grad_norm": 0.41020235419273376,
+      "learning_rate": 4.917369862958494e-06,
+      "loss": 0.622,
+      "step": 532
+    },
+    {
+      "epoch": 0.5961968680089486,
+      "grad_norm": 0.40504705905914307,
+      "learning_rate": 4.916989006932811e-06,
+      "loss": 0.621,
+      "step": 533
+    },
+    {
+      "epoch": 0.5973154362416108,
+      "grad_norm": 0.3829837441444397,
+      "learning_rate": 4.9166072900270725e-06,
+      "loss": 0.5942,
+      "step": 534
+    },
+    {
+      "epoch": 0.598434004474273,
+      "grad_norm": 0.4082834720611572,
+      "learning_rate": 4.9162247123772375e-06,
+      "loss": 0.5923,
+      "step": 535
+    },
+    {
+      "epoch": 0.5995525727069351,
+      "grad_norm": 0.40038296580314636,
+      "learning_rate": 4.915841274119572e-06,
+      "loss": 0.6057,
+      "step": 536
+    },
+    {
+      "epoch": 0.6006711409395973,
+      "grad_norm": 0.40687569975852966,
+      "learning_rate": 4.91545697539065e-06,
+      "loss": 0.6343,
+      "step": 537
+    },
+    {
+      "epoch": 0.6017897091722595,
+      "grad_norm": 0.386262983083725,
+      "learning_rate": 4.9150718163273494e-06,
+      "loss": 0.6372,
+      "step": 538
+    },
+    {
+      "epoch": 0.6029082774049217,
+      "grad_norm": 0.39570850133895874,
+      "learning_rate": 4.914685797066855e-06,
+      "loss": 0.6157,
+      "step": 539
+    },
+    {
+      "epoch": 0.6040268456375839,
+      "grad_norm": 0.40055716037750244,
+      "learning_rate": 4.9142989177466594e-06,
+      "loss": 0.6141,
+      "step": 540
+    },
+    {
+      "epoch": 0.6051454138702461,
+      "grad_norm": 0.4038466811180115,
+      "learning_rate": 4.913911178504562e-06,
+      "loss": 0.6286,
+      "step": 541
+    },
+    {
+      "epoch": 0.6062639821029083,
+      "grad_norm": 0.38774847984313965,
+      "learning_rate": 4.913522579478664e-06,
+      "loss": 0.6343,
+      "step": 542
+    },
+    {
+      "epoch": 0.6073825503355704,
+      "grad_norm": 0.39426755905151367,
+      "learning_rate": 4.913133120807379e-06,
+      "loss": 0.6121,
+      "step": 543
+    },
+    {
+      "epoch": 0.6085011185682326,
+      "grad_norm": 0.4076898396015167,
+      "learning_rate": 4.912742802629423e-06,
+      "loss": 0.6273,
+      "step": 544
+    },
+    {
+      "epoch": 0.6096196868008948,
+      "grad_norm": 0.3859540820121765,
+      "learning_rate": 4.91235162508382e-06,
+      "loss": 0.6314,
+      "step": 545
+    },
+    {
+      "epoch": 0.610738255033557,
+      "grad_norm": 0.3914327621459961,
+      "learning_rate": 4.911959588309897e-06,
+      "loss": 0.6027,
+      "step": 546
+    },
+    {
+      "epoch": 0.6118568232662193,
+      "grad_norm": 0.3892766833305359,
+      "learning_rate": 4.9115666924472906e-06,
+      "loss": 0.5922,
+      "step": 547
+    },
+    {
+      "epoch": 0.6129753914988815,
+      "grad_norm": 0.3921322226524353,
+      "learning_rate": 4.911172937635942e-06,
+      "loss": 0.6066,
+      "step": 548
+    },
+    {
+      "epoch": 0.6140939597315436,
+      "grad_norm": 0.3972843885421753,
+      "learning_rate": 4.910778324016098e-06,
+      "loss": 0.614,
+      "step": 549
+    },
+    {
+      "epoch": 0.6152125279642058,
+      "grad_norm": 0.4027954638004303,
+      "learning_rate": 4.9103828517283105e-06,
+      "loss": 0.6174,
+      "step": 550
+    },
+    {
+      "epoch": 0.616331096196868,
+      "grad_norm": 0.40479257702827454,
+      "learning_rate": 4.909986520913441e-06,
+      "loss": 0.6114,
+      "step": 551
+    },
+    {
+      "epoch": 0.6174496644295302,
+      "grad_norm": 0.4246085584163666,
+      "learning_rate": 4.909589331712651e-06,
+      "loss": 0.6145,
+      "step": 552
+    },
+    {
+      "epoch": 0.6185682326621924,
+      "grad_norm": 0.4173775017261505,
+      "learning_rate": 4.909191284267413e-06,
+      "loss": 0.6375,
+      "step": 553
+    },
+    {
+      "epoch": 0.6196868008948546,
+      "grad_norm": 0.4135677218437195,
+      "learning_rate": 4.908792378719502e-06,
+      "loss": 0.6444,
+      "step": 554
+    },
+    {
+      "epoch": 0.6208053691275168,
+      "grad_norm": 0.40163061022758484,
+      "learning_rate": 4.9083926152110004e-06,
+      "loss": 0.6128,
+      "step": 555
+    },
+    {
+      "epoch": 0.6219239373601789,
+      "grad_norm": 0.41246625781059265,
+      "learning_rate": 4.907991993884295e-06,
+      "loss": 0.6229,
+      "step": 556
+    },
+    {
+      "epoch": 0.6230425055928411,
+      "grad_norm": 0.4114304780960083,
+      "learning_rate": 4.907590514882079e-06,
+      "loss": 0.6028,
+      "step": 557
+    },
+    {
+      "epoch": 0.6241610738255033,
+      "grad_norm": 0.40224742889404297,
+      "learning_rate": 4.90718817834735e-06,
+      "loss": 0.5896,
+      "step": 558
+    },
+    {
+      "epoch": 0.6252796420581656,
+      "grad_norm": 0.397650808095932,
+      "learning_rate": 4.906784984423411e-06,
+      "loss": 0.6309,
+      "step": 559
+    },
+    {
+      "epoch": 0.6263982102908278,
+      "grad_norm": 0.4087318480014801,
+      "learning_rate": 4.906380933253874e-06,
+      "loss": 0.6002,
+      "step": 560
+    },
+    {
+      "epoch": 0.62751677852349,
+      "grad_norm": 0.3988543450832367,
+      "learning_rate": 4.90597602498265e-06,
+      "loss": 0.6415,
+      "step": 561
+    },
+    {
+      "epoch": 0.6286353467561522,
+      "grad_norm": 0.38457274436950684,
+      "learning_rate": 4.905570259753961e-06,
+      "loss": 0.6105,
+      "step": 562
+    },
+    {
+      "epoch": 0.6297539149888143,
+      "grad_norm": 0.38756313920021057,
+      "learning_rate": 4.905163637712331e-06,
+      "loss": 0.5953,
+      "step": 563
+    },
+    {
+      "epoch": 0.6308724832214765,
+      "grad_norm": 0.4071662127971649,
+      "learning_rate": 4.90475615900259e-06,
+      "loss": 0.635,
+      "step": 564
+    },
+    {
+      "epoch": 0.6319910514541387,
+      "grad_norm": 0.4213521182537079,
+      "learning_rate": 4.904347823769875e-06,
+      "loss": 0.6141,
+      "step": 565
+    },
+    {
+      "epoch": 0.6331096196868009,
+      "grad_norm": 0.4104982018470764,
+      "learning_rate": 4.9039386321596235e-06,
+      "loss": 0.6235,
+      "step": 566
+    },
+    {
+      "epoch": 0.6342281879194631,
+      "grad_norm": 0.41318175196647644,
+      "learning_rate": 4.903528584317583e-06,
+      "loss": 0.6315,
+      "step": 567
+    },
+    {
+      "epoch": 0.6353467561521253,
+      "grad_norm": 0.39332863688468933,
+      "learning_rate": 4.903117680389802e-06,
+      "loss": 0.5807,
+      "step": 568
+    },
+    {
+      "epoch": 0.6364653243847874,
+      "grad_norm": 0.4188497066497803,
+      "learning_rate": 4.902705920522638e-06,
+      "loss": 0.6176,
+      "step": 569
+    },
+    {
+      "epoch": 0.6375838926174496,
+      "grad_norm": 0.41399574279785156,
+      "learning_rate": 4.9022933048627496e-06,
+      "loss": 0.6067,
+      "step": 570
+    },
+    {
+      "epoch": 0.6387024608501118,
+      "grad_norm": 0.4197136461734772,
+      "learning_rate": 4.901879833557102e-06,
+      "loss": 0.6182,
+      "step": 571
+    },
+    {
+      "epoch": 0.639821029082774,
+      "grad_norm": 0.41715213656425476,
+      "learning_rate": 4.9014655067529645e-06,
+      "loss": 0.6088,
+      "step": 572
+    },
+    {
+      "epoch": 0.6409395973154363,
+      "grad_norm": 0.40957003831863403,
+      "learning_rate": 4.901050324597912e-06,
+      "loss": 0.5942,
+      "step": 573
+    },
+    {
+      "epoch": 0.6420581655480985,
+      "grad_norm": 0.4082324206829071,
+      "learning_rate": 4.9006342872398235e-06,
+      "loss": 0.6389,
+      "step": 574
+    },
+    {
+      "epoch": 0.6431767337807607,
+      "grad_norm": 0.41336655616760254,
+      "learning_rate": 4.900217394826882e-06,
+      "loss": 0.6122,
+      "step": 575
+    },
+    {
+      "epoch": 0.6442953020134228,
+      "grad_norm": 0.40878477692604065,
+      "learning_rate": 4.899799647507577e-06,
+      "loss": 0.6372,
+      "step": 576
+    },
+    {
+      "epoch": 0.645413870246085,
+      "grad_norm": 0.4028140604496002,
+      "learning_rate": 4.899381045430701e-06,
+      "loss": 0.5949,
+      "step": 577
+    },
+    {
+      "epoch": 0.6465324384787472,
+      "grad_norm": 0.42413756251335144,
+      "learning_rate": 4.89896158874535e-06,
+      "loss": 0.6217,
+      "step": 578
+    },
+    {
+      "epoch": 0.6476510067114094,
+      "grad_norm": 0.4108542501926422,
+      "learning_rate": 4.898541277600927e-06,
+      "loss": 0.6283,
+      "step": 579
+    },
+    {
+      "epoch": 0.6487695749440716,
+      "grad_norm": 0.4101778566837311,
+      "learning_rate": 4.898120112147135e-06,
+      "loss": 0.6028,
+      "step": 580
+    },
+    {
+      "epoch": 0.6498881431767338,
+      "grad_norm": 0.4104720652103424,
+      "learning_rate": 4.897698092533988e-06,
+      "loss": 0.6481,
+      "step": 581
+    },
+    {
+      "epoch": 0.6510067114093959,
+      "grad_norm": 0.4004007577896118,
+      "learning_rate": 4.897275218911799e-06,
+      "loss": 0.6042,
+      "step": 582
+    },
+    {
+      "epoch": 0.6521252796420581,
+      "grad_norm": 0.4042847752571106,
+      "learning_rate": 4.896851491431185e-06,
+      "loss": 0.6076,
+      "step": 583
+    },
+    {
+      "epoch": 0.6532438478747203,
+      "grad_norm": 0.40685543417930603,
+      "learning_rate": 4.89642691024307e-06,
+      "loss": 0.6066,
+      "step": 584
+    },
+    {
+      "epoch": 0.6543624161073825,
+      "grad_norm": 0.41699886322021484,
+      "learning_rate": 4.896001475498682e-06,
+      "loss": 0.6091,
+      "step": 585
+    },
+    {
+      "epoch": 0.6554809843400448,
+      "grad_norm": 0.3905144929885864,
+      "learning_rate": 4.89557518734955e-06,
+      "loss": 0.6272,
+      "step": 586
+    },
+    {
+      "epoch": 0.656599552572707,
+      "grad_norm": 0.40033814311027527,
+      "learning_rate": 4.895148045947509e-06,
+      "loss": 0.6183,
+      "step": 587
+    },
+    {
+      "epoch": 0.6577181208053692,
+      "grad_norm": 0.39996397495269775,
+      "learning_rate": 4.894720051444698e-06,
+      "loss": 0.5996,
+      "step": 588
+    },
+    {
+      "epoch": 0.6588366890380313,
+      "grad_norm": 0.42592981457710266,
+      "learning_rate": 4.894291203993561e-06,
+      "loss": 0.6506,
+      "step": 589
+    },
+    {
+      "epoch": 0.6599552572706935,
+      "grad_norm": 0.40710797905921936,
+      "learning_rate": 4.8938615037468405e-06,
+      "loss": 0.6044,
+      "step": 590
+    },
+    {
+      "epoch": 0.6610738255033557,
+      "grad_norm": 0.427405446767807,
+      "learning_rate": 4.893430950857591e-06,
+      "loss": 0.6236,
+      "step": 591
+    },
+    {
+      "epoch": 0.6621923937360179,
+      "grad_norm": 0.40190666913986206,
+      "learning_rate": 4.892999545479163e-06,
+      "loss": 0.6031,
+      "step": 592
+    },
+    {
+      "epoch": 0.6633109619686801,
+      "grad_norm": 0.4019568860530853,
+      "learning_rate": 4.8925672877652155e-06,
+      "loss": 0.6232,
+      "step": 593
+    },
+    {
+      "epoch": 0.6644295302013423,
+      "grad_norm": 0.40001606941223145,
+      "learning_rate": 4.892134177869709e-06,
+      "loss": 0.6141,
+      "step": 594
+    },
+    {
+      "epoch": 0.6655480984340044,
+      "grad_norm": 0.40650853514671326,
+      "learning_rate": 4.891700215946909e-06,
+      "loss": 0.6011,
+      "step": 595
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.41019386053085327,
+      "learning_rate": 4.8912654021513815e-06,
+      "loss": 0.6262,
+      "step": 596
+    },
+    {
+      "epoch": 0.6677852348993288,
+      "grad_norm": 0.4064581096172333,
+      "learning_rate": 4.890829736638e-06,
+      "loss": 0.6329,
+      "step": 597
+    },
+    {
+      "epoch": 0.668903803131991,
+      "grad_norm": 0.4091740846633911,
+      "learning_rate": 4.890393219561938e-06,
+      "loss": 0.6193,
+      "step": 598
+    },
+    {
+      "epoch": 0.6700223713646533,
+      "grad_norm": 0.407552570104599,
+      "learning_rate": 4.889955851078674e-06,
+      "loss": 0.6535,
+      "step": 599
+    },
+    {
+      "epoch": 0.6711409395973155,
+      "grad_norm": 0.3950585126876831,
+      "learning_rate": 4.889517631343988e-06,
+      "loss": 0.6033,
+      "step": 600
+    },
+    {
+      "epoch": 0.6722595078299777,
+      "grad_norm": 0.39515334367752075,
+      "learning_rate": 4.889078560513968e-06,
+      "loss": 0.6006,
+      "step": 601
+    },
+    {
+      "epoch": 0.6733780760626398,
+      "grad_norm": 0.4188336431980133,
+      "learning_rate": 4.888638638744999e-06,
+      "loss": 0.6333,
+      "step": 602
+    },
+    {
+      "epoch": 0.674496644295302,
+      "grad_norm": 0.4050522446632385,
+      "learning_rate": 4.888197866193772e-06,
+      "loss": 0.6329,
+      "step": 603
+    },
+    {
+      "epoch": 0.6756152125279642,
+      "grad_norm": 0.39428868889808655,
+      "learning_rate": 4.887756243017282e-06,
+      "loss": 0.6007,
+      "step": 604
+    },
+    {
+      "epoch": 0.6767337807606264,
+      "grad_norm": 0.402410626411438,
+      "learning_rate": 4.887313769372823e-06,
+      "loss": 0.5885,
+      "step": 605
+    },
+    {
+      "epoch": 0.6778523489932886,
+      "grad_norm": 0.4062318205833435,
+      "learning_rate": 4.886870445417998e-06,
+      "loss": 0.6312,
+      "step": 606
+    },
+    {
+      "epoch": 0.6789709172259508,
+      "grad_norm": 0.4123631715774536,
+      "learning_rate": 4.886426271310708e-06,
+      "loss": 0.619,
+      "step": 607
+    },
+    {
+      "epoch": 0.680089485458613,
+      "grad_norm": 0.40353336930274963,
+      "learning_rate": 4.885981247209159e-06,
+      "loss": 0.6324,
+      "step": 608
+    },
+    {
+      "epoch": 0.6812080536912751,
+      "grad_norm": 0.40726402401924133,
+      "learning_rate": 4.885535373271858e-06,
+      "loss": 0.5819,
+      "step": 609
+    },
+    {
+      "epoch": 0.6823266219239373,
+      "grad_norm": 0.4142349660396576,
+      "learning_rate": 4.885088649657618e-06,
+      "loss": 0.6175,
+      "step": 610
+    },
+    {
+      "epoch": 0.6834451901565995,
+      "grad_norm": 0.4165160059928894,
+      "learning_rate": 4.884641076525549e-06,
+      "loss": 0.597,
+      "step": 611
+    },
+    {
+      "epoch": 0.6845637583892618,
+      "grad_norm": 0.4037843942642212,
+      "learning_rate": 4.884192654035069e-06,
+      "loss": 0.6183,
+      "step": 612
+    },
+    {
+      "epoch": 0.685682326621924,
+      "grad_norm": 0.41777777671813965,
+      "learning_rate": 4.883743382345898e-06,
+      "loss": 0.6063,
+      "step": 613
+    },
+    {
+      "epoch": 0.6868008948545862,
+      "grad_norm": 0.41021421551704407,
+      "learning_rate": 4.883293261618054e-06,
+      "loss": 0.6134,
+      "step": 614
+    },
+    {
+      "epoch": 0.6879194630872483,
+      "grad_norm": 0.4215847849845886,
+      "learning_rate": 4.882842292011863e-06,
+      "loss": 0.6458,
+      "step": 615
+    },
+    {
+      "epoch": 0.6890380313199105,
+      "grad_norm": 0.42801633477211,
+      "learning_rate": 4.882390473687949e-06,
+      "loss": 0.6259,
+      "step": 616
+    },
+    {
+      "epoch": 0.6901565995525727,
+      "grad_norm": 0.40879589319229126,
+      "learning_rate": 4.881937806807241e-06,
+      "loss": 0.6208,
+      "step": 617
+    },
+    {
+      "epoch": 0.6912751677852349,
+      "grad_norm": 0.39453622698783875,
+      "learning_rate": 4.881484291530969e-06,
+      "loss": 0.5966,
+      "step": 618
+    },
+    {
+      "epoch": 0.6923937360178971,
+      "grad_norm": 0.3992539048194885,
+      "learning_rate": 4.881029928020666e-06,
+      "loss": 0.5976,
+      "step": 619
+    },
+    {
+      "epoch": 0.6935123042505593,
+      "grad_norm": 0.4175397753715515,
+      "learning_rate": 4.880574716438166e-06,
+      "loss": 0.6261,
+      "step": 620
+    },
+    {
+      "epoch": 0.6946308724832215,
+      "grad_norm": 0.40533408522605896,
+      "learning_rate": 4.880118656945606e-06,
+      "loss": 0.5945,
+      "step": 621
+    },
+    {
+      "epoch": 0.6957494407158836,
+      "grad_norm": 0.4089728593826294,
+      "learning_rate": 4.879661749705424e-06,
+      "loss": 0.6226,
+      "step": 622
+    },
+    {
+      "epoch": 0.6968680089485458,
+      "grad_norm": 0.4341566562652588,
+      "learning_rate": 4.879203994880362e-06,
+      "loss": 0.6463,
+      "step": 623
+    },
+    {
+      "epoch": 0.697986577181208,
+      "grad_norm": 0.44256189465522766,
+      "learning_rate": 4.878745392633462e-06,
+      "loss": 0.653,
+      "step": 624
+    },
+    {
+      "epoch": 0.6991051454138703,
+      "grad_norm": 0.4098159372806549,
+      "learning_rate": 4.878285943128067e-06,
+      "loss": 0.5808,
+      "step": 625
+    },
+    {
+      "epoch": 0.7002237136465325,
+      "grad_norm": 0.43130752444267273,
+      "learning_rate": 4.8778256465278245e-06,
+      "loss": 0.6261,
+      "step": 626
+    },
+    {
+      "epoch": 0.7013422818791947,
+      "grad_norm": 0.4110218286514282,
+      "learning_rate": 4.877364502996682e-06,
+      "loss": 0.5954,
+      "step": 627
+    },
+    {
+      "epoch": 0.7024608501118568,
+      "grad_norm": 0.42106136679649353,
+      "learning_rate": 4.87690251269889e-06,
+      "loss": 0.6026,
+      "step": 628
+    },
+    {
+      "epoch": 0.703579418344519,
+      "grad_norm": 0.4233524203300476,
+      "learning_rate": 4.876439675798997e-06,
+      "loss": 0.6432,
+      "step": 629
+    },
+    {
+      "epoch": 0.7046979865771812,
+      "grad_norm": 0.42107197642326355,
+      "learning_rate": 4.87597599246186e-06,
+      "loss": 0.6198,
+      "step": 630
+    },
+    {
+      "epoch": 0.7058165548098434,
+      "grad_norm": 0.43851831555366516,
+      "learning_rate": 4.875511462852628e-06,
+      "loss": 0.6293,
+      "step": 631
+    },
+    {
+      "epoch": 0.7069351230425056,
+      "grad_norm": 0.41345685720443726,
+      "learning_rate": 4.87504608713676e-06,
+      "loss": 0.6178,
+      "step": 632
+    },
+    {
+      "epoch": 0.7080536912751678,
+      "grad_norm": 0.41011178493499756,
+      "learning_rate": 4.874579865480013e-06,
+      "loss": 0.6441,
+      "step": 633
+    },
+    {
+      "epoch": 0.70917225950783,
+      "grad_norm": 0.41372135281562805,
+      "learning_rate": 4.874112798048442e-06,
+      "loss": 0.6142,
+      "step": 634
+    },
+    {
+      "epoch": 0.7102908277404921,
+      "grad_norm": 0.41231900453567505,
+      "learning_rate": 4.8736448850084105e-06,
+      "loss": 0.6277,
+      "step": 635
+    },
+    {
+      "epoch": 0.7114093959731543,
+      "grad_norm": 0.4147928059101105,
+      "learning_rate": 4.873176126526578e-06,
+      "loss": 0.6197,
+      "step": 636
+    },
+    {
+      "epoch": 0.7125279642058165,
+      "grad_norm": 0.4046717882156372,
+      "learning_rate": 4.8727065227699035e-06,
+      "loss": 0.6138,
+      "step": 637
+    },
+    {
+      "epoch": 0.7136465324384788,
+      "grad_norm": 0.4150887727737427,
+      "learning_rate": 4.872236073905654e-06,
+      "loss": 0.616,
+      "step": 638
+    },
+    {
+      "epoch": 0.714765100671141,
+      "grad_norm": 0.41429632902145386,
+      "learning_rate": 4.87176478010139e-06,
+      "loss": 0.6153,
+      "step": 639
+    },
+    {
+      "epoch": 0.7158836689038032,
+      "grad_norm": 0.41153407096862793,
+      "learning_rate": 4.8712926415249785e-06,
+      "loss": 0.6171,
+      "step": 640
+    },
+    {
+      "epoch": 0.7170022371364653,
+      "grad_norm": 0.4178698658943176,
+      "learning_rate": 4.870819658344584e-06,
+      "loss": 0.6417,
+      "step": 641
+    },
+    {
+      "epoch": 0.7181208053691275,
+      "grad_norm": 0.40587952733039856,
+      "learning_rate": 4.870345830728675e-06,
+      "loss": 0.6206,
+      "step": 642
+    },
+    {
+      "epoch": 0.7192393736017897,
+      "grad_norm": 0.42633864283561707,
+      "learning_rate": 4.869871158846016e-06,
+      "loss": 0.6246,
+      "step": 643
+    },
+    {
+      "epoch": 0.7203579418344519,
+      "grad_norm": 0.41023534536361694,
+      "learning_rate": 4.8693956428656766e-06,
+      "loss": 0.601,
+      "step": 644
+    },
+    {
+      "epoch": 0.7214765100671141,
+      "grad_norm": 0.40645042061805725,
+      "learning_rate": 4.868919282957024e-06,
+      "loss": 0.6193,
+      "step": 645
+    },
+    {
+      "epoch": 0.7225950782997763,
+      "grad_norm": 0.40088531374931335,
+      "learning_rate": 4.86844207928973e-06,
+      "loss": 0.5869,
+      "step": 646
+    },
+    {
+      "epoch": 0.7237136465324385,
+      "grad_norm": 0.4136696755886078,
+      "learning_rate": 4.8679640320337625e-06,
+      "loss": 0.6413,
+      "step": 647
+    },
+    {
+      "epoch": 0.7248322147651006,
+      "grad_norm": 0.40026187896728516,
+      "learning_rate": 4.867485141359394e-06,
+      "loss": 0.6075,
+      "step": 648
+    },
+    {
+      "epoch": 0.7259507829977628,
+      "grad_norm": 0.40911242365837097,
+      "learning_rate": 4.867005407437192e-06,
+      "loss": 0.6411,
+      "step": 649
+    },
+    {
+      "epoch": 0.727069351230425,
+      "grad_norm": 0.42306697368621826,
+      "learning_rate": 4.866524830438029e-06,
+      "loss": 0.6376,
+      "step": 650
+    },
+    {
+      "epoch": 0.7281879194630873,
+      "grad_norm": 0.40857061743736267,
+      "learning_rate": 4.866043410533077e-06,
+      "loss": 0.6071,
+      "step": 651
+    },
+    {
+      "epoch": 0.7293064876957495,
+      "grad_norm": 0.41601142287254333,
+      "learning_rate": 4.8655611478938055e-06,
+      "loss": 0.6079,
+      "step": 652
+    },
+    {
+      "epoch": 0.7304250559284117,
+      "grad_norm": 0.40857282280921936,
+      "learning_rate": 4.8650780426919895e-06,
+      "loss": 0.6246,
+      "step": 653
+    },
+    {
+      "epoch": 0.7315436241610739,
+      "grad_norm": 0.4063502252101898,
+      "learning_rate": 4.864594095099697e-06,
+      "loss": 0.6105,
+      "step": 654
+    },
+    {
+      "epoch": 0.732662192393736,
+      "grad_norm": 0.40278729796409607,
+      "learning_rate": 4.864109305289303e-06,
+      "loss": 0.5936,
+      "step": 655
+    },
+    {
+      "epoch": 0.7337807606263982,
+      "grad_norm": 0.4201098382472992,
+      "learning_rate": 4.863623673433478e-06,
+      "loss": 0.6081,
+      "step": 656
+    },
+    {
+      "epoch": 0.7348993288590604,
+      "grad_norm": 0.40003877878189087,
+      "learning_rate": 4.863137199705192e-06,
+      "loss": 0.6085,
+      "step": 657
+    },
+    {
+      "epoch": 0.7360178970917226,
+      "grad_norm": 0.41234898567199707,
+      "learning_rate": 4.86264988427772e-06,
+      "loss": 0.6252,
+      "step": 658
+    },
+    {
+      "epoch": 0.7371364653243848,
+      "grad_norm": 0.4233507513999939,
+      "learning_rate": 4.862161727324632e-06,
+      "loss": 0.5987,
+      "step": 659
+    },
+    {
+      "epoch": 0.738255033557047,
+      "grad_norm": 0.4099391996860504,
+      "learning_rate": 4.861672729019798e-06,
+      "loss": 0.6293,
+      "step": 660
+    },
+    {
+      "epoch": 0.7393736017897091,
+      "grad_norm": 0.4255772829055786,
+      "learning_rate": 4.861182889537389e-06,
+      "loss": 0.6268,
+      "step": 661
+    },
+    {
+      "epoch": 0.7404921700223713,
+      "grad_norm": 0.4317517578601837,
+      "learning_rate": 4.860692209051877e-06,
+      "loss": 0.6444,
+      "step": 662
+    },
+    {
+      "epoch": 0.7416107382550335,
+      "grad_norm": 0.4352816939353943,
+      "learning_rate": 4.86020068773803e-06,
+      "loss": 0.6304,
+      "step": 663
+    },
+    {
+      "epoch": 0.7427293064876958,
+      "grad_norm": 0.3987254500389099,
+      "learning_rate": 4.859708325770919e-06,
+      "loss": 0.611,
+      "step": 664
+    },
+    {
+      "epoch": 0.743847874720358,
+      "grad_norm": 0.4213384985923767,
+      "learning_rate": 4.859215123325912e-06,
+      "loss": 0.6292,
+      "step": 665
+    },
+    {
+      "epoch": 0.7449664429530202,
+      "grad_norm": 0.4062172472476959,
+      "learning_rate": 4.8587210805786785e-06,
+      "loss": 0.6197,
+      "step": 666
+    },
+    {
+      "epoch": 0.7460850111856824,
+      "grad_norm": 0.41443362832069397,
+      "learning_rate": 4.858226197705183e-06,
+      "loss": 0.6414,
+      "step": 667
+    },
+    {
+      "epoch": 0.7472035794183445,
+      "grad_norm": 0.4183506667613983,
+      "learning_rate": 4.857730474881696e-06,
+      "loss": 0.6294,
+      "step": 668
+    },
+    {
+      "epoch": 0.7483221476510067,
+      "grad_norm": 0.42685073614120483,
+      "learning_rate": 4.857233912284781e-06,
+      "loss": 0.6264,
+      "step": 669
+    },
+    {
+      "epoch": 0.7494407158836689,
+      "grad_norm": 0.4143792390823364,
+      "learning_rate": 4.856736510091304e-06,
+      "loss": 0.6575,
+      "step": 670
+    },
+    {
+      "epoch": 0.7505592841163311,
+      "grad_norm": 0.4124217629432678,
+      "learning_rate": 4.8562382684784284e-06,
+      "loss": 0.6295,
+      "step": 671
+    },
+    {
+      "epoch": 0.7516778523489933,
+      "grad_norm": 0.4060792624950409,
+      "learning_rate": 4.855739187623619e-06,
+      "loss": 0.5983,
+      "step": 672
+    },
+    {
+      "epoch": 0.7527964205816555,
+      "grad_norm": 0.4100533723831177,
+      "learning_rate": 4.855239267704635e-06,
+      "loss": 0.6271,
+      "step": 673
+    },
+    {
+      "epoch": 0.7539149888143176,
+      "grad_norm": 0.4047471582889557,
+      "learning_rate": 4.854738508899538e-06,
+      "loss": 0.5843,
+      "step": 674
+    },
+    {
+      "epoch": 0.7550335570469798,
+      "grad_norm": 0.41550201177597046,
+      "learning_rate": 4.854236911386689e-06,
+      "loss": 0.6015,
+      "step": 675
+    },
+    {
+      "epoch": 0.756152125279642,
+      "grad_norm": 0.4035356044769287,
+      "learning_rate": 4.853734475344745e-06,
+      "loss": 0.6085,
+      "step": 676
+    },
+    {
+      "epoch": 0.7572706935123042,
+      "grad_norm": 0.4054676294326782,
+      "learning_rate": 4.853231200952665e-06,
+      "loss": 0.5879,
+      "step": 677
+    },
+    {
+      "epoch": 0.7583892617449665,
+      "grad_norm": 0.4165349304676056,
+      "learning_rate": 4.852727088389702e-06,
+      "loss": 0.6065,
+      "step": 678
+    },
+    {
+      "epoch": 0.7595078299776287,
+      "grad_norm": 0.41854768991470337,
+      "learning_rate": 4.8522221378354125e-06,
+      "loss": 0.6115,
+      "step": 679
+    },
+    {
+      "epoch": 0.7606263982102909,
+      "grad_norm": 0.4189227223396301,
+      "learning_rate": 4.851716349469647e-06,
+      "loss": 0.6174,
+      "step": 680
+    },
+    {
+      "epoch": 0.761744966442953,
+      "grad_norm": 0.44432833790779114,
+      "learning_rate": 4.851209723472559e-06,
+      "loss": 0.6382,
+      "step": 681
+    },
+    {
+      "epoch": 0.7628635346756152,
+      "grad_norm": 0.4199828803539276,
+      "learning_rate": 4.8507022600245954e-06,
+      "loss": 0.6125,
+      "step": 682
+    },
+    {
+      "epoch": 0.7639821029082774,
+      "grad_norm": 0.44079893827438354,
+      "learning_rate": 4.850193959306506e-06,
+      "loss": 0.6263,
+      "step": 683
+    },
+    {
+      "epoch": 0.7651006711409396,
+      "grad_norm": 0.41406047344207764,
+      "learning_rate": 4.8496848214993355e-06,
+      "loss": 0.5979,
+      "step": 684
+    },
+    {
+      "epoch": 0.7662192393736018,
+      "grad_norm": 0.43209850788116455,
+      "learning_rate": 4.849174846784428e-06,
+      "loss": 0.6451,
+      "step": 685
+    },
+    {
+      "epoch": 0.767337807606264,
+      "grad_norm": 0.4180072844028473,
+      "learning_rate": 4.848664035343425e-06,
+      "loss": 0.6009,
+      "step": 686
+    },
+    {
+      "epoch": 0.7684563758389261,
+      "grad_norm": 0.4092356860637665,
+      "learning_rate": 4.8481523873582685e-06,
+      "loss": 0.6431,
+      "step": 687
+    },
+    {
+      "epoch": 0.7695749440715883,
+      "grad_norm": 0.41440829634666443,
+      "learning_rate": 4.847639903011196e-06,
+      "loss": 0.6001,
+      "step": 688
+    },
+    {
+      "epoch": 0.7706935123042505,
+      "grad_norm": 0.4246008098125458,
+      "learning_rate": 4.8471265824847415e-06,
+      "loss": 0.6137,
+      "step": 689
+    },
+    {
+      "epoch": 0.7718120805369127,
+      "grad_norm": 0.4177666902542114,
+      "learning_rate": 4.846612425961742e-06,
+      "loss": 0.6026,
+      "step": 690
+    },
+    {
+      "epoch": 0.772930648769575,
+      "grad_norm": 0.4130840003490448,
+      "learning_rate": 4.846097433625327e-06,
+      "loss": 0.6183,
+      "step": 691
+    },
+    {
+      "epoch": 0.7740492170022372,
+      "grad_norm": 0.406780868768692,
+      "learning_rate": 4.845581605658926e-06,
+      "loss": 0.5992,
+      "step": 692
+    },
+    {
+      "epoch": 0.7751677852348994,
+      "grad_norm": 0.42086103558540344,
+      "learning_rate": 4.845064942246267e-06,
+      "loss": 0.6057,
+      "step": 693
+    },
+    {
+      "epoch": 0.7762863534675615,
+      "grad_norm": 0.4122505486011505,
+      "learning_rate": 4.844547443571374e-06,
+      "loss": 0.6134,
+      "step": 694
+    },
+    {
+      "epoch": 0.7774049217002237,
+      "grad_norm": 0.43634387850761414,
+      "learning_rate": 4.8440291098185686e-06,
+      "loss": 0.6044,
+      "step": 695
+    },
+    {
+      "epoch": 0.7785234899328859,
+      "grad_norm": 0.4160690903663635,
+      "learning_rate": 4.843509941172471e-06,
+      "loss": 0.6046,
+      "step": 696
+    },
+    {
+      "epoch": 0.7796420581655481,
+      "grad_norm": 0.41897231340408325,
+      "learning_rate": 4.842989937817997e-06,
+      "loss": 0.6186,
+      "step": 697
+    },
+    {
+      "epoch": 0.7807606263982103,
+      "grad_norm": 0.4187341034412384,
+      "learning_rate": 4.842469099940361e-06,
+      "loss": 0.6266,
+      "step": 698
+    },
+    {
+      "epoch": 0.7818791946308725,
+      "grad_norm": 0.4075968563556671,
+      "learning_rate": 4.841947427725076e-06,
+      "loss": 0.5772,
+      "step": 699
+    },
+    {
+      "epoch": 0.7829977628635347,
+      "grad_norm": 0.4157114028930664,
+      "learning_rate": 4.841424921357948e-06,
+      "loss": 0.5999,
+      "step": 700
+    },
+    {
+      "epoch": 0.7841163310961968,
+      "grad_norm": 0.4198933243751526,
+      "learning_rate": 4.840901581025083e-06,
+      "loss": 0.6273,
+      "step": 701
+    },
+    {
+      "epoch": 0.785234899328859,
+      "grad_norm": 0.42646607756614685,
+      "learning_rate": 4.840377406912887e-06,
+      "loss": 0.6074,
+      "step": 702
+    },
+    {
+      "epoch": 0.7863534675615212,
+      "grad_norm": 0.42644554376602173,
+      "learning_rate": 4.839852399208056e-06,
+      "loss": 0.5872,
+      "step": 703
+    },
+    {
+      "epoch": 0.7874720357941835,
+      "grad_norm": 0.43172845244407654,
+      "learning_rate": 4.839326558097587e-06,
+      "loss": 0.633,
+      "step": 704
+    },
+    {
+      "epoch": 0.7885906040268457,
+      "grad_norm": 0.4165332317352295,
+      "learning_rate": 4.838799883768775e-06,
+      "loss": 0.6206,
+      "step": 705
+    },
+    {
+      "epoch": 0.7897091722595079,
+      "grad_norm": 0.4209877550601959,
+      "learning_rate": 4.83827237640921e-06,
+      "loss": 0.6015,
+      "step": 706
+    },
+    {
+      "epoch": 0.79082774049217,
+      "grad_norm": 0.4267021715641022,
+      "learning_rate": 4.837744036206777e-06,
+      "loss": 0.5975,
+      "step": 707
+    },
+    {
+      "epoch": 0.7919463087248322,
+      "grad_norm": 0.4415457546710968,
+      "learning_rate": 4.837214863349662e-06,
+      "loss": 0.6251,
+      "step": 708
+    },
+    {
+      "epoch": 0.7930648769574944,
+      "grad_norm": 0.43104031682014465,
+      "learning_rate": 4.836684858026343e-06,
+      "loss": 0.6048,
+      "step": 709
+    },
+    {
+      "epoch": 0.7941834451901566,
+      "grad_norm": 0.41736820340156555,
+      "learning_rate": 4.8361540204255985e-06,
+      "loss": 0.5948,
+      "step": 710
+    },
+    {
+      "epoch": 0.7953020134228188,
+      "grad_norm": 0.4202009439468384,
+      "learning_rate": 4.835622350736499e-06,
+      "loss": 0.6099,
+      "step": 711
+    },
+    {
+      "epoch": 0.796420581655481,
+      "grad_norm": 0.42279568314552307,
+      "learning_rate": 4.8350898491484175e-06,
+      "loss": 0.6247,
+      "step": 712
+    },
+    {
+      "epoch": 0.7975391498881432,
+      "grad_norm": 0.4266239404678345,
+      "learning_rate": 4.8345565158510176e-06,
+      "loss": 0.6136,
+      "step": 713
+    },
+    {
+      "epoch": 0.7986577181208053,
+      "grad_norm": 0.42605841159820557,
+      "learning_rate": 4.83402235103426e-06,
+      "loss": 0.6001,
+      "step": 714
+    },
+    {
+      "epoch": 0.7997762863534675,
+      "grad_norm": 0.42846307158470154,
+      "learning_rate": 4.8334873548884055e-06,
+      "loss": 0.5941,
+      "step": 715
+    },
+    {
+      "epoch": 0.8008948545861297,
+      "grad_norm": 0.44009047746658325,
+      "learning_rate": 4.832951527604007e-06,
+      "loss": 0.622,
+      "step": 716
+    },
+    {
+      "epoch": 0.802013422818792,
+      "grad_norm": 0.44512951374053955,
+      "learning_rate": 4.8324148693719145e-06,
+      "loss": 0.6507,
+      "step": 717
+    },
+    {
+      "epoch": 0.8031319910514542,
+      "grad_norm": 0.455010324716568,
+      "learning_rate": 4.831877380383276e-06,
+      "loss": 0.6201,
+      "step": 718
+    },
+    {
+      "epoch": 0.8042505592841164,
+      "grad_norm": 0.43456459045410156,
+      "learning_rate": 4.83133906082953e-06,
+      "loss": 0.623,
+      "step": 719
+    },
+    {
+      "epoch": 0.8053691275167785,
+      "grad_norm": 0.42063653469085693,
+      "learning_rate": 4.830799910902418e-06,
+      "loss": 0.5841,
+      "step": 720
+    },
+    {
+      "epoch": 0.8064876957494407,
+      "grad_norm": 0.41323843598365784,
+      "learning_rate": 4.8302599307939725e-06,
+      "loss": 0.6127,
+      "step": 721
+    },
+    {
+      "epoch": 0.8076062639821029,
+      "grad_norm": 0.41982001066207886,
+      "learning_rate": 4.829719120696523e-06,
+      "loss": 0.6274,
+      "step": 722
+    },
+    {
+      "epoch": 0.8087248322147651,
+      "grad_norm": 0.43330860137939453,
+      "learning_rate": 4.829177480802694e-06,
+      "loss": 0.6416,
+      "step": 723
+    },
+    {
+      "epoch": 0.8098434004474273,
+      "grad_norm": 0.4351330101490021,
+      "learning_rate": 4.828635011305407e-06,
+      "loss": 0.6399,
+      "step": 724
+    },
+    {
+      "epoch": 0.8109619686800895,
+      "grad_norm": 0.4017598032951355,
+      "learning_rate": 4.828091712397878e-06,
+      "loss": 0.5817,
+      "step": 725
+    },
+    {
+      "epoch": 0.8120805369127517,
+      "grad_norm": 0.42594751715660095,
+      "learning_rate": 4.827547584273618e-06,
+      "loss": 0.6438,
+      "step": 726
+    },
+    {
+      "epoch": 0.8131991051454138,
+      "grad_norm": 0.409135639667511,
+      "learning_rate": 4.827002627126433e-06,
+      "loss": 0.5797,
+      "step": 727
+    },
+    {
+      "epoch": 0.814317673378076,
+      "grad_norm": 0.4304857850074768,
+      "learning_rate": 4.826456841150428e-06,
+      "loss": 0.6173,
+      "step": 728
+    },
+    {
+      "epoch": 0.8154362416107382,
+      "grad_norm": 0.446872740983963,
+      "learning_rate": 4.825910226539997e-06,
+      "loss": 0.6059,
+      "step": 729
+    },
+    {
+      "epoch": 0.8165548098434005,
+      "grad_norm": 0.42625290155410767,
+      "learning_rate": 4.8253627834898355e-06,
+      "loss": 0.5994,
+      "step": 730
+    },
+    {
+      "epoch": 0.8176733780760627,
+      "grad_norm": 0.42183029651641846,
+      "learning_rate": 4.824814512194929e-06,
+      "loss": 0.6202,
+      "step": 731
+    },
+    {
+      "epoch": 0.8187919463087249,
+      "grad_norm": 0.4235664904117584,
+      "learning_rate": 4.824265412850559e-06,
+      "loss": 0.6263,
+      "step": 732
+    },
+    {
+      "epoch": 0.819910514541387,
+      "grad_norm": 0.4118615686893463,
+      "learning_rate": 4.823715485652307e-06,
+      "loss": 0.6058,
+      "step": 733
+    },
+    {
+      "epoch": 0.8210290827740492,
+      "grad_norm": 0.43514224886894226,
+      "learning_rate": 4.823164730796042e-06,
+      "loss": 0.6092,
+      "step": 734
+    },
+    {
+      "epoch": 0.8221476510067114,
+      "grad_norm": 0.41756734251976013,
+      "learning_rate": 4.8226131484779325e-06,
+      "loss": 0.6281,
+      "step": 735
+    },
+    {
+      "epoch": 0.8232662192393736,
+      "grad_norm": 0.438475638628006,
+      "learning_rate": 4.822060738894439e-06,
+      "loss": 0.6122,
+      "step": 736
+    },
+    {
+      "epoch": 0.8243847874720358,
+      "grad_norm": 0.426792174577713,
+      "learning_rate": 4.821507502242321e-06,
+      "loss": 0.6407,
+      "step": 737
+    },
+    {
+      "epoch": 0.825503355704698,
+      "grad_norm": 0.42697012424468994,
+      "learning_rate": 4.820953438718626e-06,
+      "loss": 0.5996,
+      "step": 738
+    },
+    {
+      "epoch": 0.8266219239373602,
+      "grad_norm": 0.42373016476631165,
+      "learning_rate": 4.820398548520702e-06,
+      "loss": 0.6075,
+      "step": 739
+    },
+    {
+      "epoch": 0.8277404921700223,
+      "grad_norm": 0.42235615849494934,
+      "learning_rate": 4.81984283184619e-06,
+      "loss": 0.608,
+      "step": 740
+    },
+    {
+      "epoch": 0.8288590604026845,
+      "grad_norm": 0.41180866956710815,
+      "learning_rate": 4.819286288893022e-06,
+      "loss": 0.6127,
+      "step": 741
+    },
+    {
+      "epoch": 0.8299776286353467,
+      "grad_norm": 0.4207548499107361,
+      "learning_rate": 4.818728919859426e-06,
+      "loss": 0.6131,
+      "step": 742
+    },
+    {
+      "epoch": 0.831096196868009,
+      "grad_norm": 0.4295390546321869,
+      "learning_rate": 4.818170724943928e-06,
+      "loss": 0.629,
+      "step": 743
+    },
+    {
+      "epoch": 0.8322147651006712,
+      "grad_norm": 0.4099291265010834,
+      "learning_rate": 4.817611704345344e-06,
+      "loss": 0.6055,
+      "step": 744
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 0.4142029583454132,
+      "learning_rate": 4.817051858262785e-06,
+      "loss": 0.6127,
+      "step": 745
+    },
+    {
+      "epoch": 0.8344519015659956,
+      "grad_norm": 0.41662877798080444,
+      "learning_rate": 4.816491186895656e-06,
+      "loss": 0.6171,
+      "step": 746
+    },
+    {
+      "epoch": 0.8355704697986577,
+      "grad_norm": 0.4345078766345978,
+      "learning_rate": 4.815929690443657e-06,
+      "loss": 0.6091,
+      "step": 747
+    },
+    {
+      "epoch": 0.8366890380313199,
+      "grad_norm": 0.43010810017585754,
+      "learning_rate": 4.8153673691067806e-06,
+      "loss": 0.626,
+      "step": 748
+    },
+    {
+      "epoch": 0.8378076062639821,
+      "grad_norm": 0.4256346821784973,
+      "learning_rate": 4.814804223085313e-06,
+      "loss": 0.6216,
+      "step": 749
+    },
+    {
+      "epoch": 0.8389261744966443,
+      "grad_norm": 0.42812031507492065,
+      "learning_rate": 4.814240252579836e-06,
+      "loss": 0.6138,
+      "step": 750
+    },
+    {
+      "epoch": 0.8400447427293065,
+      "grad_norm": 0.41682377457618713,
+      "learning_rate": 4.813675457791224e-06,
+      "loss": 0.5783,
+      "step": 751
+    },
+    {
+      "epoch": 0.8411633109619687,
+      "grad_norm": 0.4257197380065918,
+      "learning_rate": 4.8131098389206435e-06,
+      "loss": 0.6006,
+      "step": 752
+    },
+    {
+      "epoch": 0.8422818791946308,
+      "grad_norm": 0.42872053384780884,
+      "learning_rate": 4.812543396169557e-06,
+      "loss": 0.6272,
+      "step": 753
+    },
+    {
+      "epoch": 0.843400447427293,
+      "grad_norm": 0.4263060986995697,
+      "learning_rate": 4.81197612973972e-06,
+      "loss": 0.6093,
+      "step": 754
+    },
+    {
+      "epoch": 0.8445190156599552,
+      "grad_norm": 0.4151028096675873,
+      "learning_rate": 4.811408039833178e-06,
+      "loss": 0.5773,
+      "step": 755
+    },
+    {
+      "epoch": 0.8456375838926175,
+      "grad_norm": 0.4382091164588928,
+      "learning_rate": 4.810839126652275e-06,
+      "loss": 0.596,
+      "step": 756
+    },
+    {
+      "epoch": 0.8467561521252797,
+      "grad_norm": 0.4290934205055237,
+      "learning_rate": 4.810269390399646e-06,
+      "loss": 0.5904,
+      "step": 757
+    },
+    {
+      "epoch": 0.8478747203579419,
+      "grad_norm": 0.4299798011779785,
+      "learning_rate": 4.809698831278217e-06,
+      "loss": 0.6449,
+      "step": 758
+    },
+    {
+      "epoch": 0.8489932885906041,
+      "grad_norm": 0.4188365042209625,
+      "learning_rate": 4.809127449491211e-06,
+      "loss": 0.6007,
+      "step": 759
+    },
+    {
+      "epoch": 0.8501118568232662,
+      "grad_norm": 0.41594186425209045,
+      "learning_rate": 4.808555245242141e-06,
+      "loss": 0.5888,
+      "step": 760
+    },
+    {
+      "epoch": 0.8512304250559284,
+      "grad_norm": 0.4184630513191223,
+      "learning_rate": 4.807982218734814e-06,
+      "loss": 0.6495,
+      "step": 761
+    },
+    {
+      "epoch": 0.8523489932885906,
+      "grad_norm": 0.418849378824234,
+      "learning_rate": 4.80740837017333e-06,
+      "loss": 0.6128,
+      "step": 762
+    },
+    {
+      "epoch": 0.8534675615212528,
+      "grad_norm": 0.42030590772628784,
+      "learning_rate": 4.8068336997620804e-06,
+      "loss": 0.6294,
+      "step": 763
+    },
+    {
+      "epoch": 0.854586129753915,
+      "grad_norm": 0.4204208552837372,
+      "learning_rate": 4.806258207705753e-06,
+      "loss": 0.6279,
+      "step": 764
+    },
+    {
+      "epoch": 0.8557046979865772,
+      "grad_norm": 0.41544729471206665,
+      "learning_rate": 4.805681894209324e-06,
+      "loss": 0.6235,
+      "step": 765
+    },
+    {
+      "epoch": 0.8568232662192393,
+      "grad_norm": 0.430301308631897,
+      "learning_rate": 4.805104759478065e-06,
+      "loss": 0.5876,
+      "step": 766
+    },
+    {
+      "epoch": 0.8579418344519015,
+      "grad_norm": 0.4640876054763794,
+      "learning_rate": 4.804526803717539e-06,
+      "loss": 0.6264,
+      "step": 767
+    },
+    {
+      "epoch": 0.8590604026845637,
+      "grad_norm": 0.43245360255241394,
+      "learning_rate": 4.8039480271336005e-06,
+      "loss": 0.5871,
+      "step": 768
+    },
+    {
+      "epoch": 0.860178970917226,
+      "grad_norm": 0.41617804765701294,
+      "learning_rate": 4.803368429932399e-06,
+      "loss": 0.6218,
+      "step": 769
+    },
+    {
+      "epoch": 0.8612975391498882,
+      "grad_norm": 0.4325237274169922,
+      "learning_rate": 4.8027880123203726e-06,
+      "loss": 0.5874,
+      "step": 770
+    },
+    {
+      "epoch": 0.8624161073825504,
+      "grad_norm": 0.4480580687522888,
+      "learning_rate": 4.802206774504255e-06,
+      "loss": 0.6093,
+      "step": 771
+    },
+    {
+      "epoch": 0.8635346756152126,
+      "grad_norm": 0.44137299060821533,
+      "learning_rate": 4.801624716691072e-06,
+      "loss": 0.6031,
+      "step": 772
+    },
+    {
+      "epoch": 0.8646532438478747,
+      "grad_norm": 0.4269614815711975,
+      "learning_rate": 4.801041839088139e-06,
+      "loss": 0.5963,
+      "step": 773
+    },
+    {
+      "epoch": 0.8657718120805369,
+      "grad_norm": 0.4120911657810211,
+      "learning_rate": 4.800458141903064e-06,
+      "loss": 0.5959,
+      "step": 774
+    },
+    {
+      "epoch": 0.8668903803131991,
+      "grad_norm": 0.4332381784915924,
+      "learning_rate": 4.799873625343747e-06,
+      "loss": 0.6007,
+      "step": 775
+    },
+    {
+      "epoch": 0.8680089485458613,
+      "grad_norm": 0.4361085295677185,
+      "learning_rate": 4.7992882896183825e-06,
+      "loss": 0.6012,
+      "step": 776
+    },
+    {
+      "epoch": 0.8691275167785235,
+      "grad_norm": 0.43251463770866394,
+      "learning_rate": 4.798702134935454e-06,
+      "loss": 0.5799,
+      "step": 777
+    },
+    {
+      "epoch": 0.8702460850111857,
+      "grad_norm": 0.4305305778980255,
+      "learning_rate": 4.798115161503735e-06,
+      "loss": 0.6068,
+      "step": 778
+    },
+    {
+      "epoch": 0.8713646532438478,
+      "grad_norm": 0.4410499930381775,
+      "learning_rate": 4.797527369532296e-06,
+      "loss": 0.6486,
+      "step": 779
+    },
+    {
+      "epoch": 0.87248322147651,
+      "grad_norm": 0.43271416425704956,
+      "learning_rate": 4.796938759230494e-06,
+      "loss": 0.6367,
+      "step": 780
+    },
+    {
+      "epoch": 0.8736017897091722,
+      "grad_norm": 0.430905282497406,
+      "learning_rate": 4.7963493308079815e-06,
+      "loss": 0.5753,
+      "step": 781
+    },
+    {
+      "epoch": 0.8747203579418344,
+      "grad_norm": 0.4054125249385834,
+      "learning_rate": 4.7957590844746986e-06,
+      "loss": 0.5806,
+      "step": 782
+    },
+    {
+      "epoch": 0.8758389261744967,
+      "grad_norm": 0.4322145879268646,
+      "learning_rate": 4.795168020440878e-06,
+      "loss": 0.5989,
+      "step": 783
+    },
+    {
+      "epoch": 0.8769574944071589,
+      "grad_norm": 0.433716744184494,
+      "learning_rate": 4.7945761389170464e-06,
+      "loss": 0.6284,
+      "step": 784
+    },
+    {
+      "epoch": 0.8780760626398211,
+      "grad_norm": 0.4301673471927643,
+      "learning_rate": 4.793983440114018e-06,
+      "loss": 0.6469,
+      "step": 785
+    },
+    {
+      "epoch": 0.8791946308724832,
+      "grad_norm": 0.4279182553291321,
+      "learning_rate": 4.7933899242428986e-06,
+      "loss": 0.6032,
+      "step": 786
+    },
+    {
+      "epoch": 0.8803131991051454,
+      "grad_norm": 0.4133903682231903,
+      "learning_rate": 4.792795591515087e-06,
+      "loss": 0.5745,
+      "step": 787
+    },
+    {
+      "epoch": 0.8814317673378076,
+      "grad_norm": 0.45094674825668335,
+      "learning_rate": 4.792200442142273e-06,
+      "loss": 0.6212,
+      "step": 788
+    },
+    {
+      "epoch": 0.8825503355704698,
+      "grad_norm": 0.4304371774196625,
+      "learning_rate": 4.7916044763364344e-06,
+      "loss": 0.61,
+      "step": 789
+    },
+    {
+      "epoch": 0.883668903803132,
+      "grad_norm": 0.4232103228569031,
+      "learning_rate": 4.791007694309842e-06,
+      "loss": 0.5942,
+      "step": 790
+    },
+    {
+      "epoch": 0.8847874720357942,
+      "grad_norm": 0.41822549700737,
+      "learning_rate": 4.790410096275057e-06,
+      "loss": 0.5829,
+      "step": 791
+    },
+    {
+      "epoch": 0.8859060402684564,
+      "grad_norm": 0.44473201036453247,
+      "learning_rate": 4.789811682444931e-06,
+      "loss": 0.6359,
+      "step": 792
+    },
+    {
+      "epoch": 0.8870246085011185,
+      "grad_norm": 0.42226001620292664,
+      "learning_rate": 4.7892124530326065e-06,
+      "loss": 0.5966,
+      "step": 793
+    },
+    {
+      "epoch": 0.8881431767337807,
+      "grad_norm": 0.4254050552845001,
+      "learning_rate": 4.788612408251517e-06,
+      "loss": 0.6211,
+      "step": 794
+    },
+    {
+      "epoch": 0.889261744966443,
+      "grad_norm": 0.42909836769104004,
+      "learning_rate": 4.788011548315383e-06,
+      "loss": 0.6039,
+      "step": 795
+    },
+    {
+      "epoch": 0.8903803131991052,
+      "grad_norm": 0.4296148419380188,
+      "learning_rate": 4.78740987343822e-06,
+      "loss": 0.6055,
+      "step": 796
+    },
+    {
+      "epoch": 0.8914988814317674,
+      "grad_norm": 0.4225505292415619,
+      "learning_rate": 4.786807383834332e-06,
+      "loss": 0.5947,
+      "step": 797
+    },
+    {
+      "epoch": 0.8926174496644296,
+      "grad_norm": 0.4271566867828369,
+      "learning_rate": 4.786204079718314e-06,
+      "loss": 0.6002,
+      "step": 798
+    },
+    {
+      "epoch": 0.8937360178970917,
+      "grad_norm": 0.42522063851356506,
+      "learning_rate": 4.785599961305048e-06,
+      "loss": 0.6231,
+      "step": 799
+    },
+    {
+      "epoch": 0.8948545861297539,
+      "grad_norm": 0.4384607970714569,
+      "learning_rate": 4.784995028809707e-06,
+      "loss": 0.6072,
+      "step": 800
+    },
+    {
+      "epoch": 0.8959731543624161,
+      "grad_norm": 0.4194418787956238,
+      "learning_rate": 4.784389282447759e-06,
+      "loss": 0.5979,
+      "step": 801
+    },
+    {
+      "epoch": 0.8970917225950783,
+      "grad_norm": 0.43825826048851013,
+      "learning_rate": 4.7837827224349544e-06,
+      "loss": 0.6256,
+      "step": 802
+    },
+    {
+      "epoch": 0.8982102908277405,
+      "grad_norm": 0.43381622433662415,
+      "learning_rate": 4.783175348987339e-06,
+      "loss": 0.5932,
+      "step": 803
+    },
+    {
+      "epoch": 0.8993288590604027,
+      "grad_norm": 0.4468502700328827,
+      "learning_rate": 4.7825671623212456e-06,
+      "loss": 0.618,
+      "step": 804
+    },
+    {
+      "epoch": 0.9004474272930649,
+      "grad_norm": 0.4352877140045166,
+      "learning_rate": 4.781958162653298e-06,
+      "loss": 0.5898,
+      "step": 805
+    },
+    {
+      "epoch": 0.901565995525727,
+      "grad_norm": 0.4242689907550812,
+      "learning_rate": 4.781348350200408e-06,
+      "loss": 0.5856,
+      "step": 806
+    },
+    {
+      "epoch": 0.9026845637583892,
+      "grad_norm": 0.4262087941169739,
+      "learning_rate": 4.780737725179778e-06,
+      "loss": 0.5994,
+      "step": 807
+    },
+    {
+      "epoch": 0.9038031319910514,
+      "grad_norm": 0.42303264141082764,
+      "learning_rate": 4.780126287808899e-06,
+      "loss": 0.6106,
+      "step": 808
+    },
+    {
+      "epoch": 0.9049217002237137,
+      "grad_norm": 0.43589121103286743,
+      "learning_rate": 4.779514038305555e-06,
+      "loss": 0.6251,
+      "step": 809
+    },
+    {
+      "epoch": 0.9060402684563759,
+      "grad_norm": 0.43768516182899475,
+      "learning_rate": 4.778900976887813e-06,
+      "loss": 0.6124,
+      "step": 810
+    },
+    {
+      "epoch": 0.9071588366890381,
+      "grad_norm": 0.4439849257469177,
+      "learning_rate": 4.778287103774033e-06,
+      "loss": 0.6397,
+      "step": 811
+    },
+    {
+      "epoch": 0.9082774049217002,
+      "grad_norm": 0.44813254475593567,
+      "learning_rate": 4.777672419182863e-06,
+      "loss": 0.6213,
+      "step": 812
+    },
+    {
+      "epoch": 0.9093959731543624,
+      "grad_norm": 0.4133831858634949,
+      "learning_rate": 4.777056923333244e-06,
+      "loss": 0.6138,
+      "step": 813
+    },
+    {
+      "epoch": 0.9105145413870246,
+      "grad_norm": 0.4255264699459076,
+      "learning_rate": 4.7764406164444e-06,
+      "loss": 0.6143,
+      "step": 814
+    },
+    {
+      "epoch": 0.9116331096196868,
+      "grad_norm": 0.42810630798339844,
+      "learning_rate": 4.775823498735845e-06,
+      "loss": 0.6253,
+      "step": 815
+    },
+    {
+      "epoch": 0.912751677852349,
+      "grad_norm": 0.42162856459617615,
+      "learning_rate": 4.775205570427386e-06,
+      "loss": 0.602,
+      "step": 816
+    },
+    {
+      "epoch": 0.9138702460850112,
+      "grad_norm": 0.4342280328273773,
+      "learning_rate": 4.7745868317391135e-06,
+      "loss": 0.6088,
+      "step": 817
+    },
+    {
+      "epoch": 0.9149888143176734,
+      "grad_norm": 0.42438629269599915,
+      "learning_rate": 4.773967282891411e-06,
+      "loss": 0.5788,
+      "step": 818
+    },
+    {
+      "epoch": 0.9161073825503355,
+      "grad_norm": 0.437950074672699,
+      "learning_rate": 4.7733469241049475e-06,
+      "loss": 0.6277,
+      "step": 819
+    },
+    {
+      "epoch": 0.9172259507829977,
+      "grad_norm": 0.4286377429962158,
+      "learning_rate": 4.772725755600682e-06,
+      "loss": 0.6024,
+      "step": 820
+    },
+    {
+      "epoch": 0.9183445190156599,
+      "grad_norm": 0.4317566156387329,
+      "learning_rate": 4.772103777599861e-06,
+      "loss": 0.6048,
+      "step": 821
+    },
+    {
+      "epoch": 0.9194630872483222,
+      "grad_norm": 0.4509202837944031,
+      "learning_rate": 4.771480990324021e-06,
+      "loss": 0.6219,
+      "step": 822
+    },
+    {
+      "epoch": 0.9205816554809844,
+      "grad_norm": 0.4387308955192566,
+      "learning_rate": 4.7708573939949845e-06,
+      "loss": 0.6082,
+      "step": 823
+    },
+    {
+      "epoch": 0.9217002237136466,
+      "grad_norm": 0.457883358001709,
+      "learning_rate": 4.770232988834864e-06,
+      "loss": 0.6112,
+      "step": 824
+    },
+    {
+      "epoch": 0.9228187919463087,
+      "grad_norm": 0.44200408458709717,
+      "learning_rate": 4.769607775066058e-06,
+      "loss": 0.6146,
+      "step": 825
+    },
+    {
+      "epoch": 0.9239373601789709,
+      "grad_norm": 0.44704675674438477,
+      "learning_rate": 4.768981752911256e-06,
+      "loss": 0.5921,
+      "step": 826
+    },
+    {
+      "epoch": 0.9250559284116331,
+      "grad_norm": 0.4367467164993286,
+      "learning_rate": 4.768354922593433e-06,
+      "loss": 0.6075,
+      "step": 827
+    },
+    {
+      "epoch": 0.9261744966442953,
+      "grad_norm": 0.4321734309196472,
+      "learning_rate": 4.767727284335852e-06,
+      "loss": 0.6041,
+      "step": 828
+    },
+    {
+      "epoch": 0.9272930648769575,
+      "grad_norm": 0.42991605401039124,
+      "learning_rate": 4.767098838362065e-06,
+      "loss": 0.5804,
+      "step": 829
+    },
+    {
+      "epoch": 0.9284116331096197,
+      "grad_norm": 0.43791651725769043,
+      "learning_rate": 4.766469584895912e-06,
+      "loss": 0.6005,
+      "step": 830
+    },
+    {
+      "epoch": 0.9295302013422819,
+      "grad_norm": 0.41972237825393677,
+      "learning_rate": 4.765839524161518e-06,
+      "loss": 0.582,
+      "step": 831
+    },
+    {
+      "epoch": 0.930648769574944,
+      "grad_norm": 0.4424271881580353,
+      "learning_rate": 4.765208656383299e-06,
+      "loss": 0.5978,
+      "step": 832
+    },
+    {
+      "epoch": 0.9317673378076062,
+      "grad_norm": 0.45667219161987305,
+      "learning_rate": 4.7645769817859554e-06,
+      "loss": 0.6208,
+      "step": 833
+    },
+    {
+      "epoch": 0.9328859060402684,
+      "grad_norm": 0.4423377811908722,
+      "learning_rate": 4.763944500594476e-06,
+      "loss": 0.6061,
+      "step": 834
+    },
+    {
+      "epoch": 0.9340044742729307,
+      "grad_norm": 0.4316536784172058,
+      "learning_rate": 4.7633112130341385e-06,
+      "loss": 0.6116,
+      "step": 835
+    },
+    {
+      "epoch": 0.9351230425055929,
+      "grad_norm": 0.4591672718524933,
+      "learning_rate": 4.762677119330505e-06,
+      "loss": 0.5729,
+      "step": 836
+    },
+    {
+      "epoch": 0.9362416107382551,
+      "grad_norm": 0.4469880759716034,
+      "learning_rate": 4.762042219709427e-06,
+      "loss": 0.6025,
+      "step": 837
+    },
+    {
+      "epoch": 0.9373601789709173,
+      "grad_norm": 0.4560692012310028,
+      "learning_rate": 4.761406514397042e-06,
+      "loss": 0.6103,
+      "step": 838
+    },
+    {
+      "epoch": 0.9384787472035794,
+      "grad_norm": 0.4428820013999939,
+      "learning_rate": 4.760770003619775e-06,
+      "loss": 0.6258,
+      "step": 839
+    },
+    {
+      "epoch": 0.9395973154362416,
+      "grad_norm": 0.44238874316215515,
+      "learning_rate": 4.760132687604338e-06,
+      "loss": 0.6032,
+      "step": 840
+    },
+    {
+      "epoch": 0.9407158836689038,
+      "grad_norm": 0.46432724595069885,
+      "learning_rate": 4.759494566577727e-06,
+      "loss": 0.6266,
+      "step": 841
+    },
+    {
+      "epoch": 0.941834451901566,
+      "grad_norm": 0.42941901087760925,
+      "learning_rate": 4.75885564076723e-06,
+      "loss": 0.5927,
+      "step": 842
+    },
+    {
+      "epoch": 0.9429530201342282,
+      "grad_norm": 0.43781232833862305,
+      "learning_rate": 4.758215910400418e-06,
+      "loss": 0.5967,
+      "step": 843
+    },
+    {
+      "epoch": 0.9440715883668904,
+      "grad_norm": 0.45641666650772095,
+      "learning_rate": 4.757575375705149e-06,
+      "loss": 0.6423,
+      "step": 844
+    },
+    {
+      "epoch": 0.9451901565995525,
+      "grad_norm": 0.43784114718437195,
+      "learning_rate": 4.756934036909567e-06,
+      "loss": 0.606,
+      "step": 845
+    },
+    {
+      "epoch": 0.9463087248322147,
+      "grad_norm": 0.4379528760910034,
+      "learning_rate": 4.756291894242106e-06,
+      "loss": 0.6201,
+      "step": 846
+    },
+    {
+      "epoch": 0.9474272930648769,
+      "grad_norm": 0.42831459641456604,
+      "learning_rate": 4.755648947931479e-06,
+      "loss": 0.6121,
+      "step": 847
+    },
+    {
+      "epoch": 0.9485458612975392,
+      "grad_norm": 0.43790462613105774,
+      "learning_rate": 4.7550051982066945e-06,
+      "loss": 0.5785,
+      "step": 848
+    },
+    {
+      "epoch": 0.9496644295302014,
+      "grad_norm": 0.4407269358634949,
+      "learning_rate": 4.75436064529704e-06,
+      "loss": 0.6127,
+      "step": 849
+    },
+    {
+      "epoch": 0.9507829977628636,
+      "grad_norm": 0.4252265393733978,
+      "learning_rate": 4.753715289432092e-06,
+      "loss": 0.6129,
+      "step": 850
+    },
+    {
+      "epoch": 0.9519015659955258,
+      "grad_norm": 0.4376990795135498,
+      "learning_rate": 4.753069130841712e-06,
+      "loss": 0.614,
+      "step": 851
+    },
+    {
+      "epoch": 0.9530201342281879,
+      "grad_norm": 0.43123552203178406,
+      "learning_rate": 4.752422169756048e-06,
+      "loss": 0.6169,
+      "step": 852
+    },
+    {
+      "epoch": 0.9541387024608501,
+      "grad_norm": 0.4513196349143982,
+      "learning_rate": 4.7517744064055345e-06,
+      "loss": 0.6381,
+      "step": 853
+    },
+    {
+      "epoch": 0.9552572706935123,
+      "grad_norm": 0.44663751125335693,
+      "learning_rate": 4.751125841020891e-06,
+      "loss": 0.605,
+      "step": 854
+    },
+    {
+      "epoch": 0.9563758389261745,
+      "grad_norm": 0.44196903705596924,
+      "learning_rate": 4.750476473833123e-06,
+      "loss": 0.6163,
+      "step": 855
+    },
+    {
+      "epoch": 0.9574944071588367,
+      "grad_norm": 0.40786847472190857,
+      "learning_rate": 4.74982630507352e-06,
+      "loss": 0.5624,
+      "step": 856
+    },
+    {
+      "epoch": 0.9586129753914989,
+      "grad_norm": 0.43438002467155457,
+      "learning_rate": 4.749175334973659e-06,
+      "loss": 0.6183,
+      "step": 857
+    },
+    {
+      "epoch": 0.959731543624161,
+      "grad_norm": 0.43120619654655457,
+      "learning_rate": 4.748523563765401e-06,
+      "loss": 0.6097,
+      "step": 858
+    },
+    {
+      "epoch": 0.9608501118568232,
+      "grad_norm": 0.4761989414691925,
+      "learning_rate": 4.747870991680895e-06,
+      "loss": 0.6029,
+      "step": 859
+    },
+    {
+      "epoch": 0.9619686800894854,
+      "grad_norm": 0.44484785199165344,
+      "learning_rate": 4.747217618952571e-06,
+      "loss": 0.5955,
+      "step": 860
+    },
+    {
+      "epoch": 0.9630872483221476,
+      "grad_norm": 0.4473284184932709,
+      "learning_rate": 4.746563445813148e-06,
+      "loss": 0.6367,
+      "step": 861
+    },
+    {
+      "epoch": 0.9642058165548099,
+      "grad_norm": 0.4486042857170105,
+      "learning_rate": 4.745908472495628e-06,
+      "loss": 0.5917,
+      "step": 862
+    },
+    {
+      "epoch": 0.9653243847874721,
+      "grad_norm": 0.45661088824272156,
+      "learning_rate": 4.745252699233298e-06,
+      "loss": 0.61,
+      "step": 863
+    },
+    {
+      "epoch": 0.9664429530201343,
+      "grad_norm": 0.4235589802265167,
+      "learning_rate": 4.744596126259731e-06,
+      "loss": 0.5887,
+      "step": 864
+    },
+    {
+      "epoch": 0.9675615212527964,
+      "grad_norm": 0.45463138818740845,
+      "learning_rate": 4.743938753808785e-06,
+      "loss": 0.6295,
+      "step": 865
+    },
+    {
+      "epoch": 0.9686800894854586,
+      "grad_norm": 0.4576405882835388,
+      "learning_rate": 4.743280582114601e-06,
+      "loss": 0.6301,
+      "step": 866
+    },
+    {
+      "epoch": 0.9697986577181208,
+      "grad_norm": 0.4380008280277252,
+      "learning_rate": 4.742621611411606e-06,
+      "loss": 0.619,
+      "step": 867
+    },
+    {
+      "epoch": 0.970917225950783,
+      "grad_norm": 0.4485333263874054,
+      "learning_rate": 4.7419618419345124e-06,
+      "loss": 0.6311,
+      "step": 868
+    },
+    {
+      "epoch": 0.9720357941834452,
+      "grad_norm": 0.44375064969062805,
+      "learning_rate": 4.741301273918314e-06,
+      "loss": 0.6095,
+      "step": 869
+    },
+    {
+      "epoch": 0.9731543624161074,
+      "grad_norm": 0.4440104365348816,
+      "learning_rate": 4.740639907598293e-06,
+      "loss": 0.6173,
+      "step": 870
+    },
+    {
+      "epoch": 0.9742729306487695,
+      "grad_norm": 0.4545641243457794,
+      "learning_rate": 4.739977743210014e-06,
+      "loss": 0.6046,
+      "step": 871
+    },
+    {
+      "epoch": 0.9753914988814317,
+      "grad_norm": 0.41214534640312195,
+      "learning_rate": 4.739314780989324e-06,
+      "loss": 0.6072,
+      "step": 872
+    },
+    {
+      "epoch": 0.9765100671140939,
+      "grad_norm": 0.4223095178604126,
+      "learning_rate": 4.738651021172357e-06,
+      "loss": 0.5878,
+      "step": 873
+    },
+    {
+      "epoch": 0.9776286353467561,
+      "grad_norm": 0.42885127663612366,
+      "learning_rate": 4.7379864639955304e-06,
+      "loss": 0.577,
+      "step": 874
+    },
+    {
+      "epoch": 0.9787472035794184,
+      "grad_norm": 0.42921069264411926,
+      "learning_rate": 4.737321109695546e-06,
+      "loss": 0.5844,
+      "step": 875
+    },
+    {
+      "epoch": 0.9798657718120806,
+      "grad_norm": 0.43462637066841125,
+      "learning_rate": 4.736654958509387e-06,
+      "loss": 0.6135,
+      "step": 876
+    },
+    {
+      "epoch": 0.9809843400447428,
+      "grad_norm": 0.43558555841445923,
+      "learning_rate": 4.735988010674324e-06,
+      "loss": 0.6255,
+      "step": 877
+    },
+    {
+      "epoch": 0.9821029082774049,
+      "grad_norm": 0.44332823157310486,
+      "learning_rate": 4.735320266427909e-06,
+      "loss": 0.6266,
+      "step": 878
+    },
+    {
+      "epoch": 0.9832214765100671,
+      "grad_norm": 0.4158824682235718,
+      "learning_rate": 4.734651726007978e-06,
+      "loss": 0.585,
+      "step": 879
+    },
+    {
+      "epoch": 0.9843400447427293,
+      "grad_norm": 0.4264974296092987,
+      "learning_rate": 4.733982389652652e-06,
+      "loss": 0.5871,
+      "step": 880
+    },
+    {
+      "epoch": 0.9854586129753915,
+      "grad_norm": 0.44846397638320923,
+      "learning_rate": 4.733312257600332e-06,
+      "loss": 0.6441,
+      "step": 881
+    },
+    {
+      "epoch": 0.9865771812080537,
+      "grad_norm": 0.46592265367507935,
+      "learning_rate": 4.732641330089707e-06,
+      "loss": 0.6326,
+      "step": 882
+    },
+    {
+      "epoch": 0.9876957494407159,
+      "grad_norm": 0.423447847366333,
+      "learning_rate": 4.731969607359747e-06,
+      "loss": 0.5922,
+      "step": 883
+    },
+    {
+      "epoch": 0.9888143176733781,
+      "grad_norm": 0.43406492471694946,
+      "learning_rate": 4.731297089649704e-06,
+      "loss": 0.6234,
+      "step": 884
+    },
+    {
+      "epoch": 0.9899328859060402,
+      "grad_norm": 0.443352073431015,
+      "learning_rate": 4.730623777199115e-06,
+      "loss": 0.6397,
+      "step": 885
+    },
+    {
+      "epoch": 0.9910514541387024,
+      "grad_norm": 0.4311400353908539,
+      "learning_rate": 4.7299496702478e-06,
+      "loss": 0.6073,
+      "step": 886
+    },
+    {
+      "epoch": 0.9921700223713646,
+      "grad_norm": 0.4217356741428375,
+      "learning_rate": 4.729274769035861e-06,
+      "loss": 0.6177,
+      "step": 887
+    },
+    {
+      "epoch": 0.9932885906040269,
+      "grad_norm": 0.45120054483413696,
+      "learning_rate": 4.728599073803685e-06,
+      "loss": 0.6181,
+      "step": 888
+    },
+    {
+      "epoch": 0.9944071588366891,
+      "grad_norm": 0.4567187428474426,
+      "learning_rate": 4.7279225847919375e-06,
+      "loss": 0.5839,
+      "step": 889
+    },
+    {
+      "epoch": 0.9955257270693513,
+      "grad_norm": 0.4429774880409241,
+      "learning_rate": 4.727245302241572e-06,
+      "loss": 0.6033,
+      "step": 890
+    },
+    {
+      "epoch": 0.9966442953020134,
+      "grad_norm": 0.44755569100379944,
+      "learning_rate": 4.726567226393821e-06,
+      "loss": 0.5877,
+      "step": 891
+    },
+    {
+      "epoch": 0.9977628635346756,
+      "grad_norm": 0.45908474922180176,
+      "learning_rate": 4.725888357490201e-06,
+      "loss": 0.6017,
+      "step": 892
+    },
+    {
+      "epoch": 0.9988814317673378,
+      "grad_norm": 0.4454707205295563,
+      "learning_rate": 4.725208695772511e-06,
+      "loss": 0.6007,
+      "step": 893
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.43688732385635376,
+      "learning_rate": 4.7245282414828305e-06,
+      "loss": 0.6202,
+      "step": 894
+    },
+    {
+      "epoch": 1.0011185682326622,
+      "grad_norm": 0.43844303488731384,
+      "learning_rate": 4.723846994863524e-06,
+      "loss": 0.5886,
+      "step": 895
+    },
+    {
+      "epoch": 1.0022371364653244,
+      "grad_norm": 0.4296126067638397,
+      "learning_rate": 4.7231649561572376e-06,
+      "loss": 0.5759,
+      "step": 896
+    },
+    {
+      "epoch": 1.0033557046979866,
+      "grad_norm": 0.4716373682022095,
+      "learning_rate": 4.722482125606898e-06,
+      "loss": 0.6078,
+      "step": 897
+    },
+    {
+      "epoch": 1.0044742729306488,
+      "grad_norm": 0.45032379031181335,
+      "learning_rate": 4.721798503455716e-06,
+      "loss": 0.6145,
+      "step": 898
+    },
+    {
+      "epoch": 1.005592841163311,
+      "grad_norm": 0.42808797955513,
+      "learning_rate": 4.721114089947181e-06,
+      "loss": 0.5446,
+      "step": 899
+    },
+    {
+      "epoch": 1.0067114093959733,
+      "grad_norm": 0.4501192271709442,
+      "learning_rate": 4.7204288853250694e-06,
+      "loss": 0.5951,
+      "step": 900
+    },
+    {
+      "epoch": 1.0078299776286352,
+      "grad_norm": 0.43108487129211426,
+      "learning_rate": 4.719742889833434e-06,
+      "loss": 0.5826,
+      "step": 901
+    },
+    {
+      "epoch": 1.0089485458612975,
+      "grad_norm": 0.44014376401901245,
+      "learning_rate": 4.7190561037166135e-06,
+      "loss": 0.6138,
+      "step": 902
+    },
+    {
+      "epoch": 1.0100671140939597,
+      "grad_norm": 0.4504956305027008,
+      "learning_rate": 4.718368527219226e-06,
+      "loss": 0.6125,
+      "step": 903
+    },
+    {
+      "epoch": 1.0111856823266219,
+      "grad_norm": 0.4215417504310608,
+      "learning_rate": 4.717680160586172e-06,
+      "loss": 0.5761,
+      "step": 904
+    },
+    {
+      "epoch": 1.012304250559284,
+      "grad_norm": 0.44780582189559937,
+      "learning_rate": 4.716991004062632e-06,
+      "loss": 0.6056,
+      "step": 905
+    },
+    {
+      "epoch": 1.0134228187919463,
+      "grad_norm": 0.462046355009079,
+      "learning_rate": 4.7163010578940695e-06,
+      "loss": 0.6188,
+      "step": 906
+    },
+    {
+      "epoch": 1.0145413870246085,
+      "grad_norm": 0.42324090003967285,
+      "learning_rate": 4.715610322326229e-06,
+      "loss": 0.6086,
+      "step": 907
+    },
+    {
+      "epoch": 1.0156599552572707,
+      "grad_norm": 0.4330340623855591,
+      "learning_rate": 4.714918797605135e-06,
+      "loss": 0.5894,
+      "step": 908
+    },
+    {
+      "epoch": 1.016778523489933,
+      "grad_norm": 0.4506567418575287,
+      "learning_rate": 4.714226483977095e-06,
+      "loss": 0.6048,
+      "step": 909
+    },
+    {
+      "epoch": 1.0178970917225951,
+      "grad_norm": 0.4513271152973175,
+      "learning_rate": 4.713533381688695e-06,
+      "loss": 0.6139,
+      "step": 910
+    },
+    {
+      "epoch": 1.0190156599552573,
+      "grad_norm": 0.46496498584747314,
+      "learning_rate": 4.712839490986804e-06,
+      "loss": 0.607,
+      "step": 911
+    },
+    {
+      "epoch": 1.0201342281879195,
+      "grad_norm": 0.4421592354774475,
+      "learning_rate": 4.7121448121185716e-06,
+      "loss": 0.6044,
+      "step": 912
+    },
+    {
+      "epoch": 1.0212527964205818,
+      "grad_norm": 0.4436202943325043,
+      "learning_rate": 4.711449345331427e-06,
+      "loss": 0.6216,
+      "step": 913
+    },
+    {
+      "epoch": 1.0223713646532437,
+      "grad_norm": 0.43512076139450073,
+      "learning_rate": 4.7107530908730815e-06,
+      "loss": 0.5738,
+      "step": 914
+    },
+    {
+      "epoch": 1.023489932885906,
+      "grad_norm": 0.4467792510986328,
+      "learning_rate": 4.710056048991525e-06,
+      "loss": 0.6205,
+      "step": 915
+    },
+    {
+      "epoch": 1.0246085011185682,
+      "grad_norm": 0.4396659731864929,
+      "learning_rate": 4.709358219935028e-06,
+      "loss": 0.5902,
+      "step": 916
+    },
+    {
+      "epoch": 1.0257270693512304,
+      "grad_norm": 0.4392754137516022,
+      "learning_rate": 4.708659603952146e-06,
+      "loss": 0.6032,
+      "step": 917
+    },
+    {
+      "epoch": 1.0268456375838926,
+      "grad_norm": 0.4519026577472687,
+      "learning_rate": 4.707960201291708e-06,
+      "loss": 0.6024,
+      "step": 918
+    },
+    {
+      "epoch": 1.0279642058165548,
+      "grad_norm": 0.45630815625190735,
+      "learning_rate": 4.707260012202826e-06,
+      "loss": 0.5928,
+      "step": 919
+    },
+    {
+      "epoch": 1.029082774049217,
+      "grad_norm": 0.43156692385673523,
+      "learning_rate": 4.706559036934896e-06,
+      "loss": 0.5819,
+      "step": 920
+    },
+    {
+      "epoch": 1.0302013422818792,
+      "grad_norm": 0.4461636543273926,
+      "learning_rate": 4.705857275737587e-06,
+      "loss": 0.5727,
+      "step": 921
+    },
+    {
+      "epoch": 1.0313199105145414,
+      "grad_norm": 0.43878501653671265,
+      "learning_rate": 4.705154728860853e-06,
+      "loss": 0.5784,
+      "step": 922
+    },
+    {
+      "epoch": 1.0324384787472036,
+      "grad_norm": 0.42566847801208496,
+      "learning_rate": 4.704451396554925e-06,
+      "loss": 0.5607,
+      "step": 923
+    },
+    {
+      "epoch": 1.0335570469798658,
+      "grad_norm": 0.43960490822792053,
+      "learning_rate": 4.703747279070318e-06,
+      "loss": 0.6085,
+      "step": 924
+    },
+    {
+      "epoch": 1.034675615212528,
+      "grad_norm": 0.4411300718784332,
+      "learning_rate": 4.7030423766578194e-06,
+      "loss": 0.5821,
+      "step": 925
+    },
+    {
+      "epoch": 1.0357941834451903,
+      "grad_norm": 0.43742841482162476,
+      "learning_rate": 4.702336689568503e-06,
+      "loss": 0.5791,
+      "step": 926
+    },
+    {
+      "epoch": 1.0369127516778525,
+      "grad_norm": 0.4498448669910431,
+      "learning_rate": 4.70163021805372e-06,
+      "loss": 0.5803,
+      "step": 927
+    },
+    {
+      "epoch": 1.0380313199105144,
+      "grad_norm": 0.44952791929244995,
+      "learning_rate": 4.7009229623650986e-06,
+      "loss": 0.5954,
+      "step": 928
+    },
+    {
+      "epoch": 1.0391498881431767,
+      "grad_norm": 0.44882333278656006,
+      "learning_rate": 4.7002149227545505e-06,
+      "loss": 0.5866,
+      "step": 929
+    },
+    {
+      "epoch": 1.0402684563758389,
+      "grad_norm": 0.43540847301483154,
+      "learning_rate": 4.699506099474261e-06,
+      "loss": 0.5742,
+      "step": 930
+    },
+    {
+      "epoch": 1.041387024608501,
+      "grad_norm": 0.44182834029197693,
+      "learning_rate": 4.6987964927767015e-06,
+      "loss": 0.5712,
+      "step": 931
+    },
+    {
+      "epoch": 1.0425055928411633,
+      "grad_norm": 0.43853750824928284,
+      "learning_rate": 4.6980861029146174e-06,
+      "loss": 0.5847,
+      "step": 932
+    },
+    {
+      "epoch": 1.0436241610738255,
+      "grad_norm": 0.4608674943447113,
+      "learning_rate": 4.697374930141034e-06,
+      "loss": 0.6076,
+      "step": 933
+    },
+    {
+      "epoch": 1.0447427293064877,
+      "grad_norm": 0.45616960525512695,
+      "learning_rate": 4.696662974709256e-06,
+      "loss": 0.6088,
+      "step": 934
+    },
+    {
+      "epoch": 1.04586129753915,
+      "grad_norm": 0.44517964124679565,
+      "learning_rate": 4.695950236872867e-06,
+      "loss": 0.5795,
+      "step": 935
+    },
+    {
+      "epoch": 1.0469798657718121,
+      "grad_norm": 0.44067806005477905,
+      "learning_rate": 4.69523671688573e-06,
+      "loss": 0.6035,
+      "step": 936
+    },
+    {
+      "epoch": 1.0480984340044743,
+      "grad_norm": 0.44141000509262085,
+      "learning_rate": 4.694522415001984e-06,
+      "loss": 0.5676,
+      "step": 937
+    },
+    {
+      "epoch": 1.0492170022371365,
+      "grad_norm": 0.47550198435783386,
+      "learning_rate": 4.693807331476049e-06,
+      "loss": 0.5774,
+      "step": 938
+    },
+    {
+      "epoch": 1.0503355704697988,
+      "grad_norm": 0.4472981095314026,
+      "learning_rate": 4.6930914665626215e-06,
+      "loss": 0.5884,
+      "step": 939
+    },
+    {
+      "epoch": 1.0514541387024607,
+      "grad_norm": 0.4382828176021576,
+      "learning_rate": 4.692374820516679e-06,
+      "loss": 0.5941,
+      "step": 940
+    },
+    {
+      "epoch": 1.052572706935123,
+      "grad_norm": 0.4618227481842041,
+      "learning_rate": 4.691657393593475e-06,
+      "loss": 0.6139,
+      "step": 941
+    },
+    {
+      "epoch": 1.0536912751677852,
+      "grad_norm": 0.44602257013320923,
+      "learning_rate": 4.690939186048541e-06,
+      "loss": 0.5824,
+      "step": 942
+    },
+    {
+      "epoch": 1.0548098434004474,
+      "grad_norm": 0.4528485834598541,
+      "learning_rate": 4.690220198137688e-06,
+      "loss": 0.6059,
+      "step": 943
+    },
+    {
+      "epoch": 1.0559284116331096,
+      "grad_norm": 0.4452129304409027,
+      "learning_rate": 4.689500430117005e-06,
+      "loss": 0.5792,
+      "step": 944
+    },
+    {
+      "epoch": 1.0570469798657718,
+      "grad_norm": 0.4420737028121948,
+      "learning_rate": 4.688779882242855e-06,
+      "loss": 0.5834,
+      "step": 945
+    },
+    {
+      "epoch": 1.058165548098434,
+      "grad_norm": 0.46812739968299866,
+      "learning_rate": 4.6880585547718845e-06,
+      "loss": 0.6169,
+      "step": 946
+    },
+    {
+      "epoch": 1.0592841163310962,
+      "grad_norm": 0.4635341465473175,
+      "learning_rate": 4.687336447961015e-06,
+      "loss": 0.5954,
+      "step": 947
+    },
+    {
+      "epoch": 1.0604026845637584,
+      "grad_norm": 0.45534729957580566,
+      "learning_rate": 4.686613562067444e-06,
+      "loss": 0.5906,
+      "step": 948
+    },
+    {
+      "epoch": 1.0615212527964206,
+      "grad_norm": 0.4358477294445038,
+      "learning_rate": 4.685889897348649e-06,
+      "loss": 0.5701,
+      "step": 949
+    },
+    {
+      "epoch": 1.0626398210290828,
+      "grad_norm": 0.4336836338043213,
+      "learning_rate": 4.685165454062385e-06,
+      "loss": 0.5552,
+      "step": 950
+    },
+    {
+      "epoch": 1.063758389261745,
+      "grad_norm": 0.4392858147621155,
+      "learning_rate": 4.684440232466682e-06,
+      "loss": 0.572,
+      "step": 951
+    },
+    {
+      "epoch": 1.0648769574944073,
+      "grad_norm": 0.4643237292766571,
+      "learning_rate": 4.683714232819848e-06,
+      "loss": 0.6037,
+      "step": 952
+    },
+    {
+      "epoch": 1.0659955257270695,
+      "grad_norm": 0.45478659868240356,
+      "learning_rate": 4.682987455380469e-06,
+      "loss": 0.6188,
+      "step": 953
+    },
+    {
+      "epoch": 1.0671140939597314,
+      "grad_norm": 0.4337981939315796,
+      "learning_rate": 4.682259900407409e-06,
+      "loss": 0.5715,
+      "step": 954
+    },
+    {
+      "epoch": 1.0682326621923937,
+      "grad_norm": 0.4659992456436157,
+      "learning_rate": 4.6815315681598065e-06,
+      "loss": 0.5939,
+      "step": 955
+    },
+    {
+      "epoch": 1.0693512304250559,
+      "grad_norm": 0.4571925103664398,
+      "learning_rate": 4.680802458897078e-06,
+      "loss": 0.605,
+      "step": 956
+    },
+    {
+      "epoch": 1.070469798657718,
+      "grad_norm": 0.4513433575630188,
+      "learning_rate": 4.6800725728789164e-06,
+      "loss": 0.5856,
+      "step": 957
+    },
+    {
+      "epoch": 1.0715883668903803,
+      "grad_norm": 0.45815905928611755,
+      "learning_rate": 4.6793419103652914e-06,
+      "loss": 0.6006,
+      "step": 958
+    },
+    {
+      "epoch": 1.0727069351230425,
+      "grad_norm": 0.4600979685783386,
+      "learning_rate": 4.678610471616451e-06,
+      "loss": 0.5863,
+      "step": 959
+    },
+    {
+      "epoch": 1.0738255033557047,
+      "grad_norm": 0.4419516623020172,
+      "learning_rate": 4.677878256892917e-06,
+      "loss": 0.5796,
+      "step": 960
+    },
+    {
+      "epoch": 1.074944071588367,
+      "grad_norm": 0.458377867937088,
+      "learning_rate": 4.677145266455489e-06,
+      "loss": 0.5719,
+      "step": 961
+    },
+    {
+      "epoch": 1.0760626398210291,
+      "grad_norm": 0.445962131023407,
+      "learning_rate": 4.676411500565241e-06,
+      "loss": 0.585,
+      "step": 962
+    },
+    {
+      "epoch": 1.0771812080536913,
+      "grad_norm": 0.4639750123023987,
+      "learning_rate": 4.675676959483528e-06,
+      "loss": 0.6117,
+      "step": 963
+    },
+    {
+      "epoch": 1.0782997762863535,
+      "grad_norm": 0.4576922059059143,
+      "learning_rate": 4.6749416434719755e-06,
+      "loss": 0.592,
+      "step": 964
+    },
+    {
+      "epoch": 1.0794183445190157,
+      "grad_norm": 0.45511624217033386,
+      "learning_rate": 4.674205552792487e-06,
+      "loss": 0.5888,
+      "step": 965
+    },
+    {
+      "epoch": 1.0805369127516777,
+      "grad_norm": 0.4598134756088257,
+      "learning_rate": 4.673468687707244e-06,
+      "loss": 0.5705,
+      "step": 966
+    },
+    {
+      "epoch": 1.08165548098434,
+      "grad_norm": 0.4522278606891632,
+      "learning_rate": 4.672731048478702e-06,
+      "loss": 0.5741,
+      "step": 967
+    },
+    {
+      "epoch": 1.0827740492170022,
+      "grad_norm": 0.4419742822647095,
+      "learning_rate": 4.671992635369592e-06,
+      "loss": 0.5832,
+      "step": 968
+    },
+    {
+      "epoch": 1.0838926174496644,
+      "grad_norm": 0.4407685399055481,
+      "learning_rate": 4.67125344864292e-06,
+      "loss": 0.5978,
+      "step": 969
+    },
+    {
+      "epoch": 1.0850111856823266,
+      "grad_norm": 0.4517025053501129,
+      "learning_rate": 4.67051348856197e-06,
+      "loss": 0.6115,
+      "step": 970
+    },
+    {
+      "epoch": 1.0861297539149888,
+      "grad_norm": 0.4652605354785919,
+      "learning_rate": 4.6697727553903e-06,
+      "loss": 0.5838,
+      "step": 971
+    },
+    {
+      "epoch": 1.087248322147651,
+      "grad_norm": 0.45211100578308105,
+      "learning_rate": 4.6690312493917424e-06,
+      "loss": 0.5905,
+      "step": 972
+    },
+    {
+      "epoch": 1.0883668903803132,
+      "grad_norm": 0.45999446511268616,
+      "learning_rate": 4.668288970830407e-06,
+      "loss": 0.589,
+      "step": 973
+    },
+    {
+      "epoch": 1.0894854586129754,
+      "grad_norm": 0.4458955228328705,
+      "learning_rate": 4.667545919970676e-06,
+      "loss": 0.5927,
+      "step": 974
+    },
+    {
+      "epoch": 1.0906040268456376,
+      "grad_norm": 0.46779292821884155,
+      "learning_rate": 4.666802097077211e-06,
+      "loss": 0.5955,
+      "step": 975
+    },
+    {
+      "epoch": 1.0917225950782998,
+      "grad_norm": 0.4617951810359955,
+      "learning_rate": 4.666057502414942e-06,
+      "loss": 0.6117,
+      "step": 976
+    },
+    {
+      "epoch": 1.092841163310962,
+      "grad_norm": 0.45327237248420715,
+      "learning_rate": 4.665312136249082e-06,
+      "loss": 0.5863,
+      "step": 977
+    },
+    {
+      "epoch": 1.0939597315436242,
+      "grad_norm": 0.4517632722854614,
+      "learning_rate": 4.664565998845112e-06,
+      "loss": 0.5832,
+      "step": 978
+    },
+    {
+      "epoch": 1.0950782997762865,
+      "grad_norm": 0.43310609459877014,
+      "learning_rate": 4.663819090468791e-06,
+      "loss": 0.5647,
+      "step": 979
+    },
+    {
+      "epoch": 1.0961968680089484,
+      "grad_norm": 0.47837111353874207,
+      "learning_rate": 4.6630714113861505e-06,
+      "loss": 0.6195,
+      "step": 980
+    },
+    {
+      "epoch": 1.0973154362416107,
+      "grad_norm": 0.4558872878551483,
+      "learning_rate": 4.662322961863501e-06,
+      "loss": 0.5933,
+      "step": 981
+    },
+    {
+      "epoch": 1.0984340044742729,
+      "grad_norm": 0.45069193840026855,
+      "learning_rate": 4.661573742167421e-06,
+      "loss": 0.6066,
+      "step": 982
+    },
+    {
+      "epoch": 1.099552572706935,
+      "grad_norm": 0.44924792647361755,
+      "learning_rate": 4.660823752564769e-06,
+      "loss": 0.5773,
+      "step": 983
+    },
+    {
+      "epoch": 1.1006711409395973,
+      "grad_norm": 0.44701021909713745,
+      "learning_rate": 4.660072993322674e-06,
+      "loss": 0.5663,
+      "step": 984
+    },
+    {
+      "epoch": 1.1017897091722595,
+      "grad_norm": 0.4506109952926636,
+      "learning_rate": 4.659321464708541e-06,
+      "loss": 0.593,
+      "step": 985
+    },
+    {
+      "epoch": 1.1029082774049217,
+      "grad_norm": 0.4586380422115326,
+      "learning_rate": 4.658569166990048e-06,
+      "loss": 0.6014,
+      "step": 986
+    },
+    {
+      "epoch": 1.104026845637584,
+      "grad_norm": 0.45552054047584534,
+      "learning_rate": 4.657816100435147e-06,
+      "loss": 0.6043,
+      "step": 987
+    },
+    {
+      "epoch": 1.1051454138702461,
+      "grad_norm": 0.4566228985786438,
+      "learning_rate": 4.657062265312065e-06,
+      "loss": 0.6088,
+      "step": 988
+    },
+    {
+      "epoch": 1.1062639821029083,
+      "grad_norm": 0.45366841554641724,
+      "learning_rate": 4.6563076618893045e-06,
+      "loss": 0.5918,
+      "step": 989
+    },
+    {
+      "epoch": 1.1073825503355705,
+      "grad_norm": 0.46025100350379944,
+      "learning_rate": 4.6555522904356344e-06,
+      "loss": 0.5876,
+      "step": 990
+    },
+    {
+      "epoch": 1.1085011185682327,
+      "grad_norm": 0.46731919050216675,
+      "learning_rate": 4.654796151220106e-06,
+      "loss": 0.5992,
+      "step": 991
+    },
+    {
+      "epoch": 1.109619686800895,
+      "grad_norm": 0.45841652154922485,
+      "learning_rate": 4.654039244512036e-06,
+      "loss": 0.5833,
+      "step": 992
+    },
+    {
+      "epoch": 1.110738255033557,
+      "grad_norm": 0.46612823009490967,
+      "learning_rate": 4.653281570581023e-06,
+      "loss": 0.6137,
+      "step": 993
+    },
+    {
+      "epoch": 1.1118568232662192,
+      "grad_norm": 0.4462451934814453,
+      "learning_rate": 4.6525231296969305e-06,
+      "loss": 0.545,
+      "step": 994
+    },
+    {
+      "epoch": 1.1129753914988814,
+      "grad_norm": 0.4724692404270172,
+      "learning_rate": 4.651763922129901e-06,
+      "loss": 0.6144,
+      "step": 995
+    },
+    {
+      "epoch": 1.1140939597315436,
+      "grad_norm": 0.47240084409713745,
+      "learning_rate": 4.651003948150349e-06,
+      "loss": 0.6037,
+      "step": 996
+    },
+    {
+      "epoch": 1.1152125279642058,
+      "grad_norm": 0.44295522570610046,
+      "learning_rate": 4.650243208028958e-06,
+      "loss": 0.5571,
+      "step": 997
+    },
+    {
+      "epoch": 1.116331096196868,
+      "grad_norm": 0.4727005958557129,
+      "learning_rate": 4.649481702036691e-06,
+      "loss": 0.5851,
+      "step": 998
+    },
+    {
+      "epoch": 1.1174496644295302,
+      "grad_norm": 0.4520561397075653,
+      "learning_rate": 4.648719430444777e-06,
+      "loss": 0.6016,
+      "step": 999
+    },
+    {
+      "epoch": 1.1185682326621924,
+      "grad_norm": 0.5022072792053223,
+      "learning_rate": 4.647956393524723e-06,
+      "loss": 0.5993,
+      "step": 1000
+    },
+    {
+      "epoch": 1.1196868008948546,
+      "grad_norm": 0.45666182041168213,
+      "learning_rate": 4.647192591548305e-06,
+      "loss": 0.6092,
+      "step": 1001
+    },
+    {
+      "epoch": 1.1208053691275168,
+      "grad_norm": 0.4551488161087036,
+      "learning_rate": 4.646428024787575e-06,
+      "loss": 0.5938,
+      "step": 1002
+    },
+    {
+      "epoch": 1.121923937360179,
+      "grad_norm": 0.4763842821121216,
+      "learning_rate": 4.645662693514853e-06,
+      "loss": 0.6016,
+      "step": 1003
+    },
+    {
+      "epoch": 1.1230425055928412,
+      "grad_norm": 0.4607378840446472,
+      "learning_rate": 4.644896598002736e-06,
+      "loss": 0.5785,
+      "step": 1004
+    },
+    {
+      "epoch": 1.1241610738255035,
+      "grad_norm": 0.4715465009212494,
+      "learning_rate": 4.64412973852409e-06,
+      "loss": 0.588,
+      "step": 1005
+    },
+    {
+      "epoch": 1.1252796420581657,
+      "grad_norm": 0.47461339831352234,
+      "learning_rate": 4.643362115352053e-06,
+      "loss": 0.6296,
+      "step": 1006
+    },
+    {
+      "epoch": 1.1263982102908277,
+      "grad_norm": 0.4571302831172943,
+      "learning_rate": 4.642593728760038e-06,
+      "loss": 0.5493,
+      "step": 1007
+    },
+    {
+      "epoch": 1.1275167785234899,
+      "grad_norm": 0.46194958686828613,
+      "learning_rate": 4.641824579021726e-06,
+      "loss": 0.5884,
+      "step": 1008
+    },
+    {
+      "epoch": 1.128635346756152,
+      "grad_norm": 0.4678124487400055,
+      "learning_rate": 4.6410546664110736e-06,
+      "loss": 0.5967,
+      "step": 1009
+    },
+    {
+      "epoch": 1.1297539149888143,
+      "grad_norm": 0.45504289865493774,
+      "learning_rate": 4.640283991202306e-06,
+      "loss": 0.5696,
+      "step": 1010
+    },
+    {
+      "epoch": 1.1308724832214765,
+      "grad_norm": 0.46907755732536316,
+      "learning_rate": 4.639512553669921e-06,
+      "loss": 0.6261,
+      "step": 1011
+    },
+    {
+      "epoch": 1.1319910514541387,
+      "grad_norm": 0.46805256605148315,
+      "learning_rate": 4.63874035408869e-06,
+      "loss": 0.5696,
+      "step": 1012
+    },
+    {
+      "epoch": 1.133109619686801,
+      "grad_norm": 0.4621639549732208,
+      "learning_rate": 4.637967392733652e-06,
+      "loss": 0.563,
+      "step": 1013
+    },
+    {
+      "epoch": 1.1342281879194631,
+      "grad_norm": 0.43852928280830383,
+      "learning_rate": 4.6371936698801215e-06,
+      "loss": 0.5434,
+      "step": 1014
+    },
+    {
+      "epoch": 1.1353467561521253,
+      "grad_norm": 0.46553272008895874,
+      "learning_rate": 4.636419185803681e-06,
+      "loss": 0.5825,
+      "step": 1015
+    },
+    {
+      "epoch": 1.1364653243847875,
+      "grad_norm": 0.46866393089294434,
+      "learning_rate": 4.635643940780184e-06,
+      "loss": 0.5723,
+      "step": 1016
+    },
+    {
+      "epoch": 1.1375838926174497,
+      "grad_norm": 0.46932780742645264,
+      "learning_rate": 4.634867935085758e-06,
+      "loss": 0.5972,
+      "step": 1017
+    },
+    {
+      "epoch": 1.138702460850112,
+      "grad_norm": 0.45771515369415283,
+      "learning_rate": 4.634091168996801e-06,
+      "loss": 0.5871,
+      "step": 1018
+    },
+    {
+      "epoch": 1.139821029082774,
+      "grad_norm": 0.461588978767395,
+      "learning_rate": 4.633313642789976e-06,
+      "loss": 0.6246,
+      "step": 1019
+    },
+    {
+      "epoch": 1.1409395973154361,
+      "grad_norm": 0.47002744674682617,
+      "learning_rate": 4.632535356742226e-06,
+      "loss": 0.581,
+      "step": 1020
+    },
+    {
+      "epoch": 1.1420581655480984,
+      "grad_norm": 0.46632397174835205,
+      "learning_rate": 4.631756311130757e-06,
+      "loss": 0.6203,
+      "step": 1021
+    },
+    {
+      "epoch": 1.1431767337807606,
+      "grad_norm": 0.46423229575157166,
+      "learning_rate": 4.6309765062330504e-06,
+      "loss": 0.6074,
+      "step": 1022
+    },
+    {
+      "epoch": 1.1442953020134228,
+      "grad_norm": 0.47234588861465454,
+      "learning_rate": 4.630195942326855e-06,
+      "loss": 0.6076,
+      "step": 1023
+    },
+    {
+      "epoch": 1.145413870246085,
+      "grad_norm": 0.46395570039749146,
+      "learning_rate": 4.62941461969019e-06,
+      "loss": 0.5775,
+      "step": 1024
+    },
+    {
+      "epoch": 1.1465324384787472,
+      "grad_norm": 0.456720232963562,
+      "learning_rate": 4.628632538601349e-06,
+      "loss": 0.5924,
+      "step": 1025
+    },
+    {
+      "epoch": 1.1476510067114094,
+      "grad_norm": 0.4563044309616089,
+      "learning_rate": 4.62784969933889e-06,
+      "loss": 0.5883,
+      "step": 1026
+    },
+    {
+      "epoch": 1.1487695749440716,
+      "grad_norm": 0.46938541531562805,
+      "learning_rate": 4.627066102181645e-06,
+      "loss": 0.5602,
+      "step": 1027
+    },
+    {
+      "epoch": 1.1498881431767338,
+      "grad_norm": 0.48113343119621277,
+      "learning_rate": 4.626281747408713e-06,
+      "loss": 0.609,
+      "step": 1028
+    },
+    {
+      "epoch": 1.151006711409396,
+      "grad_norm": 0.46784207224845886,
+      "learning_rate": 4.6254966352994666e-06,
+      "loss": 0.6018,
+      "step": 1029
+    },
+    {
+      "epoch": 1.1521252796420582,
+      "grad_norm": 0.46613121032714844,
+      "learning_rate": 4.624710766133544e-06,
+      "loss": 0.5579,
+      "step": 1030
+    },
+    {
+      "epoch": 1.1532438478747205,
+      "grad_norm": 0.4715366065502167,
+      "learning_rate": 4.6239241401908575e-06,
+      "loss": 0.5767,
+      "step": 1031
+    },
+    {
+      "epoch": 1.1543624161073827,
+      "grad_norm": 0.48206454515457153,
+      "learning_rate": 4.623136757751584e-06,
+      "loss": 0.6118,
+      "step": 1032
+    },
+    {
+      "epoch": 1.1554809843400446,
+      "grad_norm": 0.46678969264030457,
+      "learning_rate": 4.622348619096174e-06,
+      "loss": 0.5876,
+      "step": 1033
+    },
+    {
+      "epoch": 1.1565995525727069,
+      "grad_norm": 0.4533562958240509,
+      "learning_rate": 4.621559724505346e-06,
+      "loss": 0.6038,
+      "step": 1034
+    },
+    {
+      "epoch": 1.157718120805369,
+      "grad_norm": 0.473532497882843,
+      "learning_rate": 4.620770074260084e-06,
+      "loss": 0.5802,
+      "step": 1035
+    },
+    {
+      "epoch": 1.1588366890380313,
+      "grad_norm": 0.4734630882740021,
+      "learning_rate": 4.6199796686416505e-06,
+      "loss": 0.5906,
+      "step": 1036
+    },
+    {
+      "epoch": 1.1599552572706935,
+      "grad_norm": 0.48316991329193115,
+      "learning_rate": 4.6191885079315665e-06,
+      "loss": 0.561,
+      "step": 1037
+    },
+    {
+      "epoch": 1.1610738255033557,
+      "grad_norm": 0.4646005630493164,
+      "learning_rate": 4.618396592411628e-06,
+      "loss": 0.5901,
+      "step": 1038
+    },
+    {
+      "epoch": 1.162192393736018,
+      "grad_norm": 0.4684715270996094,
+      "learning_rate": 4.617603922363899e-06,
+      "loss": 0.5915,
+      "step": 1039
+    },
+    {
+      "epoch": 1.1633109619686801,
+      "grad_norm": 0.46257463097572327,
+      "learning_rate": 4.6168104980707105e-06,
+      "loss": 0.577,
+      "step": 1040
+    },
+    {
+      "epoch": 1.1644295302013423,
+      "grad_norm": 0.4703442454338074,
+      "learning_rate": 4.616016319814664e-06,
+      "loss": 0.564,
+      "step": 1041
+    },
+    {
+      "epoch": 1.1655480984340045,
+      "grad_norm": 0.4818289577960968,
+      "learning_rate": 4.615221387878631e-06,
+      "loss": 0.5955,
+      "step": 1042
+    },
+    {
+      "epoch": 1.1666666666666667,
+      "grad_norm": 0.4658922553062439,
+      "learning_rate": 4.614425702545745e-06,
+      "loss": 0.5944,
+      "step": 1043
+    },
+    {
+      "epoch": 1.167785234899329,
+      "grad_norm": 0.4584915041923523,
+      "learning_rate": 4.613629264099416e-06,
+      "loss": 0.5625,
+      "step": 1044
+    },
+    {
+      "epoch": 1.168903803131991,
+      "grad_norm": 0.4565418064594269,
+      "learning_rate": 4.6128320728233164e-06,
+      "loss": 0.5942,
+      "step": 1045
+    },
+    {
+      "epoch": 1.1700223713646531,
+      "grad_norm": 0.48082879185676575,
+      "learning_rate": 4.61203412900139e-06,
+      "loss": 0.6366,
+      "step": 1046
+    },
+    {
+      "epoch": 1.1711409395973154,
+      "grad_norm": 0.47267577052116394,
+      "learning_rate": 4.611235432917845e-06,
+      "loss": 0.5804,
+      "step": 1047
+    },
+    {
+      "epoch": 1.1722595078299776,
+      "grad_norm": 0.48486819863319397,
+      "learning_rate": 4.610435984857163e-06,
+      "loss": 0.5755,
+      "step": 1048
+    },
+    {
+      "epoch": 1.1733780760626398,
+      "grad_norm": 0.48791539669036865,
+      "learning_rate": 4.609635785104089e-06,
+      "loss": 0.6168,
+      "step": 1049
+    },
+    {
+      "epoch": 1.174496644295302,
+      "grad_norm": 0.45250606536865234,
+      "learning_rate": 4.6088348339436365e-06,
+      "loss": 0.5833,
+      "step": 1050
+    },
+    {
+      "epoch": 1.1756152125279642,
+      "grad_norm": 0.4777594804763794,
+      "learning_rate": 4.608033131661089e-06,
+      "loss": 0.6048,
+      "step": 1051
+    },
+    {
+      "epoch": 1.1767337807606264,
+      "grad_norm": 0.47242894768714905,
+      "learning_rate": 4.607230678541993e-06,
+      "loss": 0.5945,
+      "step": 1052
+    },
+    {
+      "epoch": 1.1778523489932886,
+      "grad_norm": 0.47580021619796753,
+      "learning_rate": 4.606427474872169e-06,
+      "loss": 0.593,
+      "step": 1053
+    },
+    {
+      "epoch": 1.1789709172259508,
+      "grad_norm": 0.45570215582847595,
+      "learning_rate": 4.605623520937698e-06,
+      "loss": 0.5617,
+      "step": 1054
+    },
+    {
+      "epoch": 1.180089485458613,
+      "grad_norm": 0.4642491936683655,
+      "learning_rate": 4.604818817024932e-06,
+      "loss": 0.604,
+      "step": 1055
+    },
+    {
+      "epoch": 1.1812080536912752,
+      "grad_norm": 0.46808069944381714,
+      "learning_rate": 4.60401336342049e-06,
+      "loss": 0.5829,
+      "step": 1056
+    },
+    {
+      "epoch": 1.1823266219239374,
+      "grad_norm": 0.482730507850647,
+      "learning_rate": 4.603207160411257e-06,
+      "loss": 0.6246,
+      "step": 1057
+    },
+    {
+      "epoch": 1.1834451901565997,
+      "grad_norm": 0.4554171562194824,
+      "learning_rate": 4.6024002082843855e-06,
+      "loss": 0.5797,
+      "step": 1058
+    },
+    {
+      "epoch": 1.1845637583892619,
+      "grad_norm": 0.4731457233428955,
+      "learning_rate": 4.6015925073272945e-06,
+      "loss": 0.5883,
+      "step": 1059
+    },
+    {
+      "epoch": 1.1856823266219239,
+      "grad_norm": 0.45422542095184326,
+      "learning_rate": 4.600784057827672e-06,
+      "loss": 0.5914,
+      "step": 1060
+    },
+    {
+      "epoch": 1.186800894854586,
+      "grad_norm": 0.4871719181537628,
+      "learning_rate": 4.599974860073466e-06,
+      "loss": 0.6146,
+      "step": 1061
+    },
+    {
+      "epoch": 1.1879194630872483,
+      "grad_norm": 0.47315844893455505,
+      "learning_rate": 4.5991649143529e-06,
+      "loss": 0.6044,
+      "step": 1062
+    },
+    {
+      "epoch": 1.1890380313199105,
+      "grad_norm": 0.45965805649757385,
+      "learning_rate": 4.5983542209544565e-06,
+      "loss": 0.5628,
+      "step": 1063
+    },
+    {
+      "epoch": 1.1901565995525727,
+      "grad_norm": 0.4573105573654175,
+      "learning_rate": 4.5975427801668886e-06,
+      "loss": 0.581,
+      "step": 1064
+    },
+    {
+      "epoch": 1.191275167785235,
+      "grad_norm": 0.45358559489250183,
+      "learning_rate": 4.596730592279214e-06,
+      "loss": 0.5851,
+      "step": 1065
+    },
+    {
+      "epoch": 1.192393736017897,
+      "grad_norm": 0.45561692118644714,
+      "learning_rate": 4.595917657580716e-06,
+      "loss": 0.5636,
+      "step": 1066
+    },
+    {
+      "epoch": 1.1935123042505593,
+      "grad_norm": 0.45746201276779175,
+      "learning_rate": 4.5951039763609464e-06,
+      "loss": 0.5943,
+      "step": 1067
+    },
+    {
+      "epoch": 1.1946308724832215,
+      "grad_norm": 0.4627397656440735,
+      "learning_rate": 4.594289548909719e-06,
+      "loss": 0.598,
+      "step": 1068
+    },
+    {
+      "epoch": 1.1957494407158837,
+      "grad_norm": 0.4608605206012726,
+      "learning_rate": 4.593474375517118e-06,
+      "loss": 0.5861,
+      "step": 1069
+    },
+    {
+      "epoch": 1.196868008948546,
+      "grad_norm": 0.45715343952178955,
+      "learning_rate": 4.592658456473489e-06,
+      "loss": 0.6024,
+      "step": 1070
+    },
+    {
+      "epoch": 1.197986577181208,
+      "grad_norm": 0.4674588143825531,
+      "learning_rate": 4.591841792069445e-06,
+      "loss": 0.5779,
+      "step": 1071
+    },
+    {
+      "epoch": 1.1991051454138701,
+      "grad_norm": 0.4705957770347595,
+      "learning_rate": 4.591024382595864e-06,
+      "loss": 0.588,
+      "step": 1072
+    },
+    {
+      "epoch": 1.2002237136465324,
+      "grad_norm": 0.47805118560791016,
+      "learning_rate": 4.590206228343892e-06,
+      "loss": 0.5796,
+      "step": 1073
+    },
+    {
+      "epoch": 1.2013422818791946,
+      "grad_norm": 0.45588305592536926,
+      "learning_rate": 4.589387329604936e-06,
+      "loss": 0.5653,
+      "step": 1074
+    },
+    {
+      "epoch": 1.2024608501118568,
+      "grad_norm": 0.47811266779899597,
+      "learning_rate": 4.588567686670672e-06,
+      "loss": 0.6207,
+      "step": 1075
+    },
+    {
+      "epoch": 1.203579418344519,
+      "grad_norm": 0.46762150526046753,
+      "learning_rate": 4.587747299833039e-06,
+      "loss": 0.5697,
+      "step": 1076
+    },
+    {
+      "epoch": 1.2046979865771812,
+      "grad_norm": 0.46018633246421814,
+      "learning_rate": 4.586926169384239e-06,
+      "loss": 0.5772,
+      "step": 1077
+    },
+    {
+      "epoch": 1.2058165548098434,
+      "grad_norm": 0.46897298097610474,
+      "learning_rate": 4.586104295616744e-06,
+      "loss": 0.5733,
+      "step": 1078
+    },
+    {
+      "epoch": 1.2069351230425056,
+      "grad_norm": 0.466189980506897,
+      "learning_rate": 4.585281678823288e-06,
+      "loss": 0.6027,
+      "step": 1079
+    },
+    {
+      "epoch": 1.2080536912751678,
+      "grad_norm": 0.4684281647205353,
+      "learning_rate": 4.584458319296868e-06,
+      "loss": 0.5766,
+      "step": 1080
+    },
+    {
+      "epoch": 1.20917225950783,
+      "grad_norm": 0.4513910710811615,
+      "learning_rate": 4.583634217330747e-06,
+      "loss": 0.5891,
+      "step": 1081
+    },
+    {
+      "epoch": 1.2102908277404922,
+      "grad_norm": 0.476510226726532,
+      "learning_rate": 4.5828093732184545e-06,
+      "loss": 0.601,
+      "step": 1082
+    },
+    {
+      "epoch": 1.2114093959731544,
+      "grad_norm": 0.47996675968170166,
+      "learning_rate": 4.581983787253781e-06,
+      "loss": 0.598,
+      "step": 1083
+    },
+    {
+      "epoch": 1.2125279642058167,
+      "grad_norm": 0.4757002890110016,
+      "learning_rate": 4.581157459730783e-06,
+      "loss": 0.5561,
+      "step": 1084
+    },
+    {
+      "epoch": 1.2136465324384789,
+      "grad_norm": 0.45650506019592285,
+      "learning_rate": 4.5803303909437805e-06,
+      "loss": 0.5702,
+      "step": 1085
+    },
+    {
+      "epoch": 1.2147651006711409,
+      "grad_norm": 0.4679136872291565,
+      "learning_rate": 4.579502581187358e-06,
+      "loss": 0.5854,
+      "step": 1086
+    },
+    {
+      "epoch": 1.215883668903803,
+      "grad_norm": 0.46434950828552246,
+      "learning_rate": 4.578674030756364e-06,
+      "loss": 0.5623,
+      "step": 1087
+    },
+    {
+      "epoch": 1.2170022371364653,
+      "grad_norm": 0.4812433123588562,
+      "learning_rate": 4.57784473994591e-06,
+      "loss": 0.6006,
+      "step": 1088
+    },
+    {
+      "epoch": 1.2181208053691275,
+      "grad_norm": 0.4599030017852783,
+      "learning_rate": 4.577014709051372e-06,
+      "loss": 0.5614,
+      "step": 1089
+    },
+    {
+      "epoch": 1.2192393736017897,
+      "grad_norm": 0.48703038692474365,
+      "learning_rate": 4.576183938368389e-06,
+      "loss": 0.5987,
+      "step": 1090
+    },
+    {
+      "epoch": 1.220357941834452,
+      "grad_norm": 0.48404622077941895,
+      "learning_rate": 4.575352428192865e-06,
+      "loss": 0.592,
+      "step": 1091
+    },
+    {
+      "epoch": 1.221476510067114,
+      "grad_norm": 0.47787612676620483,
+      "learning_rate": 4.574520178820965e-06,
+      "loss": 0.5733,
+      "step": 1092
+    },
+    {
+      "epoch": 1.2225950782997763,
+      "grad_norm": 0.4672394394874573,
+      "learning_rate": 4.573687190549119e-06,
+      "loss": 0.5736,
+      "step": 1093
+    },
+    {
+      "epoch": 1.2237136465324385,
+      "grad_norm": 0.47233250737190247,
+      "learning_rate": 4.57285346367402e-06,
+      "loss": 0.6045,
+      "step": 1094
+    },
+    {
+      "epoch": 1.2248322147651007,
+      "grad_norm": 0.4758702516555786,
+      "learning_rate": 4.572018998492623e-06,
+      "loss": 0.5734,
+      "step": 1095
+    },
+    {
+      "epoch": 1.225950782997763,
+      "grad_norm": 0.4715832769870758,
+      "learning_rate": 4.571183795302147e-06,
+      "loss": 0.599,
+      "step": 1096
+    },
+    {
+      "epoch": 1.227069351230425,
+      "grad_norm": 0.4783090651035309,
+      "learning_rate": 4.570347854400074e-06,
+      "loss": 0.608,
+      "step": 1097
+    },
+    {
+      "epoch": 1.2281879194630871,
+      "grad_norm": 0.4721231162548065,
+      "learning_rate": 4.569511176084148e-06,
+      "loss": 0.5964,
+      "step": 1098
+    },
+    {
+      "epoch": 1.2293064876957494,
+      "grad_norm": 0.47731295228004456,
+      "learning_rate": 4.568673760652377e-06,
+      "loss": 0.5968,
+      "step": 1099
+    },
+    {
+      "epoch": 1.2304250559284116,
+      "grad_norm": 0.4623177945613861,
+      "learning_rate": 4.5678356084030286e-06,
+      "loss": 0.5669,
+      "step": 1100
+    },
+    {
+      "epoch": 1.2315436241610738,
+      "grad_norm": 0.46821466088294983,
+      "learning_rate": 4.566996719634636e-06,
+      "loss": 0.5895,
+      "step": 1101
+    },
+    {
+      "epoch": 1.232662192393736,
+      "grad_norm": 0.47642552852630615,
+      "learning_rate": 4.566157094645994e-06,
+      "loss": 0.5645,
+      "step": 1102
+    },
+    {
+      "epoch": 1.2337807606263982,
+      "grad_norm": 0.46859392523765564,
+      "learning_rate": 4.565316733736159e-06,
+      "loss": 0.6204,
+      "step": 1103
+    },
+    {
+      "epoch": 1.2348993288590604,
+      "grad_norm": 0.47033944725990295,
+      "learning_rate": 4.564475637204449e-06,
+      "loss": 0.5884,
+      "step": 1104
+    },
+    {
+      "epoch": 1.2360178970917226,
+      "grad_norm": 0.500777006149292,
+      "learning_rate": 4.563633805350443e-06,
+      "loss": 0.5922,
+      "step": 1105
+    },
+    {
+      "epoch": 1.2371364653243848,
+      "grad_norm": 0.49536949396133423,
+      "learning_rate": 4.562791238473988e-06,
+      "loss": 0.5953,
+      "step": 1106
+    },
+    {
+      "epoch": 1.238255033557047,
+      "grad_norm": 0.47423887252807617,
+      "learning_rate": 4.5619479368751855e-06,
+      "loss": 0.5772,
+      "step": 1107
+    },
+    {
+      "epoch": 1.2393736017897092,
+      "grad_norm": 0.5001300573348999,
+      "learning_rate": 4.561103900854401e-06,
+      "loss": 0.6128,
+      "step": 1108
+    },
+    {
+      "epoch": 1.2404921700223714,
+      "grad_norm": 0.4864799380302429,
+      "learning_rate": 4.560259130712264e-06,
+      "loss": 0.5821,
+      "step": 1109
+    },
+    {
+      "epoch": 1.2416107382550337,
+      "grad_norm": 0.5224222540855408,
+      "learning_rate": 4.559413626749662e-06,
+      "loss": 0.6192,
+      "step": 1110
+    },
+    {
+      "epoch": 1.2427293064876959,
+      "grad_norm": 0.4829416275024414,
+      "learning_rate": 4.558567389267748e-06,
+      "loss": 0.5825,
+      "step": 1111
+    },
+    {
+      "epoch": 1.2438478747203578,
+      "grad_norm": 0.46932196617126465,
+      "learning_rate": 4.557720418567931e-06,
+      "loss": 0.5887,
+      "step": 1112
+    },
+    {
+      "epoch": 1.24496644295302,
+      "grad_norm": 0.47189292311668396,
+      "learning_rate": 4.556872714951886e-06,
+      "loss": 0.5777,
+      "step": 1113
+    },
+    {
+      "epoch": 1.2460850111856823,
+      "grad_norm": 0.47725993394851685,
+      "learning_rate": 4.5560242787215444e-06,
+      "loss": 0.5801,
+      "step": 1114
+    },
+    {
+      "epoch": 1.2472035794183445,
+      "grad_norm": 0.4747818112373352,
+      "learning_rate": 4.555175110179104e-06,
+      "loss": 0.5833,
+      "step": 1115
+    },
+    {
+      "epoch": 1.2483221476510067,
+      "grad_norm": 0.48167556524276733,
+      "learning_rate": 4.554325209627019e-06,
+      "loss": 0.5917,
+      "step": 1116
+    },
+    {
+      "epoch": 1.249440715883669,
+      "grad_norm": 0.4757511615753174,
+      "learning_rate": 4.553474577368006e-06,
+      "loss": 0.5955,
+      "step": 1117
+    },
+    {
+      "epoch": 1.250559284116331,
+      "grad_norm": 0.4677585959434509,
+      "learning_rate": 4.552623213705043e-06,
+      "loss": 0.5864,
+      "step": 1118
+    },
+    {
+      "epoch": 1.2516778523489933,
+      "grad_norm": 0.46574166417121887,
+      "learning_rate": 4.551771118941367e-06,
+      "loss": 0.562,
+      "step": 1119
+    },
+    {
+      "epoch": 1.2527964205816555,
+      "grad_norm": 0.46505433320999146,
+      "learning_rate": 4.5509182933804754e-06,
+      "loss": 0.5829,
+      "step": 1120
+    },
+    {
+      "epoch": 1.2539149888143177,
+      "grad_norm": 0.47223830223083496,
+      "learning_rate": 4.550064737326127e-06,
+      "loss": 0.5833,
+      "step": 1121
+    },
+    {
+      "epoch": 1.25503355704698,
+      "grad_norm": 0.4903958737850189,
+      "learning_rate": 4.549210451082342e-06,
+      "loss": 0.5612,
+      "step": 1122
+    },
+    {
+      "epoch": 1.256152125279642,
+      "grad_norm": 0.4613335132598877,
+      "learning_rate": 4.548355434953397e-06,
+      "loss": 0.5826,
+      "step": 1123
+    },
+    {
+      "epoch": 1.2572706935123041,
+      "grad_norm": 0.47294437885284424,
+      "learning_rate": 4.5474996892438296e-06,
+      "loss": 0.5834,
+      "step": 1124
+    },
+    {
+      "epoch": 1.2583892617449663,
+      "grad_norm": 0.4580760598182678,
+      "learning_rate": 4.546643214258441e-06,
+      "loss": 0.6002,
+      "step": 1125
+    },
+    {
+      "epoch": 1.2595078299776286,
+      "grad_norm": 0.4971761703491211,
+      "learning_rate": 4.545786010302287e-06,
+      "loss": 0.6082,
+      "step": 1126
+    },
+    {
+      "epoch": 1.2606263982102908,
+      "grad_norm": 0.47834956645965576,
+      "learning_rate": 4.544928077680687e-06,
+      "loss": 0.5697,
+      "step": 1127
+    },
+    {
+      "epoch": 1.261744966442953,
+      "grad_norm": 0.47802045941352844,
+      "learning_rate": 4.5440694166992175e-06,
+      "loss": 0.5709,
+      "step": 1128
+    },
+    {
+      "epoch": 1.2628635346756152,
+      "grad_norm": 0.46790894865989685,
+      "learning_rate": 4.5432100276637156e-06,
+      "loss": 0.574,
+      "step": 1129
+    },
+    {
+      "epoch": 1.2639821029082774,
+      "grad_norm": 0.4805370271205902,
+      "learning_rate": 4.542349910880277e-06,
+      "loss": 0.5665,
+      "step": 1130
+    },
+    {
+      "epoch": 1.2651006711409396,
+      "grad_norm": 0.4909112751483917,
+      "learning_rate": 4.5414890666552575e-06,
+      "loss": 0.5853,
+      "step": 1131
+    },
+    {
+      "epoch": 1.2662192393736018,
+      "grad_norm": 0.4830676317214966,
+      "learning_rate": 4.540627495295271e-06,
+      "loss": 0.5977,
+      "step": 1132
+    },
+    {
+      "epoch": 1.267337807606264,
+      "grad_norm": 0.48109954595565796,
+      "learning_rate": 4.539765197107191e-06,
+      "loss": 0.6083,
+      "step": 1133
+    },
+    {
+      "epoch": 1.2684563758389262,
+      "grad_norm": 0.48772624135017395,
+      "learning_rate": 4.538902172398151e-06,
+      "loss": 0.5858,
+      "step": 1134
+    },
+    {
+      "epoch": 1.2695749440715884,
+      "grad_norm": 0.4889027774333954,
+      "learning_rate": 4.53803842147554e-06,
+      "loss": 0.6107,
+      "step": 1135
+    },
+    {
+      "epoch": 1.2706935123042506,
+      "grad_norm": 0.4655384123325348,
+      "learning_rate": 4.5371739446470085e-06,
+      "loss": 0.575,
+      "step": 1136
+    },
+    {
+      "epoch": 1.2718120805369129,
+      "grad_norm": 0.5289780497550964,
+      "learning_rate": 4.536308742220466e-06,
+      "loss": 0.5875,
+      "step": 1137
+    },
+    {
+      "epoch": 1.272930648769575,
+      "grad_norm": 0.44917261600494385,
+      "learning_rate": 4.535442814504077e-06,
+      "loss": 0.5555,
+      "step": 1138
+    },
+    {
+      "epoch": 1.274049217002237,
+      "grad_norm": 0.45869240164756775,
+      "learning_rate": 4.534576161806269e-06,
+      "loss": 0.5673,
+      "step": 1139
+    },
+    {
+      "epoch": 1.2751677852348993,
+      "grad_norm": 0.47333377599716187,
+      "learning_rate": 4.533708784435722e-06,
+      "loss": 0.5843,
+      "step": 1140
+    },
+    {
+      "epoch": 1.2762863534675615,
+      "grad_norm": 0.4835321307182312,
+      "learning_rate": 4.5328406827013806e-06,
+      "loss": 0.6049,
+      "step": 1141
+    },
+    {
+      "epoch": 1.2774049217002237,
+      "grad_norm": 0.4710775315761566,
+      "learning_rate": 4.531971856912443e-06,
+      "loss": 0.5723,
+      "step": 1142
+    },
+    {
+      "epoch": 1.278523489932886,
+      "grad_norm": 0.48879000544548035,
+      "learning_rate": 4.531102307378366e-06,
+      "loss": 0.5847,
+      "step": 1143
+    },
+    {
+      "epoch": 1.279642058165548,
+      "grad_norm": 0.476485013961792,
+      "learning_rate": 4.530232034408864e-06,
+      "loss": 0.583,
+      "step": 1144
+    },
+    {
+      "epoch": 1.2807606263982103,
+      "grad_norm": 0.48762884736061096,
+      "learning_rate": 4.529361038313912e-06,
+      "loss": 0.6017,
+      "step": 1145
+    },
+    {
+      "epoch": 1.2818791946308725,
+      "grad_norm": 0.4625467360019684,
+      "learning_rate": 4.528489319403737e-06,
+      "loss": 0.5835,
+      "step": 1146
+    },
+    {
+      "epoch": 1.2829977628635347,
+      "grad_norm": 0.47126954793930054,
+      "learning_rate": 4.52761687798883e-06,
+      "loss": 0.5503,
+      "step": 1147
+    },
+    {
+      "epoch": 1.284116331096197,
+      "grad_norm": 0.46396777033805847,
+      "learning_rate": 4.526743714379934e-06,
+      "loss": 0.5669,
+      "step": 1148
+    },
+    {
+      "epoch": 1.285234899328859,
+      "grad_norm": 0.4775722920894623,
+      "learning_rate": 4.5258698288880535e-06,
+      "loss": 0.5727,
+      "step": 1149
+    },
+    {
+      "epoch": 1.2863534675615211,
+      "grad_norm": 0.5126634240150452,
+      "learning_rate": 4.524995221824445e-06,
+      "loss": 0.6249,
+      "step": 1150
+    },
+    {
+      "epoch": 1.2874720357941833,
+      "grad_norm": 0.47814854979515076,
+      "learning_rate": 4.524119893500627e-06,
+      "loss": 0.5881,
+      "step": 1151
+    },
+    {
+      "epoch": 1.2885906040268456,
+      "grad_norm": 0.47940847277641296,
+      "learning_rate": 4.523243844228372e-06,
+      "loss": 0.562,
+      "step": 1152
+    },
+    {
+      "epoch": 1.2897091722595078,
+      "grad_norm": 0.48836034536361694,
+      "learning_rate": 4.52236707431971e-06,
+      "loss": 0.5812,
+      "step": 1153
+    },
+    {
+      "epoch": 1.29082774049217,
+      "grad_norm": 0.47201383113861084,
+      "learning_rate": 4.521489584086929e-06,
+      "loss": 0.5998,
+      "step": 1154
+    },
+    {
+      "epoch": 1.2919463087248322,
+      "grad_norm": 0.4711892604827881,
+      "learning_rate": 4.52061137384257e-06,
+      "loss": 0.5868,
+      "step": 1155
+    },
+    {
+      "epoch": 1.2930648769574944,
+      "grad_norm": 0.49769148230552673,
+      "learning_rate": 4.519732443899435e-06,
+      "loss": 0.6405,
+      "step": 1156
+    },
+    {
+      "epoch": 1.2941834451901566,
+      "grad_norm": 0.4788062870502472,
+      "learning_rate": 4.51885279457058e-06,
+      "loss": 0.5838,
+      "step": 1157
+    },
+    {
+      "epoch": 1.2953020134228188,
+      "grad_norm": 0.47234484553337097,
+      "learning_rate": 4.5179724261693154e-06,
+      "loss": 0.5794,
+      "step": 1158
+    },
+    {
+      "epoch": 1.296420581655481,
+      "grad_norm": 0.464893639087677,
+      "learning_rate": 4.517091339009212e-06,
+      "loss": 0.5741,
+      "step": 1159
+    },
+    {
+      "epoch": 1.2975391498881432,
+      "grad_norm": 0.47124040126800537,
+      "learning_rate": 4.516209533404092e-06,
+      "loss": 0.5974,
+      "step": 1160
+    },
+    {
+      "epoch": 1.2986577181208054,
+      "grad_norm": 0.470726877450943,
+      "learning_rate": 4.5153270096680395e-06,
+      "loss": 0.5686,
+      "step": 1161
+    },
+    {
+      "epoch": 1.2997762863534676,
+      "grad_norm": 0.4629981815814972,
+      "learning_rate": 4.514443768115386e-06,
+      "loss": 0.5794,
+      "step": 1162
+    },
+    {
+      "epoch": 1.3008948545861299,
+      "grad_norm": 0.4796271026134491,
+      "learning_rate": 4.513559809060727e-06,
+      "loss": 0.5741,
+      "step": 1163
+    },
+    {
+      "epoch": 1.302013422818792,
+      "grad_norm": 0.4912780523300171,
+      "learning_rate": 4.512675132818908e-06,
+      "loss": 0.5726,
+      "step": 1164
+    },
+    {
+      "epoch": 1.3031319910514543,
+      "grad_norm": 0.4748340845108032,
+      "learning_rate": 4.511789739705033e-06,
+      "loss": 0.5788,
+      "step": 1165
+    },
+    {
+      "epoch": 1.3042505592841163,
+      "grad_norm": 0.47445476055145264,
+      "learning_rate": 4.510903630034458e-06,
+      "loss": 0.5924,
+      "step": 1166
+    },
+    {
+      "epoch": 1.3053691275167785,
+      "grad_norm": 0.46978822350502014,
+      "learning_rate": 4.510016804122799e-06,
+      "loss": 0.5525,
+      "step": 1167
+    },
+    {
+      "epoch": 1.3064876957494407,
+      "grad_norm": 0.48169562220573425,
+      "learning_rate": 4.509129262285924e-06,
+      "loss": 0.5944,
+      "step": 1168
+    },
+    {
+      "epoch": 1.307606263982103,
+      "grad_norm": 0.4853328764438629,
+      "learning_rate": 4.5082410048399555e-06,
+      "loss": 0.5774,
+      "step": 1169
+    },
+    {
+      "epoch": 1.308724832214765,
+      "grad_norm": 0.4685922861099243,
+      "learning_rate": 4.507352032101273e-06,
+      "loss": 0.583,
+      "step": 1170
+    },
+    {
+      "epoch": 1.3098434004474273,
+      "grad_norm": 0.466190904378891,
+      "learning_rate": 4.5064623443865085e-06,
+      "loss": 0.5676,
+      "step": 1171
+    },
+    {
+      "epoch": 1.3109619686800895,
+      "grad_norm": 0.4621398448944092,
+      "learning_rate": 4.50557194201255e-06,
+      "loss": 0.5674,
+      "step": 1172
+    },
+    {
+      "epoch": 1.3120805369127517,
+      "grad_norm": 0.48330026865005493,
+      "learning_rate": 4.504680825296542e-06,
+      "loss": 0.588,
+      "step": 1173
+    },
+    {
+      "epoch": 1.313199105145414,
+      "grad_norm": 0.48152461647987366,
+      "learning_rate": 4.503788994555878e-06,
+      "loss": 0.5975,
+      "step": 1174
+    },
+    {
+      "epoch": 1.3143176733780761,
+      "grad_norm": 0.47018247842788696,
+      "learning_rate": 4.502896450108211e-06,
+      "loss": 0.5707,
+      "step": 1175
+    },
+    {
+      "epoch": 1.3154362416107381,
+      "grad_norm": 0.48568496108055115,
+      "learning_rate": 4.502003192271447e-06,
+      "loss": 0.5923,
+      "step": 1176
+    },
+    {
+      "epoch": 1.3165548098434003,
+      "grad_norm": 0.5025852918624878,
+      "learning_rate": 4.501109221363744e-06,
+      "loss": 0.6094,
+      "step": 1177
+    },
+    {
+      "epoch": 1.3176733780760626,
+      "grad_norm": 0.46192529797554016,
+      "learning_rate": 4.500214537703515e-06,
+      "loss": 0.5815,
+      "step": 1178
+    },
+    {
+      "epoch": 1.3187919463087248,
+      "grad_norm": 0.4682995080947876,
+      "learning_rate": 4.499319141609429e-06,
+      "loss": 0.5526,
+      "step": 1179
+    },
+    {
+      "epoch": 1.319910514541387,
+      "grad_norm": 0.4849640130996704,
+      "learning_rate": 4.498423033400408e-06,
+      "loss": 0.6011,
+      "step": 1180
+    },
+    {
+      "epoch": 1.3210290827740492,
+      "grad_norm": 0.4680434465408325,
+      "learning_rate": 4.4975262133956235e-06,
+      "loss": 0.5763,
+      "step": 1181
+    },
+    {
+      "epoch": 1.3221476510067114,
+      "grad_norm": 0.47362959384918213,
+      "learning_rate": 4.496628681914505e-06,
+      "loss": 0.5793,
+      "step": 1182
+    },
+    {
+      "epoch": 1.3232662192393736,
+      "grad_norm": 0.48274803161621094,
+      "learning_rate": 4.495730439276734e-06,
+      "loss": 0.5806,
+      "step": 1183
+    },
+    {
+      "epoch": 1.3243847874720358,
+      "grad_norm": 0.5004247426986694,
+      "learning_rate": 4.4948314858022456e-06,
+      "loss": 0.5982,
+      "step": 1184
+    },
+    {
+      "epoch": 1.325503355704698,
+      "grad_norm": 0.4769462049007416,
+      "learning_rate": 4.4939318218112284e-06,
+      "loss": 0.593,
+      "step": 1185
+    },
+    {
+      "epoch": 1.3266219239373602,
+      "grad_norm": 0.4818820357322693,
+      "learning_rate": 4.493031447624125e-06,
+      "loss": 0.5999,
+      "step": 1186
+    },
+    {
+      "epoch": 1.3277404921700224,
+      "grad_norm": 0.4788765609264374,
+      "learning_rate": 4.492130363561625e-06,
+      "loss": 0.5978,
+      "step": 1187
+    },
+    {
+      "epoch": 1.3288590604026846,
+      "grad_norm": 0.4671361446380615,
+      "learning_rate": 4.491228569944679e-06,
+      "loss": 0.5715,
+      "step": 1188
+    },
+    {
+      "epoch": 1.3299776286353469,
+      "grad_norm": 0.4770296514034271,
+      "learning_rate": 4.4903260670944855e-06,
+      "loss": 0.5913,
+      "step": 1189
+    },
+    {
+      "epoch": 1.331096196868009,
+      "grad_norm": 0.47733429074287415,
+      "learning_rate": 4.489422855332497e-06,
+      "loss": 0.5914,
+      "step": 1190
+    },
+    {
+      "epoch": 1.3322147651006713,
+      "grad_norm": 0.49386340379714966,
+      "learning_rate": 4.488518934980419e-06,
+      "loss": 0.5866,
+      "step": 1191
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.4807666838169098,
+      "learning_rate": 4.487614306360208e-06,
+      "loss": 0.5769,
+      "step": 1192
+    },
+    {
+      "epoch": 1.3344519015659955,
+      "grad_norm": 0.46346330642700195,
+      "learning_rate": 4.4867089697940735e-06,
+      "loss": 0.5889,
+      "step": 1193
+    },
+    {
+      "epoch": 1.3355704697986577,
+      "grad_norm": 0.46165117621421814,
+      "learning_rate": 4.485802925604476e-06,
+      "loss": 0.5638,
+      "step": 1194
+    },
+    {
+      "epoch": 1.3366890380313199,
+      "grad_norm": 0.4789906442165375,
+      "learning_rate": 4.484896174114132e-06,
+      "loss": 0.5984,
+      "step": 1195
+    },
+    {
+      "epoch": 1.337807606263982,
+      "grad_norm": 0.47423994541168213,
+      "learning_rate": 4.4839887156460045e-06,
+      "loss": 0.6043,
+      "step": 1196
+    },
+    {
+      "epoch": 1.3389261744966443,
+      "grad_norm": 0.45627936720848083,
+      "learning_rate": 4.4830805505233125e-06,
+      "loss": 0.5781,
+      "step": 1197
+    },
+    {
+      "epoch": 1.3400447427293065,
+      "grad_norm": 0.49102070927619934,
+      "learning_rate": 4.482171679069524e-06,
+      "loss": 0.5854,
+      "step": 1198
+    },
+    {
+      "epoch": 1.3411633109619687,
+      "grad_norm": 0.48892414569854736,
+      "learning_rate": 4.48126210160836e-06,
+      "loss": 0.6194,
+      "step": 1199
+    },
+    {
+      "epoch": 1.342281879194631,
+      "grad_norm": 0.481579065322876,
+      "learning_rate": 4.480351818463793e-06,
+      "loss": 0.5901,
+      "step": 1200
+    },
+    {
+      "epoch": 1.3434004474272931,
+      "grad_norm": 0.4813998341560364,
+      "learning_rate": 4.479440829960045e-06,
+      "loss": 0.6095,
+      "step": 1201
+    },
+    {
+      "epoch": 1.3445190156599551,
+      "grad_norm": 0.466458261013031,
+      "learning_rate": 4.478529136421593e-06,
+      "loss": 0.5481,
+      "step": 1202
+    },
+    {
+      "epoch": 1.3456375838926173,
+      "grad_norm": 0.45974788069725037,
+      "learning_rate": 4.477616738173162e-06,
+      "loss": 0.5832,
+      "step": 1203
+    },
+    {
+      "epoch": 1.3467561521252795,
+      "grad_norm": 0.4749172329902649,
+      "learning_rate": 4.476703635539728e-06,
+      "loss": 0.6021,
+      "step": 1204
+    },
+    {
+      "epoch": 1.3478747203579418,
+      "grad_norm": 0.47382864356040955,
+      "learning_rate": 4.475789828846519e-06,
+      "loss": 0.5936,
+      "step": 1205
+    },
+    {
+      "epoch": 1.348993288590604,
+      "grad_norm": 0.4810637831687927,
+      "learning_rate": 4.474875318419015e-06,
+      "loss": 0.5914,
+      "step": 1206
+    },
+    {
+      "epoch": 1.3501118568232662,
+      "grad_norm": 0.46880605816841125,
+      "learning_rate": 4.473960104582943e-06,
+      "loss": 0.569,
+      "step": 1207
+    },
+    {
+      "epoch": 1.3512304250559284,
+      "grad_norm": 0.4803348183631897,
+      "learning_rate": 4.473044187664284e-06,
+      "loss": 0.5942,
+      "step": 1208
+    },
+    {
+      "epoch": 1.3523489932885906,
+      "grad_norm": 0.4822969436645508,
+      "learning_rate": 4.472127567989268e-06,
+      "loss": 0.5893,
+      "step": 1209
+    },
+    {
+      "epoch": 1.3534675615212528,
+      "grad_norm": 0.5106329321861267,
+      "learning_rate": 4.4712102458843755e-06,
+      "loss": 0.5954,
+      "step": 1210
+    },
+    {
+      "epoch": 1.354586129753915,
+      "grad_norm": 0.4909040629863739,
+      "learning_rate": 4.470292221676336e-06,
+      "loss": 0.5968,
+      "step": 1211
+    },
+    {
+      "epoch": 1.3557046979865772,
+      "grad_norm": 0.48394322395324707,
+      "learning_rate": 4.469373495692132e-06,
+      "loss": 0.5516,
+      "step": 1212
+    },
+    {
+      "epoch": 1.3568232662192394,
+      "grad_norm": 0.46607136726379395,
+      "learning_rate": 4.4684540682589925e-06,
+      "loss": 0.5588,
+      "step": 1213
+    },
+    {
+      "epoch": 1.3579418344519016,
+      "grad_norm": 0.48817023634910583,
+      "learning_rate": 4.467533939704398e-06,
+      "loss": 0.5877,
+      "step": 1214
+    },
+    {
+      "epoch": 1.3590604026845639,
+      "grad_norm": 0.4760921895503998,
+      "learning_rate": 4.466613110356081e-06,
+      "loss": 0.5482,
+      "step": 1215
+    },
+    {
+      "epoch": 1.360178970917226,
+      "grad_norm": 0.49914559721946716,
+      "learning_rate": 4.465691580542019e-06,
+      "loss": 0.6012,
+      "step": 1216
+    },
+    {
+      "epoch": 1.3612975391498883,
+      "grad_norm": 0.4916679561138153,
+      "learning_rate": 4.464769350590441e-06,
+      "loss": 0.6107,
+      "step": 1217
+    },
+    {
+      "epoch": 1.3624161073825503,
+      "grad_norm": 0.47448763251304626,
+      "learning_rate": 4.463846420829828e-06,
+      "loss": 0.5555,
+      "step": 1218
+    },
+    {
+      "epoch": 1.3635346756152125,
+      "grad_norm": 0.48127514123916626,
+      "learning_rate": 4.462922791588906e-06,
+      "loss": 0.5826,
+      "step": 1219
+    },
+    {
+      "epoch": 1.3646532438478747,
+      "grad_norm": 0.4912678003311157,
+      "learning_rate": 4.461998463196653e-06,
+      "loss": 0.5727,
+      "step": 1220
+    },
+    {
+      "epoch": 1.3657718120805369,
+      "grad_norm": 0.4831150472164154,
+      "learning_rate": 4.461073435982295e-06,
+      "loss": 0.6161,
+      "step": 1221
+    },
+    {
+      "epoch": 1.366890380313199,
+      "grad_norm": 0.4749965965747833,
+      "learning_rate": 4.460147710275306e-06,
+      "loss": 0.6009,
+      "step": 1222
+    },
+    {
+      "epoch": 1.3680089485458613,
+      "grad_norm": 0.5110952854156494,
+      "learning_rate": 4.459221286405411e-06,
+      "loss": 0.6029,
+      "step": 1223
+    },
+    {
+      "epoch": 1.3691275167785235,
+      "grad_norm": 0.4717468321323395,
+      "learning_rate": 4.458294164702582e-06,
+      "loss": 0.5699,
+      "step": 1224
+    },
+    {
+      "epoch": 1.3702460850111857,
+      "grad_norm": 0.4701862633228302,
+      "learning_rate": 4.457366345497041e-06,
+      "loss": 0.5575,
+      "step": 1225
+    },
+    {
+      "epoch": 1.371364653243848,
+      "grad_norm": 0.47513946890830994,
+      "learning_rate": 4.456437829119256e-06,
+      "loss": 0.565,
+      "step": 1226
+    },
+    {
+      "epoch": 1.3724832214765101,
+      "grad_norm": 0.4824097752571106,
+      "learning_rate": 4.455508615899945e-06,
+      "loss": 0.5883,
+      "step": 1227
+    },
+    {
+      "epoch": 1.3736017897091721,
+      "grad_norm": 0.4734891355037689,
+      "learning_rate": 4.454578706170075e-06,
+      "loss": 0.5682,
+      "step": 1228
+    },
+    {
+      "epoch": 1.3747203579418343,
+      "grad_norm": 0.4862266182899475,
+      "learning_rate": 4.453648100260859e-06,
+      "loss": 0.5886,
+      "step": 1229
+    },
+    {
+      "epoch": 1.3758389261744965,
+      "grad_norm": 0.4855569005012512,
+      "learning_rate": 4.452716798503759e-06,
+      "loss": 0.5971,
+      "step": 1230
+    },
+    {
+      "epoch": 1.3769574944071588,
+      "grad_norm": 0.4638337194919586,
+      "learning_rate": 4.451784801230487e-06,
+      "loss": 0.578,
+      "step": 1231
+    },
+    {
+      "epoch": 1.378076062639821,
+      "grad_norm": 0.47969067096710205,
+      "learning_rate": 4.4508521087729975e-06,
+      "loss": 0.6131,
+      "step": 1232
+    },
+    {
+      "epoch": 1.3791946308724832,
+      "grad_norm": 0.45764845609664917,
+      "learning_rate": 4.449918721463497e-06,
+      "loss": 0.5439,
+      "step": 1233
+    },
+    {
+      "epoch": 1.3803131991051454,
+      "grad_norm": 0.4725312292575836,
+      "learning_rate": 4.448984639634439e-06,
+      "loss": 0.5774,
+      "step": 1234
+    },
+    {
+      "epoch": 1.3814317673378076,
+      "grad_norm": 0.4911689758300781,
+      "learning_rate": 4.448049863618522e-06,
+      "loss": 0.5891,
+      "step": 1235
+    },
+    {
+      "epoch": 1.3825503355704698,
+      "grad_norm": 0.47672075033187866,
+      "learning_rate": 4.447114393748694e-06,
+      "loss": 0.5706,
+      "step": 1236
+    },
+    {
+      "epoch": 1.383668903803132,
+      "grad_norm": 0.4902832806110382,
+      "learning_rate": 4.446178230358151e-06,
+      "loss": 0.6043,
+      "step": 1237
+    },
+    {
+      "epoch": 1.3847874720357942,
+      "grad_norm": 0.496720552444458,
+      "learning_rate": 4.445241373780333e-06,
+      "loss": 0.5831,
+      "step": 1238
+    },
+    {
+      "epoch": 1.3859060402684564,
+      "grad_norm": 0.4805045425891876,
+      "learning_rate": 4.444303824348927e-06,
+      "loss": 0.5769,
+      "step": 1239
+    },
+    {
+      "epoch": 1.3870246085011186,
+      "grad_norm": 0.48103493452072144,
+      "learning_rate": 4.44336558239787e-06,
+      "loss": 0.5778,
+      "step": 1240
+    },
+    {
+      "epoch": 1.3881431767337808,
+      "grad_norm": 0.4719166159629822,
+      "learning_rate": 4.4424266482613445e-06,
+      "loss": 0.5666,
+      "step": 1241
+    },
+    {
+      "epoch": 1.389261744966443,
+      "grad_norm": 0.4719798266887665,
+      "learning_rate": 4.4414870222737775e-06,
+      "loss": 0.5579,
+      "step": 1242
+    },
+    {
+      "epoch": 1.3903803131991053,
+      "grad_norm": 0.49560073018074036,
+      "learning_rate": 4.440546704769845e-06,
+      "loss": 0.5858,
+      "step": 1243
+    },
+    {
+      "epoch": 1.3914988814317675,
+      "grad_norm": 0.4905194044113159,
+      "learning_rate": 4.439605696084465e-06,
+      "loss": 0.5747,
+      "step": 1244
+    },
+    {
+      "epoch": 1.3926174496644295,
+      "grad_norm": 0.502269446849823,
+      "learning_rate": 4.438663996552809e-06,
+      "loss": 0.5965,
+      "step": 1245
+    },
+    {
+      "epoch": 1.3937360178970917,
+      "grad_norm": 0.47396889328956604,
+      "learning_rate": 4.437721606510288e-06,
+      "loss": 0.6069,
+      "step": 1246
+    },
+    {
+      "epoch": 1.3948545861297539,
+      "grad_norm": 0.4756862223148346,
+      "learning_rate": 4.436778526292561e-06,
+      "loss": 0.5563,
+      "step": 1247
+    },
+    {
+      "epoch": 1.395973154362416,
+      "grad_norm": 0.47935912013053894,
+      "learning_rate": 4.435834756235534e-06,
+      "loss": 0.5811,
+      "step": 1248
+    },
+    {
+      "epoch": 1.3970917225950783,
+      "grad_norm": 0.47267550230026245,
+      "learning_rate": 4.434890296675358e-06,
+      "loss": 0.5477,
+      "step": 1249
+    },
+    {
+      "epoch": 1.3982102908277405,
+      "grad_norm": 0.4694240391254425,
+      "learning_rate": 4.433945147948428e-06,
+      "loss": 0.5586,
+      "step": 1250
+    },
+    {
+      "epoch": 1.3993288590604027,
+      "grad_norm": 0.47105926275253296,
+      "learning_rate": 4.432999310391387e-06,
+      "loss": 0.5684,
+      "step": 1251
+    },
+    {
+      "epoch": 1.400447427293065,
+      "grad_norm": 0.4700409173965454,
+      "learning_rate": 4.432052784341122e-06,
+      "loss": 0.5748,
+      "step": 1252
+    },
+    {
+      "epoch": 1.4015659955257271,
+      "grad_norm": 0.47937822341918945,
+      "learning_rate": 4.431105570134766e-06,
+      "loss": 0.5638,
+      "step": 1253
+    },
+    {
+      "epoch": 1.4026845637583891,
+      "grad_norm": 0.47045087814331055,
+      "learning_rate": 4.4301576681096956e-06,
+      "loss": 0.5597,
+      "step": 1254
+    },
+    {
+      "epoch": 1.4038031319910513,
+      "grad_norm": 0.48519641160964966,
+      "learning_rate": 4.429209078603534e-06,
+      "loss": 0.555,
+      "step": 1255
+    },
+    {
+      "epoch": 1.4049217002237135,
+      "grad_norm": 0.4835051894187927,
+      "learning_rate": 4.428259801954148e-06,
+      "loss": 0.6021,
+      "step": 1256
+    },
+    {
+      "epoch": 1.4060402684563758,
+      "grad_norm": 0.49831488728523254,
+      "learning_rate": 4.427309838499651e-06,
+      "loss": 0.5739,
+      "step": 1257
+    },
+    {
+      "epoch": 1.407158836689038,
+      "grad_norm": 0.5223401784896851,
+      "learning_rate": 4.4263591885783976e-06,
+      "loss": 0.5881,
+      "step": 1258
+    },
+    {
+      "epoch": 1.4082774049217002,
+      "grad_norm": 0.47920864820480347,
+      "learning_rate": 4.425407852528991e-06,
+      "loss": 0.5981,
+      "step": 1259
+    },
+    {
+      "epoch": 1.4093959731543624,
+      "grad_norm": 0.5153250694274902,
+      "learning_rate": 4.424455830690275e-06,
+      "loss": 0.5715,
+      "step": 1260
+    },
+    {
+      "epoch": 1.4105145413870246,
+      "grad_norm": 0.48909991979599,
+      "learning_rate": 4.4235031234013425e-06,
+      "loss": 0.602,
+      "step": 1261
+    },
+    {
+      "epoch": 1.4116331096196868,
+      "grad_norm": 0.4935949444770813,
+      "learning_rate": 4.422549731001524e-06,
+      "loss": 0.6008,
+      "step": 1262
+    },
+    {
+      "epoch": 1.412751677852349,
+      "grad_norm": 0.49626025557518005,
+      "learning_rate": 4.421595653830401e-06,
+      "loss": 0.6158,
+      "step": 1263
+    },
+    {
+      "epoch": 1.4138702460850112,
+      "grad_norm": 0.4757371246814728,
+      "learning_rate": 4.420640892227793e-06,
+      "loss": 0.573,
+      "step": 1264
+    },
+    {
+      "epoch": 1.4149888143176734,
+      "grad_norm": 0.4638062119483948,
+      "learning_rate": 4.4196854465337664e-06,
+      "loss": 0.5808,
+      "step": 1265
+    },
+    {
+      "epoch": 1.4161073825503356,
+      "grad_norm": 0.4798492193222046,
+      "learning_rate": 4.418729317088631e-06,
+      "loss": 0.5706,
+      "step": 1266
+    },
+    {
+      "epoch": 1.4172259507829978,
+      "grad_norm": 0.4708186388015747,
+      "learning_rate": 4.41777250423294e-06,
+      "loss": 0.576,
+      "step": 1267
+    },
+    {
+      "epoch": 1.41834451901566,
+      "grad_norm": 0.49962857365608215,
+      "learning_rate": 4.416815008307488e-06,
+      "loss": 0.596,
+      "step": 1268
+    },
+    {
+      "epoch": 1.4194630872483223,
+      "grad_norm": 0.4764310121536255,
+      "learning_rate": 4.415856829653318e-06,
+      "loss": 0.5903,
+      "step": 1269
+    },
+    {
+      "epoch": 1.4205816554809845,
+      "grad_norm": 0.4788293242454529,
+      "learning_rate": 4.4148979686117095e-06,
+      "loss": 0.5803,
+      "step": 1270
+    },
+    {
+      "epoch": 1.4217002237136465,
+      "grad_norm": 0.45909953117370605,
+      "learning_rate": 4.41393842552419e-06,
+      "loss": 0.5782,
+      "step": 1271
+    },
+    {
+      "epoch": 1.4228187919463087,
+      "grad_norm": 0.4760737717151642,
+      "learning_rate": 4.412978200732528e-06,
+      "loss": 0.6256,
+      "step": 1272
+    },
+    {
+      "epoch": 1.4239373601789709,
+      "grad_norm": 0.485414981842041,
+      "learning_rate": 4.412017294578737e-06,
+      "loss": 0.5895,
+      "step": 1273
+    },
+    {
+      "epoch": 1.425055928411633,
+      "grad_norm": 0.4804149568080902,
+      "learning_rate": 4.411055707405068e-06,
+      "loss": 0.5755,
+      "step": 1274
+    },
+    {
+      "epoch": 1.4261744966442953,
+      "grad_norm": 0.4825088679790497,
+      "learning_rate": 4.410093439554019e-06,
+      "loss": 0.5884,
+      "step": 1275
+    },
+    {
+      "epoch": 1.4272930648769575,
+      "grad_norm": 0.49632760882377625,
+      "learning_rate": 4.409130491368331e-06,
+      "loss": 0.6117,
+      "step": 1276
+    },
+    {
+      "epoch": 1.4284116331096197,
+      "grad_norm": 0.4739936888217926,
+      "learning_rate": 4.408166863190983e-06,
+      "loss": 0.5696,
+      "step": 1277
+    },
+    {
+      "epoch": 1.429530201342282,
+      "grad_norm": 0.5045065879821777,
+      "learning_rate": 4.407202555365202e-06,
+      "loss": 0.6204,
+      "step": 1278
+    },
+    {
+      "epoch": 1.4306487695749441,
+      "grad_norm": 0.4909072816371918,
+      "learning_rate": 4.406237568234451e-06,
+      "loss": 0.5891,
+      "step": 1279
+    },
+    {
+      "epoch": 1.4317673378076063,
+      "grad_norm": 0.4925367534160614,
+      "learning_rate": 4.4052719021424395e-06,
+      "loss": 0.5963,
+      "step": 1280
+    },
+    {
+      "epoch": 1.4328859060402683,
+      "grad_norm": 0.4968665838241577,
+      "learning_rate": 4.404305557433118e-06,
+      "loss": 0.5971,
+      "step": 1281
+    },
+    {
+      "epoch": 1.4340044742729305,
+      "grad_norm": 0.49153050780296326,
+      "learning_rate": 4.403338534450675e-06,
+      "loss": 0.6104,
+      "step": 1282
+    },
+    {
+      "epoch": 1.4351230425055927,
+      "grad_norm": 0.47348082065582275,
+      "learning_rate": 4.4023708335395455e-06,
+      "loss": 0.5746,
+      "step": 1283
+    },
+    {
+      "epoch": 1.436241610738255,
+      "grad_norm": 0.47999560832977295,
+      "learning_rate": 4.401402455044405e-06,
+      "loss": 0.6035,
+      "step": 1284
+    },
+    {
+      "epoch": 1.4373601789709172,
+      "grad_norm": 0.4812726080417633,
+      "learning_rate": 4.4004333993101665e-06,
+      "loss": 0.5946,
+      "step": 1285
+    },
+    {
+      "epoch": 1.4384787472035794,
+      "grad_norm": 0.4776507318019867,
+      "learning_rate": 4.39946366668199e-06,
+      "loss": 0.6048,
+      "step": 1286
+    },
+    {
+      "epoch": 1.4395973154362416,
+      "grad_norm": 0.4812582731246948,
+      "learning_rate": 4.398493257505271e-06,
+      "loss": 0.5949,
+      "step": 1287
+    },
+    {
+      "epoch": 1.4407158836689038,
+      "grad_norm": 0.4620709717273712,
+      "learning_rate": 4.397522172125651e-06,
+      "loss": 0.5566,
+      "step": 1288
+    },
+    {
+      "epoch": 1.441834451901566,
+      "grad_norm": 0.47055062651634216,
+      "learning_rate": 4.3965504108890075e-06,
+      "loss": 0.5507,
+      "step": 1289
+    },
+    {
+      "epoch": 1.4429530201342282,
+      "grad_norm": 0.4810755252838135,
+      "learning_rate": 4.395577974141464e-06,
+      "loss": 0.5789,
+      "step": 1290
+    },
+    {
+      "epoch": 1.4440715883668904,
+      "grad_norm": 0.4749661982059479,
+      "learning_rate": 4.394604862229379e-06,
+      "loss": 0.5686,
+      "step": 1291
+    },
+    {
+      "epoch": 1.4451901565995526,
+      "grad_norm": 0.4877249598503113,
+      "learning_rate": 4.393631075499356e-06,
+      "loss": 0.5726,
+      "step": 1292
+    },
+    {
+      "epoch": 1.4463087248322148,
+      "grad_norm": 0.4861728250980377,
+      "learning_rate": 4.392656614298236e-06,
+      "loss": 0.6094,
+      "step": 1293
+    },
+    {
+      "epoch": 1.447427293064877,
+      "grad_norm": 0.4913595914840698,
+      "learning_rate": 4.391681478973103e-06,
+      "loss": 0.5971,
+      "step": 1294
+    },
+    {
+      "epoch": 1.4485458612975393,
+      "grad_norm": 0.4858112037181854,
+      "learning_rate": 4.390705669871278e-06,
+      "loss": 0.5675,
+      "step": 1295
+    },
+    {
+      "epoch": 1.4496644295302015,
+      "grad_norm": 0.5002705454826355,
+      "learning_rate": 4.389729187340323e-06,
+      "loss": 0.5944,
+      "step": 1296
+    },
+    {
+      "epoch": 1.4507829977628635,
+      "grad_norm": 0.4777258038520813,
+      "learning_rate": 4.388752031728042e-06,
+      "loss": 0.6017,
+      "step": 1297
+    },
+    {
+      "epoch": 1.4519015659955257,
+      "grad_norm": 0.4859464466571808,
+      "learning_rate": 4.387774203382476e-06,
+      "loss": 0.5914,
+      "step": 1298
+    },
+    {
+      "epoch": 1.4530201342281879,
+      "grad_norm": 0.4924688935279846,
+      "learning_rate": 4.386795702651907e-06,
+      "loss": 0.5778,
+      "step": 1299
+    },
+    {
+      "epoch": 1.45413870246085,
+      "grad_norm": 0.4761926531791687,
+      "learning_rate": 4.385816529884856e-06,
+      "loss": 0.5977,
+      "step": 1300
+    },
+    {
+      "epoch": 1.4552572706935123,
+      "grad_norm": 0.4806959331035614,
+      "learning_rate": 4.3848366854300834e-06,
+      "loss": 0.5986,
+      "step": 1301
+    },
+    {
+      "epoch": 1.4563758389261745,
+      "grad_norm": 0.4975976049900055,
+      "learning_rate": 4.383856169636589e-06,
+      "loss": 0.5788,
+      "step": 1302
+    },
+    {
+      "epoch": 1.4574944071588367,
+      "grad_norm": 0.4763396382331848,
+      "learning_rate": 4.382874982853611e-06,
+      "loss": 0.5679,
+      "step": 1303
+    },
+    {
+      "epoch": 1.458612975391499,
+      "grad_norm": 0.47432610392570496,
+      "learning_rate": 4.381893125430629e-06,
+      "loss": 0.566,
+      "step": 1304
+    },
+    {
+      "epoch": 1.4597315436241611,
+      "grad_norm": 0.4684997498989105,
+      "learning_rate": 4.380910597717357e-06,
+      "loss": 0.578,
+      "step": 1305
+    },
+    {
+      "epoch": 1.4608501118568233,
+      "grad_norm": 0.4823257327079773,
+      "learning_rate": 4.379927400063754e-06,
+      "loss": 0.5848,
+      "step": 1306
+    },
+    {
+      "epoch": 1.4619686800894853,
+      "grad_norm": 0.47135570645332336,
+      "learning_rate": 4.378943532820011e-06,
+      "loss": 0.5781,
+      "step": 1307
+    },
+    {
+      "epoch": 1.4630872483221475,
+      "grad_norm": 0.4889175593852997,
+      "learning_rate": 4.377958996336563e-06,
+      "loss": 0.5307,
+      "step": 1308
+    },
+    {
+      "epoch": 1.4642058165548097,
+      "grad_norm": 0.49853044748306274,
+      "learning_rate": 4.376973790964078e-06,
+      "loss": 0.5791,
+      "step": 1309
+    },
+    {
+      "epoch": 1.465324384787472,
+      "grad_norm": 0.47457244992256165,
+      "learning_rate": 4.375987917053468e-06,
+      "loss": 0.5552,
+      "step": 1310
+    },
+    {
+      "epoch": 1.4664429530201342,
+      "grad_norm": 0.4792843163013458,
+      "learning_rate": 4.37500137495588e-06,
+      "loss": 0.5753,
+      "step": 1311
+    },
+    {
+      "epoch": 1.4675615212527964,
+      "grad_norm": 0.4924052059650421,
+      "learning_rate": 4.3740141650226975e-06,
+      "loss": 0.6032,
+      "step": 1312
+    },
+    {
+      "epoch": 1.4686800894854586,
+      "grad_norm": 0.485517680644989,
+      "learning_rate": 4.373026287605545e-06,
+      "loss": 0.5956,
+      "step": 1313
+    },
+    {
+      "epoch": 1.4697986577181208,
+      "grad_norm": 0.5009534358978271,
+      "learning_rate": 4.372037743056283e-06,
+      "loss": 0.587,
+      "step": 1314
+    },
+    {
+      "epoch": 1.470917225950783,
+      "grad_norm": 0.5067222118377686,
+      "learning_rate": 4.371048531727009e-06,
+      "loss": 0.5663,
+      "step": 1315
+    },
+    {
+      "epoch": 1.4720357941834452,
+      "grad_norm": 0.48260119557380676,
+      "learning_rate": 4.370058653970062e-06,
+      "loss": 0.5736,
+      "step": 1316
+    },
+    {
+      "epoch": 1.4731543624161074,
+      "grad_norm": 0.4996930658817291,
+      "learning_rate": 4.369068110138013e-06,
+      "loss": 0.5796,
+      "step": 1317
+    },
+    {
+      "epoch": 1.4742729306487696,
+      "grad_norm": 0.501705527305603,
+      "learning_rate": 4.368076900583673e-06,
+      "loss": 0.6125,
+      "step": 1318
+    },
+    {
+      "epoch": 1.4753914988814318,
+      "grad_norm": 0.4803929328918457,
+      "learning_rate": 4.36708502566009e-06,
+      "loss": 0.5737,
+      "step": 1319
+    },
+    {
+      "epoch": 1.476510067114094,
+      "grad_norm": 0.4732499122619629,
+      "learning_rate": 4.366092485720549e-06,
+      "loss": 0.597,
+      "step": 1320
+    },
+    {
+      "epoch": 1.4776286353467563,
+      "grad_norm": 0.4887276589870453,
+      "learning_rate": 4.365099281118571e-06,
+      "loss": 0.5893,
+      "step": 1321
+    },
+    {
+      "epoch": 1.4787472035794185,
+      "grad_norm": 0.48169007897377014,
+      "learning_rate": 4.364105412207914e-06,
+      "loss": 0.6035,
+      "step": 1322
+    },
+    {
+      "epoch": 1.4798657718120805,
+      "grad_norm": 0.4954248070716858,
+      "learning_rate": 4.363110879342575e-06,
+      "loss": 0.6149,
+      "step": 1323
+    },
+    {
+      "epoch": 1.4809843400447427,
+      "grad_norm": 0.47244513034820557,
+      "learning_rate": 4.362115682876783e-06,
+      "loss": 0.5794,
+      "step": 1324
+    },
+    {
+      "epoch": 1.4821029082774049,
+      "grad_norm": 0.48339948058128357,
+      "learning_rate": 4.361119823165007e-06,
+      "loss": 0.5821,
+      "step": 1325
+    },
+    {
+      "epoch": 1.483221476510067,
+      "grad_norm": 0.49129050970077515,
+      "learning_rate": 4.3601233005619515e-06,
+      "loss": 0.5851,
+      "step": 1326
+    },
+    {
+      "epoch": 1.4843400447427293,
+      "grad_norm": 0.4731924533843994,
+      "learning_rate": 4.359126115422555e-06,
+      "loss": 0.55,
+      "step": 1327
+    },
+    {
+      "epoch": 1.4854586129753915,
+      "grad_norm": 0.4816952347755432,
+      "learning_rate": 4.358128268101996e-06,
+      "loss": 0.5781,
+      "step": 1328
+    },
+    {
+      "epoch": 1.4865771812080537,
+      "grad_norm": 0.4917982518672943,
+      "learning_rate": 4.357129758955685e-06,
+      "loss": 0.6017,
+      "step": 1329
+    },
+    {
+      "epoch": 1.487695749440716,
+      "grad_norm": 0.5033697485923767,
+      "learning_rate": 4.356130588339269e-06,
+      "loss": 0.5783,
+      "step": 1330
+    },
+    {
+      "epoch": 1.4888143176733781,
+      "grad_norm": 0.48702743649482727,
+      "learning_rate": 4.355130756608632e-06,
+      "loss": 0.5757,
+      "step": 1331
+    },
+    {
+      "epoch": 1.4899328859060403,
+      "grad_norm": 0.49924343824386597,
+      "learning_rate": 4.354130264119894e-06,
+      "loss": 0.5985,
+      "step": 1332
+    },
+    {
+      "epoch": 1.4910514541387023,
+      "grad_norm": 0.4950959384441376,
+      "learning_rate": 4.353129111229408e-06,
+      "loss": 0.5688,
+      "step": 1333
+    },
+    {
+      "epoch": 1.4921700223713645,
+      "grad_norm": 0.47743096947669983,
+      "learning_rate": 4.352127298293764e-06,
+      "loss": 0.5655,
+      "step": 1334
+    },
+    {
+      "epoch": 1.4932885906040267,
+      "grad_norm": 0.4949822723865509,
+      "learning_rate": 4.351124825669785e-06,
+      "loss": 0.5915,
+      "step": 1335
+    },
+    {
+      "epoch": 1.494407158836689,
+      "grad_norm": 0.48831161856651306,
+      "learning_rate": 4.350121693714531e-06,
+      "loss": 0.5983,
+      "step": 1336
+    },
+    {
+      "epoch": 1.4955257270693512,
+      "grad_norm": 0.5018520355224609,
+      "learning_rate": 4.349117902785297e-06,
+      "loss": 0.5872,
+      "step": 1337
+    },
+    {
+      "epoch": 1.4966442953020134,
+      "grad_norm": 0.49204087257385254,
+      "learning_rate": 4.3481134532396116e-06,
+      "loss": 0.5895,
+      "step": 1338
+    },
+    {
+      "epoch": 1.4977628635346756,
+      "grad_norm": 0.47245603799819946,
+      "learning_rate": 4.347108345435238e-06,
+      "loss": 0.5585,
+      "step": 1339
+    },
+    {
+      "epoch": 1.4988814317673378,
+      "grad_norm": 0.4722137749195099,
+      "learning_rate": 4.3461025797301745e-06,
+      "loss": 0.5454,
+      "step": 1340
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.4701415002346039,
+      "learning_rate": 4.345096156482655e-06,
+      "loss": 0.565,
+      "step": 1341
+    },
+    {
+      "epoch": 1.5011185682326622,
+      "grad_norm": 0.4957852065563202,
+      "learning_rate": 4.344089076051143e-06,
+      "loss": 0.6197,
+      "step": 1342
+    },
+    {
+      "epoch": 1.5022371364653244,
+      "grad_norm": 0.4819275140762329,
+      "learning_rate": 4.3430813387943405e-06,
+      "loss": 0.5635,
+      "step": 1343
+    },
+    {
+      "epoch": 1.5033557046979866,
+      "grad_norm": 0.5286908149719238,
+      "learning_rate": 4.342072945071183e-06,
+      "loss": 0.6067,
+      "step": 1344
+    },
+    {
+      "epoch": 1.5044742729306488,
+      "grad_norm": 0.4800012409687042,
+      "learning_rate": 4.3410638952408375e-06,
+      "loss": 0.58,
+      "step": 1345
+    },
+    {
+      "epoch": 1.505592841163311,
+      "grad_norm": 0.4997807443141937,
+      "learning_rate": 4.340054189662707e-06,
+      "loss": 0.5966,
+      "step": 1346
+    },
+    {
+      "epoch": 1.5067114093959733,
+      "grad_norm": 0.4921199083328247,
+      "learning_rate": 4.339043828696427e-06,
+      "loss": 0.5876,
+      "step": 1347
+    },
+    {
+      "epoch": 1.5078299776286355,
+      "grad_norm": 0.47939038276672363,
+      "learning_rate": 4.3380328127018666e-06,
+      "loss": 0.5758,
+      "step": 1348
+    },
+    {
+      "epoch": 1.5089485458612977,
+      "grad_norm": 0.5145376324653625,
+      "learning_rate": 4.337021142039127e-06,
+      "loss": 0.5815,
+      "step": 1349
+    },
+    {
+      "epoch": 1.5100671140939599,
+      "grad_norm": 0.4857310652732849,
+      "learning_rate": 4.336008817068546e-06,
+      "loss": 0.5686,
+      "step": 1350
+    },
+    {
+      "epoch": 1.5111856823266219,
+      "grad_norm": 0.4951683580875397,
+      "learning_rate": 4.33499583815069e-06,
+      "loss": 0.5872,
+      "step": 1351
+    },
+    {
+      "epoch": 1.512304250559284,
+      "grad_norm": 0.5085970163345337,
+      "learning_rate": 4.3339822056463624e-06,
+      "loss": 0.605,
+      "step": 1352
+    },
+    {
+      "epoch": 1.5134228187919463,
+      "grad_norm": 0.5023934841156006,
+      "learning_rate": 4.332967919916596e-06,
+      "loss": 0.5969,
+      "step": 1353
+    },
+    {
+      "epoch": 1.5145413870246085,
+      "grad_norm": 0.5029785633087158,
+      "learning_rate": 4.331952981322658e-06,
+      "loss": 0.5963,
+      "step": 1354
+    },
+    {
+      "epoch": 1.5156599552572707,
+      "grad_norm": 0.49423643946647644,
+      "learning_rate": 4.330937390226049e-06,
+      "loss": 0.5591,
+      "step": 1355
+    },
+    {
+      "epoch": 1.516778523489933,
+      "grad_norm": 0.48507627844810486,
+      "learning_rate": 4.3299211469885e-06,
+      "loss": 0.5816,
+      "step": 1356
+    },
+    {
+      "epoch": 1.5178970917225951,
+      "grad_norm": 0.496153861284256,
+      "learning_rate": 4.328904251971976e-06,
+      "loss": 0.5784,
+      "step": 1357
+    },
+    {
+      "epoch": 1.5190156599552571,
+      "grad_norm": 0.4916113018989563,
+      "learning_rate": 4.327886705538672e-06,
+      "loss": 0.5666,
+      "step": 1358
+    },
+    {
+      "epoch": 1.5201342281879193,
+      "grad_norm": 0.49055805802345276,
+      "learning_rate": 4.326868508051018e-06,
+      "loss": 0.5975,
+      "step": 1359
+    },
+    {
+      "epoch": 1.5212527964205815,
+      "grad_norm": 0.47937023639678955,
+      "learning_rate": 4.325849659871674e-06,
+      "loss": 0.5716,
+      "step": 1360
+    },
+    {
+      "epoch": 1.5223713646532437,
+      "grad_norm": 0.49050694704055786,
+      "learning_rate": 4.3248301613635306e-06,
+      "loss": 0.584,
+      "step": 1361
+    },
+    {
+      "epoch": 1.523489932885906,
+      "grad_norm": 0.4832146167755127,
+      "learning_rate": 4.323810012889713e-06,
+      "loss": 0.579,
+      "step": 1362
+    },
+    {
+      "epoch": 1.5246085011185682,
+      "grad_norm": 0.5009722709655762,
+      "learning_rate": 4.3227892148135755e-06,
+      "loss": 0.6014,
+      "step": 1363
+    },
+    {
+      "epoch": 1.5257270693512304,
+      "grad_norm": 0.4904753267765045,
+      "learning_rate": 4.321767767498705e-06,
+      "loss": 0.6009,
+      "step": 1364
+    },
+    {
+      "epoch": 1.5268456375838926,
+      "grad_norm": 0.5005730986595154,
+      "learning_rate": 4.32074567130892e-06,
+      "loss": 0.5975,
+      "step": 1365
+    },
+    {
+      "epoch": 1.5279642058165548,
+      "grad_norm": 0.4825516939163208,
+      "learning_rate": 4.319722926608268e-06,
+      "loss": 0.5726,
+      "step": 1366
+    },
+    {
+      "epoch": 1.529082774049217,
+      "grad_norm": 0.5070571899414062,
+      "learning_rate": 4.31869953376103e-06,
+      "loss": 0.6022,
+      "step": 1367
+    },
+    {
+      "epoch": 1.5302013422818792,
+      "grad_norm": 0.5236569046974182,
+      "learning_rate": 4.3176754931317154e-06,
+      "loss": 0.5973,
+      "step": 1368
+    },
+    {
+      "epoch": 1.5313199105145414,
+      "grad_norm": 0.49975070357322693,
+      "learning_rate": 4.316650805085068e-06,
+      "loss": 0.5881,
+      "step": 1369
+    },
+    {
+      "epoch": 1.5324384787472036,
+      "grad_norm": 0.49533969163894653,
+      "learning_rate": 4.315625469986058e-06,
+      "loss": 0.5748,
+      "step": 1370
+    },
+    {
+      "epoch": 1.5335570469798658,
+      "grad_norm": 0.4943258762359619,
+      "learning_rate": 4.314599488199889e-06,
+      "loss": 0.5832,
+      "step": 1371
+    },
+    {
+      "epoch": 1.534675615212528,
+      "grad_norm": 0.48157885670661926,
+      "learning_rate": 4.313572860091993e-06,
+      "loss": 0.5586,
+      "step": 1372
+    },
+    {
+      "epoch": 1.5357941834451903,
+      "grad_norm": 0.5046592354774475,
+      "learning_rate": 4.312545586028033e-06,
+      "loss": 0.5777,
+      "step": 1373
+    },
+    {
+      "epoch": 1.5369127516778525,
+      "grad_norm": 0.5014451146125793,
+      "learning_rate": 4.311517666373902e-06,
+      "loss": 0.5875,
+      "step": 1374
+    },
+    {
+      "epoch": 1.5380313199105147,
+      "grad_norm": 0.5050190687179565,
+      "learning_rate": 4.310489101495725e-06,
+      "loss": 0.6041,
+      "step": 1375
+    },
+    {
+      "epoch": 1.5391498881431769,
+      "grad_norm": 0.49115487933158875,
+      "learning_rate": 4.309459891759853e-06,
+      "loss": 0.5888,
+      "step": 1376
+    },
+    {
+      "epoch": 1.540268456375839,
+      "grad_norm": 0.5117396712303162,
+      "learning_rate": 4.308430037532871e-06,
+      "loss": 0.5689,
+      "step": 1377
+    },
+    {
+      "epoch": 1.541387024608501,
+      "grad_norm": 0.4883919656276703,
+      "learning_rate": 4.307399539181587e-06,
+      "loss": 0.5982,
+      "step": 1378
+    },
+    {
+      "epoch": 1.5425055928411633,
+      "grad_norm": 0.48665109276771545,
+      "learning_rate": 4.306368397073046e-06,
+      "loss": 0.5899,
+      "step": 1379
+    },
+    {
+      "epoch": 1.5436241610738255,
+      "grad_norm": 0.4853532910346985,
+      "learning_rate": 4.305336611574518e-06,
+      "loss": 0.5695,
+      "step": 1380
+    },
+    {
+      "epoch": 1.5447427293064877,
+      "grad_norm": 0.4923276901245117,
+      "learning_rate": 4.304304183053502e-06,
+      "loss": 0.5831,
+      "step": 1381
+    },
+    {
+      "epoch": 1.54586129753915,
+      "grad_norm": 0.5048404335975647,
+      "learning_rate": 4.303271111877729e-06,
+      "loss": 0.6238,
+      "step": 1382
+    },
+    {
+      "epoch": 1.5469798657718121,
+      "grad_norm": 0.49105462431907654,
+      "learning_rate": 4.302237398415156e-06,
+      "loss": 0.5976,
+      "step": 1383
+    },
+    {
+      "epoch": 1.548098434004474,
+      "grad_norm": 0.4881378412246704,
+      "learning_rate": 4.301203043033969e-06,
+      "loss": 0.5739,
+      "step": 1384
+    },
+    {
+      "epoch": 1.5492170022371363,
+      "grad_norm": 0.48858603835105896,
+      "learning_rate": 4.3001680461025844e-06,
+      "loss": 0.5792,
+      "step": 1385
+    },
+    {
+      "epoch": 1.5503355704697985,
+      "grad_norm": 0.48439061641693115,
+      "learning_rate": 4.299132407989646e-06,
+      "loss": 0.5742,
+      "step": 1386
+    },
+    {
+      "epoch": 1.5514541387024607,
+      "grad_norm": 0.47701379656791687,
+      "learning_rate": 4.298096129064026e-06,
+      "loss": 0.5844,
+      "step": 1387
+    },
+    {
+      "epoch": 1.552572706935123,
+      "grad_norm": 0.49393609166145325,
+      "learning_rate": 4.297059209694824e-06,
+      "loss": 0.5886,
+      "step": 1388
+    },
+    {
+      "epoch": 1.5536912751677852,
+      "grad_norm": 0.5089290738105774,
+      "learning_rate": 4.296021650251369e-06,
+      "loss": 0.5856,
+      "step": 1389
+    },
+    {
+      "epoch": 1.5548098434004474,
+      "grad_norm": 0.509134829044342,
+      "learning_rate": 4.294983451103219e-06,
+      "loss": 0.5939,
+      "step": 1390
+    },
+    {
+      "epoch": 1.5559284116331096,
+      "grad_norm": 0.5126070976257324,
+      "learning_rate": 4.293944612620157e-06,
+      "loss": 0.5692,
+      "step": 1391
+    },
+    {
+      "epoch": 1.5570469798657718,
+      "grad_norm": 0.5223609805107117,
+      "learning_rate": 4.2929051351721956e-06,
+      "loss": 0.599,
+      "step": 1392
+    },
+    {
+      "epoch": 1.558165548098434,
+      "grad_norm": 0.48266369104385376,
+      "learning_rate": 4.291865019129575e-06,
+      "loss": 0.5827,
+      "step": 1393
+    },
+    {
+      "epoch": 1.5592841163310962,
+      "grad_norm": 0.4978017210960388,
+      "learning_rate": 4.290824264862761e-06,
+      "loss": 0.5994,
+      "step": 1394
+    },
+    {
+      "epoch": 1.5604026845637584,
+      "grad_norm": 0.5075487494468689,
+      "learning_rate": 4.2897828727424495e-06,
+      "loss": 0.5641,
+      "step": 1395
+    },
+    {
+      "epoch": 1.5615212527964206,
+      "grad_norm": 0.5072858929634094,
+      "learning_rate": 4.2887408431395615e-06,
+      "loss": 0.5751,
+      "step": 1396
+    },
+    {
+      "epoch": 1.5626398210290828,
+      "grad_norm": 0.513059139251709,
+      "learning_rate": 4.287698176425246e-06,
+      "loss": 0.5909,
+      "step": 1397
+    },
+    {
+      "epoch": 1.563758389261745,
+      "grad_norm": 0.49780142307281494,
+      "learning_rate": 4.286654872970879e-06,
+      "loss": 0.5928,
+      "step": 1398
+    },
+    {
+      "epoch": 1.5648769574944073,
+      "grad_norm": 0.4992021918296814,
+      "learning_rate": 4.285610933148062e-06,
+      "loss": 0.583,
+      "step": 1399
+    },
+    {
+      "epoch": 1.5659955257270695,
+      "grad_norm": 0.4856356978416443,
+      "learning_rate": 4.284566357328625e-06,
+      "loss": 0.6213,
+      "step": 1400
+    },
+    {
+      "epoch": 1.5671140939597317,
+      "grad_norm": 0.48052698373794556,
+      "learning_rate": 4.283521145884625e-06,
+      "loss": 0.5753,
+      "step": 1401
+    },
+    {
+      "epoch": 1.5682326621923939,
+      "grad_norm": 0.4759695827960968,
+      "learning_rate": 4.2824752991883415e-06,
+      "loss": 0.5764,
+      "step": 1402
+    },
+    {
+      "epoch": 1.569351230425056,
+      "grad_norm": 0.49083444476127625,
+      "learning_rate": 4.2814288176122846e-06,
+      "loss": 0.5839,
+      "step": 1403
+    },
+    {
+      "epoch": 1.570469798657718,
+      "grad_norm": 0.4828498661518097,
+      "learning_rate": 4.280381701529187e-06,
+      "loss": 0.5734,
+      "step": 1404
+    },
+    {
+      "epoch": 1.5715883668903803,
+      "grad_norm": 0.4739566147327423,
+      "learning_rate": 4.27933395131201e-06,
+      "loss": 0.5958,
+      "step": 1405
+    },
+    {
+      "epoch": 1.5727069351230425,
+      "grad_norm": 0.4890132546424866,
+      "learning_rate": 4.278285567333942e-06,
+      "loss": 0.556,
+      "step": 1406
+    },
+    {
+      "epoch": 1.5738255033557047,
+      "grad_norm": 0.49774521589279175,
+      "learning_rate": 4.277236549968392e-06,
+      "loss": 0.5779,
+      "step": 1407
+    },
+    {
+      "epoch": 1.574944071588367,
+      "grad_norm": 0.5137577652931213,
+      "learning_rate": 4.276186899588999e-06,
+      "loss": 0.5947,
+      "step": 1408
+    },
+    {
+      "epoch": 1.5760626398210291,
+      "grad_norm": 0.5089177489280701,
+      "learning_rate": 4.275136616569626e-06,
+      "loss": 0.5825,
+      "step": 1409
+    },
+    {
+      "epoch": 1.5771812080536913,
+      "grad_norm": 0.49170440435409546,
+      "learning_rate": 4.2740857012843625e-06,
+      "loss": 0.5887,
+      "step": 1410
+    },
+    {
+      "epoch": 1.5782997762863533,
+      "grad_norm": 0.48788514733314514,
+      "learning_rate": 4.27303415410752e-06,
+      "loss": 0.565,
+      "step": 1411
+    },
+    {
+      "epoch": 1.5794183445190155,
+      "grad_norm": 0.4836443364620209,
+      "learning_rate": 4.2719819754136395e-06,
+      "loss": 0.5679,
+      "step": 1412
+    },
+    {
+      "epoch": 1.5805369127516777,
+      "grad_norm": 0.4887832999229431,
+      "learning_rate": 4.270929165577483e-06,
+      "loss": 0.5954,
+      "step": 1413
+    },
+    {
+      "epoch": 1.58165548098434,
+      "grad_norm": 0.5417852401733398,
+      "learning_rate": 4.26987572497404e-06,
+      "loss": 0.5747,
+      "step": 1414
+    },
+    {
+      "epoch": 1.5827740492170022,
+      "grad_norm": 0.4931381642818451,
+      "learning_rate": 4.268821653978522e-06,
+      "loss": 0.5964,
+      "step": 1415
+    },
+    {
+      "epoch": 1.5838926174496644,
+      "grad_norm": 0.48656824231147766,
+      "learning_rate": 4.267766952966369e-06,
+      "loss": 0.5919,
+      "step": 1416
+    },
+    {
+      "epoch": 1.5850111856823266,
+      "grad_norm": 0.5253188610076904,
+      "learning_rate": 4.266711622313242e-06,
+      "loss": 0.5917,
+      "step": 1417
+    },
+    {
+      "epoch": 1.5861297539149888,
+      "grad_norm": 0.5214728713035583,
+      "learning_rate": 4.2656556623950265e-06,
+      "loss": 0.5545,
+      "step": 1418
+    },
+    {
+      "epoch": 1.587248322147651,
+      "grad_norm": 0.4981340169906616,
+      "learning_rate": 4.264599073587834e-06,
+      "loss": 0.5737,
+      "step": 1419
+    },
+    {
+      "epoch": 1.5883668903803132,
+      "grad_norm": 0.49596238136291504,
+      "learning_rate": 4.263541856267999e-06,
+      "loss": 0.5841,
+      "step": 1420
+    },
+    {
+      "epoch": 1.5894854586129754,
+      "grad_norm": 0.47956663370132446,
+      "learning_rate": 4.262484010812079e-06,
+      "loss": 0.5566,
+      "step": 1421
+    },
+    {
+      "epoch": 1.5906040268456376,
+      "grad_norm": 0.4902280569076538,
+      "learning_rate": 4.261425537596857e-06,
+      "loss": 0.5907,
+      "step": 1422
+    },
+    {
+      "epoch": 1.5917225950782998,
+      "grad_norm": 0.5098991990089417,
+      "learning_rate": 4.260366436999338e-06,
+      "loss": 0.5997,
+      "step": 1423
+    },
+    {
+      "epoch": 1.592841163310962,
+      "grad_norm": 0.48146405816078186,
+      "learning_rate": 4.259306709396751e-06,
+      "loss": 0.5693,
+      "step": 1424
+    },
+    {
+      "epoch": 1.5939597315436242,
+      "grad_norm": 0.4839426577091217,
+      "learning_rate": 4.258246355166548e-06,
+      "loss": 0.5734,
+      "step": 1425
+    },
+    {
+      "epoch": 1.5950782997762865,
+      "grad_norm": 0.4717272222042084,
+      "learning_rate": 4.257185374686405e-06,
+      "loss": 0.5716,
+      "step": 1426
+    },
+    {
+      "epoch": 1.5961968680089487,
+      "grad_norm": 0.4848939776420593,
+      "learning_rate": 4.256123768334223e-06,
+      "loss": 0.5592,
+      "step": 1427
+    },
+    {
+      "epoch": 1.5973154362416109,
+      "grad_norm": 0.505172073841095,
+      "learning_rate": 4.25506153648812e-06,
+      "loss": 0.5822,
+      "step": 1428
+    },
+    {
+      "epoch": 1.598434004474273,
+      "grad_norm": 0.5064011216163635,
+      "learning_rate": 4.253998679526442e-06,
+      "loss": 0.583,
+      "step": 1429
+    },
+    {
+      "epoch": 1.599552572706935,
+      "grad_norm": 0.5151503086090088,
+      "learning_rate": 4.252935197827756e-06,
+      "loss": 0.6022,
+      "step": 1430
+    },
+    {
+      "epoch": 1.6006711409395973,
+      "grad_norm": 0.48397284746170044,
+      "learning_rate": 4.251871091770852e-06,
+      "loss": 0.5859,
+      "step": 1431
+    },
+    {
+      "epoch": 1.6017897091722595,
+      "grad_norm": 0.48410117626190186,
+      "learning_rate": 4.2508063617347415e-06,
+      "loss": 0.5724,
+      "step": 1432
+    },
+    {
+      "epoch": 1.6029082774049217,
+      "grad_norm": 0.4976502060890198,
+      "learning_rate": 4.249741008098658e-06,
+      "loss": 0.5968,
+      "step": 1433
+    },
+    {
+      "epoch": 1.604026845637584,
+      "grad_norm": 0.5036817193031311,
+      "learning_rate": 4.2486750312420585e-06,
+      "loss": 0.5922,
+      "step": 1434
+    },
+    {
+      "epoch": 1.6051454138702461,
+      "grad_norm": 0.5002356171607971,
+      "learning_rate": 4.247608431544622e-06,
+      "loss": 0.5829,
+      "step": 1435
+    },
+    {
+      "epoch": 1.6062639821029083,
+      "grad_norm": 0.4937509000301361,
+      "learning_rate": 4.246541209386247e-06,
+      "loss": 0.578,
+      "step": 1436
+    },
+    {
+      "epoch": 1.6073825503355703,
+      "grad_norm": 0.489439457654953,
+      "learning_rate": 4.245473365147056e-06,
+      "loss": 0.5793,
+      "step": 1437
+    },
+    {
+      "epoch": 1.6085011185682325,
+      "grad_norm": 0.48611974716186523,
+      "learning_rate": 4.244404899207393e-06,
+      "loss": 0.5952,
+      "step": 1438
+    },
+    {
+      "epoch": 1.6096196868008947,
+      "grad_norm": 0.48233819007873535,
+      "learning_rate": 4.2433358119478215e-06,
+      "loss": 0.586,
+      "step": 1439
+    },
+    {
+      "epoch": 1.610738255033557,
+      "grad_norm": 0.4946385324001312,
+      "learning_rate": 4.24226610374913e-06,
+      "loss": 0.5832,
+      "step": 1440
+    },
+    {
+      "epoch": 1.6118568232662192,
+      "grad_norm": 0.49822622537612915,
+      "learning_rate": 4.241195774992323e-06,
+      "loss": 0.5929,
+      "step": 1441
+    },
+    {
+      "epoch": 1.6129753914988814,
+      "grad_norm": 0.4824593663215637,
+      "learning_rate": 4.240124826058631e-06,
+      "loss": 0.5603,
+      "step": 1442
+    },
+    {
+      "epoch": 1.6140939597315436,
+      "grad_norm": 0.4935947358608246,
+      "learning_rate": 4.239053257329502e-06,
+      "loss": 0.5708,
+      "step": 1443
+    },
+    {
+      "epoch": 1.6152125279642058,
+      "grad_norm": 0.48947572708129883,
+      "learning_rate": 4.237981069186606e-06,
+      "loss": 0.5576,
+      "step": 1444
+    },
+    {
+      "epoch": 1.616331096196868,
+      "grad_norm": 0.4911848306655884,
+      "learning_rate": 4.236908262011834e-06,
+      "loss": 0.6014,
+      "step": 1445
+    },
+    {
+      "epoch": 1.6174496644295302,
+      "grad_norm": 0.5052220225334167,
+      "learning_rate": 4.2358348361872975e-06,
+      "loss": 0.6186,
+      "step": 1446
+    },
+    {
+      "epoch": 1.6185682326621924,
+      "grad_norm": 0.47427546977996826,
+      "learning_rate": 4.234760792095327e-06,
+      "loss": 0.554,
+      "step": 1447
+    },
+    {
+      "epoch": 1.6196868008948546,
+      "grad_norm": 0.495304673910141,
+      "learning_rate": 4.2336861301184754e-06,
+      "loss": 0.6066,
+      "step": 1448
+    },
+    {
+      "epoch": 1.6208053691275168,
+      "grad_norm": 0.49640989303588867,
+      "learning_rate": 4.2326108506395125e-06,
+      "loss": 0.5935,
+      "step": 1449
+    },
+    {
+      "epoch": 1.621923937360179,
+      "grad_norm": 0.4849659204483032,
+      "learning_rate": 4.231534954041432e-06,
+      "loss": 0.5707,
+      "step": 1450
+    },
+    {
+      "epoch": 1.6230425055928412,
+      "grad_norm": 0.48401832580566406,
+      "learning_rate": 4.230458440707443e-06,
+      "loss": 0.6059,
+      "step": 1451
+    },
+    {
+      "epoch": 1.6241610738255035,
+      "grad_norm": 0.48183032870292664,
+      "learning_rate": 4.2293813110209795e-06,
+      "loss": 0.5926,
+      "step": 1452
+    },
+    {
+      "epoch": 1.6252796420581657,
+      "grad_norm": 0.47466862201690674,
+      "learning_rate": 4.22830356536569e-06,
+      "loss": 0.5472,
+      "step": 1453
+    },
+    {
+      "epoch": 1.6263982102908279,
+      "grad_norm": 0.4778653383255005,
+      "learning_rate": 4.227225204125447e-06,
+      "loss": 0.5331,
+      "step": 1454
+    },
+    {
+      "epoch": 1.62751677852349,
+      "grad_norm": 0.49169987440109253,
+      "learning_rate": 4.226146227684337e-06,
+      "loss": 0.5892,
+      "step": 1455
+    },
+    {
+      "epoch": 1.6286353467561523,
+      "grad_norm": 0.4861750304698944,
+      "learning_rate": 4.225066636426669e-06,
+      "loss": 0.5902,
+      "step": 1456
+    },
+    {
+      "epoch": 1.6297539149888143,
+      "grad_norm": 0.5020110607147217,
+      "learning_rate": 4.223986430736972e-06,
+      "loss": 0.5878,
+      "step": 1457
+    },
+    {
+      "epoch": 1.6308724832214765,
+      "grad_norm": 0.5259606242179871,
+      "learning_rate": 4.2229056109999915e-06,
+      "loss": 0.5855,
+      "step": 1458
+    },
+    {
+      "epoch": 1.6319910514541387,
+      "grad_norm": 0.5020241737365723,
+      "learning_rate": 4.221824177600692e-06,
+      "loss": 0.5728,
+      "step": 1459
+    },
+    {
+      "epoch": 1.633109619686801,
+      "grad_norm": 0.4941030740737915,
+      "learning_rate": 4.220742130924257e-06,
+      "loss": 0.5834,
+      "step": 1460
+    },
+    {
+      "epoch": 1.6342281879194631,
+      "grad_norm": 0.49204495549201965,
+      "learning_rate": 4.21965947135609e-06,
+      "loss": 0.6003,
+      "step": 1461
+    },
+    {
+      "epoch": 1.6353467561521253,
+      "grad_norm": 0.5063990950584412,
+      "learning_rate": 4.218576199281809e-06,
+      "loss": 0.5886,
+      "step": 1462
+    },
+    {
+      "epoch": 1.6364653243847873,
+      "grad_norm": 0.49553683400154114,
+      "learning_rate": 4.217492315087255e-06,
+      "loss": 0.5917,
+      "step": 1463
+    },
+    {
+      "epoch": 1.6375838926174495,
+      "grad_norm": 0.48495718836784363,
+      "learning_rate": 4.216407819158482e-06,
+      "loss": 0.5649,
+      "step": 1464
+    },
+    {
+      "epoch": 1.6387024608501117,
+      "grad_norm": 0.4799686074256897,
+      "learning_rate": 4.215322711881766e-06,
+      "loss": 0.5572,
+      "step": 1465
+    },
+    {
+      "epoch": 1.639821029082774,
+      "grad_norm": 0.5015504956245422,
+      "learning_rate": 4.2142369936435986e-06,
+      "loss": 0.5829,
+      "step": 1466
+    },
+    {
+      "epoch": 1.6409395973154361,
+      "grad_norm": 0.5071560144424438,
+      "learning_rate": 4.21315066483069e-06,
+      "loss": 0.6074,
+      "step": 1467
+    },
+    {
+      "epoch": 1.6420581655480984,
+      "grad_norm": 0.5041237473487854,
+      "learning_rate": 4.212063725829966e-06,
+      "loss": 0.5996,
+      "step": 1468
+    },
+    {
+      "epoch": 1.6431767337807606,
+      "grad_norm": 0.5061616897583008,
+      "learning_rate": 4.210976177028573e-06,
+      "loss": 0.6098,
+      "step": 1469
+    },
+    {
+      "epoch": 1.6442953020134228,
+      "grad_norm": 0.488351047039032,
+      "learning_rate": 4.209888018813872e-06,
+      "loss": 0.5902,
+      "step": 1470
+    },
+    {
+      "epoch": 1.645413870246085,
+      "grad_norm": 0.494767427444458,
+      "learning_rate": 4.208799251573441e-06,
+      "loss": 0.5632,
+      "step": 1471
+    },
+    {
+      "epoch": 1.6465324384787472,
+      "grad_norm": 0.5081776976585388,
+      "learning_rate": 4.207709875695078e-06,
+      "loss": 0.6086,
+      "step": 1472
+    },
+    {
+      "epoch": 1.6476510067114094,
+      "grad_norm": 0.5012889504432678,
+      "learning_rate": 4.206619891566792e-06,
+      "loss": 0.5879,
+      "step": 1473
+    },
+    {
+      "epoch": 1.6487695749440716,
+      "grad_norm": 0.5006359219551086,
+      "learning_rate": 4.2055292995768145e-06,
+      "loss": 0.5648,
+      "step": 1474
+    },
+    {
+      "epoch": 1.6498881431767338,
+      "grad_norm": 0.4885180592536926,
+      "learning_rate": 4.204438100113592e-06,
+      "loss": 0.5629,
+      "step": 1475
+    },
+    {
+      "epoch": 1.651006711409396,
+      "grad_norm": 0.4856667220592499,
+      "learning_rate": 4.203346293565784e-06,
+      "loss": 0.598,
+      "step": 1476
+    },
+    {
+      "epoch": 1.6521252796420582,
+      "grad_norm": 0.484739750623703,
+      "learning_rate": 4.2022538803222714e-06,
+      "loss": 0.5862,
+      "step": 1477
+    },
+    {
+      "epoch": 1.6532438478747205,
+      "grad_norm": 0.4918213188648224,
+      "learning_rate": 4.2011608607721455e-06,
+      "loss": 0.5699,
+      "step": 1478
+    },
+    {
+      "epoch": 1.6543624161073827,
+      "grad_norm": 0.5036900639533997,
+      "learning_rate": 4.200067235304719e-06,
+      "loss": 0.6143,
+      "step": 1479
+    },
+    {
+      "epoch": 1.6554809843400449,
+      "grad_norm": 0.4958629310131073,
+      "learning_rate": 4.1989730043095175e-06,
+      "loss": 0.5423,
+      "step": 1480
+    },
+    {
+      "epoch": 1.656599552572707,
+      "grad_norm": 0.49514490365982056,
+      "learning_rate": 4.1978781681762825e-06,
+      "loss": 0.6092,
+      "step": 1481
+    },
+    {
+      "epoch": 1.6577181208053693,
+      "grad_norm": 0.5146138668060303,
+      "learning_rate": 4.1967827272949715e-06,
+      "loss": 0.5907,
+      "step": 1482
+    },
+    {
+      "epoch": 1.6588366890380313,
+      "grad_norm": 0.5063082575798035,
+      "learning_rate": 4.195686682055758e-06,
+      "loss": 0.5779,
+      "step": 1483
+    },
+    {
+      "epoch": 1.6599552572706935,
+      "grad_norm": 0.48410534858703613,
+      "learning_rate": 4.194590032849028e-06,
+      "loss": 0.5864,
+      "step": 1484
+    },
+    {
+      "epoch": 1.6610738255033557,
+      "grad_norm": 0.5001199841499329,
+      "learning_rate": 4.193492780065386e-06,
+      "loss": 0.5479,
+      "step": 1485
+    },
+    {
+      "epoch": 1.662192393736018,
+      "grad_norm": 0.49552416801452637,
+      "learning_rate": 4.19239492409565e-06,
+      "loss": 0.5646,
+      "step": 1486
+    },
+    {
+      "epoch": 1.6633109619686801,
+      "grad_norm": 0.4871010482311249,
+      "learning_rate": 4.191296465330853e-06,
+      "loss": 0.5601,
+      "step": 1487
+    },
+    {
+      "epoch": 1.6644295302013423,
+      "grad_norm": 0.5030876398086548,
+      "learning_rate": 4.190197404162242e-06,
+      "loss": 0.5892,
+      "step": 1488
+    },
+    {
+      "epoch": 1.6655480984340043,
+      "grad_norm": 0.4873735010623932,
+      "learning_rate": 4.18909774098128e-06,
+      "loss": 0.5981,
+      "step": 1489
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.5151247978210449,
+      "learning_rate": 4.187997476179643e-06,
+      "loss": 0.5813,
+      "step": 1490
+    },
+    {
+      "epoch": 1.6677852348993287,
+      "grad_norm": 0.4993012249469757,
+      "learning_rate": 4.1868966101492225e-06,
+      "loss": 0.5718,
+      "step": 1491
+    },
+    {
+      "epoch": 1.668903803131991,
+      "grad_norm": 0.48291707038879395,
+      "learning_rate": 4.1857951432821235e-06,
+      "loss": 0.566,
+      "step": 1492
+    },
+    {
+      "epoch": 1.6700223713646531,
+      "grad_norm": 0.49533194303512573,
+      "learning_rate": 4.184693075970665e-06,
+      "loss": 0.5956,
+      "step": 1493
+    },
+    {
+      "epoch": 1.6711409395973154,
+      "grad_norm": 0.5048168897628784,
+      "learning_rate": 4.183590408607379e-06,
+      "loss": 0.5735,
+      "step": 1494
+    },
+    {
+      "epoch": 1.6722595078299776,
+      "grad_norm": 0.4690997004508972,
+      "learning_rate": 4.1824871415850135e-06,
+      "loss": 0.5551,
+      "step": 1495
+    },
+    {
+      "epoch": 1.6733780760626398,
+      "grad_norm": 0.4947203993797302,
+      "learning_rate": 4.1813832752965275e-06,
+      "loss": 0.5877,
+      "step": 1496
+    },
+    {
+      "epoch": 1.674496644295302,
+      "grad_norm": 0.508201539516449,
+      "learning_rate": 4.180278810135096e-06,
+      "loss": 0.6059,
+      "step": 1497
+    },
+    {
+      "epoch": 1.6756152125279642,
+      "grad_norm": 0.4906938672065735,
+      "learning_rate": 4.179173746494105e-06,
+      "loss": 0.575,
+      "step": 1498
+    },
+    {
+      "epoch": 1.6767337807606264,
+      "grad_norm": 0.48704689741134644,
+      "learning_rate": 4.178068084767155e-06,
+      "loss": 0.5701,
+      "step": 1499
+    },
+    {
+      "epoch": 1.6778523489932886,
+      "grad_norm": 0.5050029754638672,
+      "learning_rate": 4.176961825348059e-06,
+      "loss": 0.5727,
+      "step": 1500
+    },
+    {
+      "epoch": 1.6789709172259508,
+      "grad_norm": 0.4842473566532135,
+      "learning_rate": 4.175854968630843e-06,
+      "loss": 0.5682,
+      "step": 1501
+    },
+    {
+      "epoch": 1.680089485458613,
+      "grad_norm": 0.5032848119735718,
+      "learning_rate": 4.1747475150097465e-06,
+      "loss": 0.5873,
+      "step": 1502
+    },
+    {
+      "epoch": 1.6812080536912752,
+      "grad_norm": 0.497781902551651,
+      "learning_rate": 4.17363946487922e-06,
+      "loss": 0.5914,
+      "step": 1503
+    },
+    {
+      "epoch": 1.6823266219239374,
+      "grad_norm": 0.5035775303840637,
+      "learning_rate": 4.172530818633929e-06,
+      "loss": 0.5519,
+      "step": 1504
+    },
+    {
+      "epoch": 1.6834451901565997,
+      "grad_norm": 0.47141513228416443,
+      "learning_rate": 4.171421576668747e-06,
+      "loss": 0.5913,
+      "step": 1505
+    },
+    {
+      "epoch": 1.6845637583892619,
+      "grad_norm": 0.49259573221206665,
+      "learning_rate": 4.170311739378765e-06,
+      "loss": 0.5769,
+      "step": 1506
+    },
+    {
+      "epoch": 1.685682326621924,
+      "grad_norm": 0.5033203363418579,
+      "learning_rate": 4.169201307159282e-06,
+      "loss": 0.5851,
+      "step": 1507
+    },
+    {
+      "epoch": 1.6868008948545863,
+      "grad_norm": 0.48197343945503235,
+      "learning_rate": 4.1680902804058095e-06,
+      "loss": 0.5502,
+      "step": 1508
+    },
+    {
+      "epoch": 1.6879194630872483,
+      "grad_norm": 0.5090881586074829,
+      "learning_rate": 4.166978659514075e-06,
+      "loss": 0.5684,
+      "step": 1509
+    },
+    {
+      "epoch": 1.6890380313199105,
+      "grad_norm": 0.4990635812282562,
+      "learning_rate": 4.1658664448800105e-06,
+      "loss": 0.5961,
+      "step": 1510
+    },
+    {
+      "epoch": 1.6901565995525727,
+      "grad_norm": 0.522132396697998,
+      "learning_rate": 4.164753636899765e-06,
+      "loss": 0.5941,
+      "step": 1511
+    },
+    {
+      "epoch": 1.691275167785235,
+      "grad_norm": 0.4907796382904053,
+      "learning_rate": 4.163640235969696e-06,
+      "loss": 0.5514,
+      "step": 1512
+    },
+    {
+      "epoch": 1.692393736017897,
+      "grad_norm": 0.49310702085494995,
+      "learning_rate": 4.1625262424863744e-06,
+      "loss": 0.5723,
+      "step": 1513
+    },
+    {
+      "epoch": 1.6935123042505593,
+      "grad_norm": 0.5064623951911926,
+      "learning_rate": 4.161411656846581e-06,
+      "loss": 0.5847,
+      "step": 1514
+    },
+    {
+      "epoch": 1.6946308724832215,
+      "grad_norm": 0.47868281602859497,
+      "learning_rate": 4.1602964794473065e-06,
+      "loss": 0.5757,
+      "step": 1515
+    },
+    {
+      "epoch": 1.6957494407158835,
+      "grad_norm": 0.5030828714370728,
+      "learning_rate": 4.159180710685753e-06,
+      "loss": 0.6001,
+      "step": 1516
+    },
+    {
+      "epoch": 1.6968680089485457,
+      "grad_norm": 0.4966081976890564,
+      "learning_rate": 4.158064350959336e-06,
+      "loss": 0.5643,
+      "step": 1517
+    },
+    {
+      "epoch": 1.697986577181208,
+      "grad_norm": 0.4922417104244232,
+      "learning_rate": 4.156947400665677e-06,
+      "loss": 0.5866,
+      "step": 1518
+    },
+    {
+      "epoch": 1.6991051454138701,
+      "grad_norm": 0.4993835985660553,
+      "learning_rate": 4.15582986020261e-06,
+      "loss": 0.5784,
+      "step": 1519
+    },
+    {
+      "epoch": 1.7002237136465324,
+      "grad_norm": 0.4877612292766571,
+      "learning_rate": 4.15471172996818e-06,
+      "loss": 0.5736,
+      "step": 1520
+    },
+    {
+      "epoch": 1.7013422818791946,
+      "grad_norm": 0.5129001140594482,
+      "learning_rate": 4.1535930103606406e-06,
+      "loss": 0.5798,
+      "step": 1521
+    },
+    {
+      "epoch": 1.7024608501118568,
+      "grad_norm": 0.48489999771118164,
+      "learning_rate": 4.152473701778455e-06,
+      "loss": 0.556,
+      "step": 1522
+    },
+    {
+      "epoch": 1.703579418344519,
+      "grad_norm": 0.4901261627674103,
+      "learning_rate": 4.1513538046202995e-06,
+      "loss": 0.578,
+      "step": 1523
+    },
+    {
+      "epoch": 1.7046979865771812,
+      "grad_norm": 0.49826791882514954,
+      "learning_rate": 4.150233319285055e-06,
+      "loss": 0.5671,
+      "step": 1524
+    },
+    {
+      "epoch": 1.7058165548098434,
+      "grad_norm": 0.49227479100227356,
+      "learning_rate": 4.149112246171817e-06,
+      "loss": 0.5561,
+      "step": 1525
+    },
+    {
+      "epoch": 1.7069351230425056,
+      "grad_norm": 0.49544984102249146,
+      "learning_rate": 4.147990585679886e-06,
+      "loss": 0.5682,
+      "step": 1526
+    },
+    {
+      "epoch": 1.7080536912751678,
+      "grad_norm": 0.49339771270751953,
+      "learning_rate": 4.146868338208775e-06,
+      "loss": 0.5607,
+      "step": 1527
+    },
+    {
+      "epoch": 1.70917225950783,
+      "grad_norm": 0.506488561630249,
+      "learning_rate": 4.1457455041582044e-06,
+      "loss": 0.5778,
+      "step": 1528
+    },
+    {
+      "epoch": 1.7102908277404922,
+      "grad_norm": 0.5283234715461731,
+      "learning_rate": 4.144622083928102e-06,
+      "loss": 0.5777,
+      "step": 1529
+    },
+    {
+      "epoch": 1.7114093959731544,
+      "grad_norm": 0.5093042254447937,
+      "learning_rate": 4.14349807791861e-06,
+      "loss": 0.5832,
+      "step": 1530
+    },
+    {
+      "epoch": 1.7125279642058167,
+      "grad_norm": 0.49732184410095215,
+      "learning_rate": 4.142373486530071e-06,
+      "loss": 0.5965,
+      "step": 1531
+    },
+    {
+      "epoch": 1.7136465324384789,
+      "grad_norm": 0.5034535527229309,
+      "learning_rate": 4.141248310163042e-06,
+      "loss": 0.5896,
+      "step": 1532
+    },
+    {
+      "epoch": 1.714765100671141,
+      "grad_norm": 0.484789103269577,
+      "learning_rate": 4.140122549218289e-06,
+      "loss": 0.5636,
+      "step": 1533
+    },
+    {
+      "epoch": 1.7158836689038033,
+      "grad_norm": 0.49596208333969116,
+      "learning_rate": 4.138996204096781e-06,
+      "loss": 0.5794,
+      "step": 1534
+    },
+    {
+      "epoch": 1.7170022371364653,
+      "grad_norm": 0.4908808767795563,
+      "learning_rate": 4.137869275199701e-06,
+      "loss": 0.5846,
+      "step": 1535
+    },
+    {
+      "epoch": 1.7181208053691275,
+      "grad_norm": 0.4921574890613556,
+      "learning_rate": 4.1367417629284356e-06,
+      "loss": 0.5834,
+      "step": 1536
+    },
+    {
+      "epoch": 1.7192393736017897,
+      "grad_norm": 0.48451635241508484,
+      "learning_rate": 4.13561366768458e-06,
+      "loss": 0.5624,
+      "step": 1537
+    },
+    {
+      "epoch": 1.720357941834452,
+      "grad_norm": 0.5193755626678467,
+      "learning_rate": 4.134484989869939e-06,
+      "loss": 0.5908,
+      "step": 1538
+    },
+    {
+      "epoch": 1.721476510067114,
+      "grad_norm": 0.497798889875412,
+      "learning_rate": 4.133355729886523e-06,
+      "loss": 0.5668,
+      "step": 1539
+    },
+    {
+      "epoch": 1.7225950782997763,
+      "grad_norm": 0.5026369690895081,
+      "learning_rate": 4.132225888136552e-06,
+      "loss": 0.5932,
+      "step": 1540
+    },
+    {
+      "epoch": 1.7237136465324385,
+      "grad_norm": 0.5131746530532837,
+      "learning_rate": 4.131095465022449e-06,
+      "loss": 0.5965,
+      "step": 1541
+    },
+    {
+      "epoch": 1.7248322147651005,
+      "grad_norm": 0.4941963851451874,
+      "learning_rate": 4.129964460946847e-06,
+      "loss": 0.5797,
+      "step": 1542
+    },
+    {
+      "epoch": 1.7259507829977627,
+      "grad_norm": 0.5222018957138062,
+      "learning_rate": 4.128832876312587e-06,
+      "loss": 0.5948,
+      "step": 1543
+    },
+    {
+      "epoch": 1.727069351230425,
+      "grad_norm": 0.5069666504859924,
+      "learning_rate": 4.127700711522715e-06,
+      "loss": 0.5855,
+      "step": 1544
+    },
+    {
+      "epoch": 1.7281879194630871,
+      "grad_norm": 0.5024001002311707,
+      "learning_rate": 4.126567966980484e-06,
+      "loss": 0.6114,
+      "step": 1545
+    },
+    {
+      "epoch": 1.7293064876957494,
+      "grad_norm": 0.49587997794151306,
+      "learning_rate": 4.125434643089353e-06,
+      "loss": 0.568,
+      "step": 1546
+    },
+    {
+      "epoch": 1.7304250559284116,
+      "grad_norm": 0.48950231075286865,
+      "learning_rate": 4.124300740252989e-06,
+      "loss": 0.5772,
+      "step": 1547
+    },
+    {
+      "epoch": 1.7315436241610738,
+      "grad_norm": 0.5003415942192078,
+      "learning_rate": 4.123166258875262e-06,
+      "loss": 0.5942,
+      "step": 1548
+    },
+    {
+      "epoch": 1.732662192393736,
+      "grad_norm": 0.5020825266838074,
+      "learning_rate": 4.1220311993602515e-06,
+      "loss": 0.5826,
+      "step": 1549
+    },
+    {
+      "epoch": 1.7337807606263982,
+      "grad_norm": 0.5122042894363403,
+      "learning_rate": 4.120895562112242e-06,
+      "loss": 0.6112,
+      "step": 1550
+    },
+    {
+      "epoch": 1.7348993288590604,
+      "grad_norm": 0.5022247433662415,
+      "learning_rate": 4.119759347535722e-06,
+      "loss": 0.556,
+      "step": 1551
+    },
+    {
+      "epoch": 1.7360178970917226,
+      "grad_norm": 0.5053969025611877,
+      "learning_rate": 4.118622556035387e-06,
+      "loss": 0.6139,
+      "step": 1552
+    },
+    {
+      "epoch": 1.7371364653243848,
+      "grad_norm": 0.5022400617599487,
+      "learning_rate": 4.11748518801614e-06,
+      "loss": 0.5976,
+      "step": 1553
+    },
+    {
+      "epoch": 1.738255033557047,
+      "grad_norm": 0.49503523111343384,
+      "learning_rate": 4.116347243883086e-06,
+      "loss": 0.5742,
+      "step": 1554
+    },
+    {
+      "epoch": 1.7393736017897092,
+      "grad_norm": 0.5141612887382507,
+      "learning_rate": 4.115208724041536e-06,
+      "loss": 0.599,
+      "step": 1555
+    },
+    {
+      "epoch": 1.7404921700223714,
+      "grad_norm": 0.484074205160141,
+      "learning_rate": 4.114069628897006e-06,
+      "loss": 0.5675,
+      "step": 1556
+    },
+    {
+      "epoch": 1.7416107382550337,
+      "grad_norm": 0.5334796905517578,
+      "learning_rate": 4.11292995885522e-06,
+      "loss": 0.6095,
+      "step": 1557
+    },
+    {
+      "epoch": 1.7427293064876959,
+      "grad_norm": 0.48460689187049866,
+      "learning_rate": 4.111789714322101e-06,
+      "loss": 0.5477,
+      "step": 1558
+    },
+    {
+      "epoch": 1.743847874720358,
+      "grad_norm": 0.5085021257400513,
+      "learning_rate": 4.110648895703782e-06,
+      "loss": 0.5941,
+      "step": 1559
+    },
+    {
+      "epoch": 1.7449664429530203,
+      "grad_norm": 0.5090224146842957,
+      "learning_rate": 4.109507503406599e-06,
+      "loss": 0.5753,
+      "step": 1560
+    },
+    {
+      "epoch": 1.7460850111856825,
+      "grad_norm": 0.5055748224258423,
+      "learning_rate": 4.108365537837088e-06,
+      "loss": 0.5914,
+      "step": 1561
+    },
+    {
+      "epoch": 1.7472035794183445,
+      "grad_norm": 0.48406150937080383,
+      "learning_rate": 4.107222999401997e-06,
+      "loss": 0.5805,
+      "step": 1562
+    },
+    {
+      "epoch": 1.7483221476510067,
+      "grad_norm": 0.48651930689811707,
+      "learning_rate": 4.106079888508272e-06,
+      "loss": 0.5848,
+      "step": 1563
+    },
+    {
+      "epoch": 1.749440715883669,
+      "grad_norm": 0.4852994978427887,
+      "learning_rate": 4.104936205563064e-06,
+      "loss": 0.5796,
+      "step": 1564
+    },
+    {
+      "epoch": 1.750559284116331,
+      "grad_norm": 0.4920569658279419,
+      "learning_rate": 4.10379195097373e-06,
+      "loss": 0.5981,
+      "step": 1565
+    },
+    {
+      "epoch": 1.7516778523489933,
+      "grad_norm": 0.5077376961708069,
+      "learning_rate": 4.1026471251478285e-06,
+      "loss": 0.6071,
+      "step": 1566
+    },
+    {
+      "epoch": 1.7527964205816555,
+      "grad_norm": 0.4932427406311035,
+      "learning_rate": 4.101501728493121e-06,
+      "loss": 0.5917,
+      "step": 1567
+    },
+    {
+      "epoch": 1.7539149888143175,
+      "grad_norm": 0.49199509620666504,
+      "learning_rate": 4.100355761417577e-06,
+      "loss": 0.5814,
+      "step": 1568
+    },
+    {
+      "epoch": 1.7550335570469797,
+      "grad_norm": 0.4920978546142578,
+      "learning_rate": 4.099209224329361e-06,
+      "loss": 0.5931,
+      "step": 1569
+    },
+    {
+      "epoch": 1.756152125279642,
+      "grad_norm": 0.5020889043807983,
+      "learning_rate": 4.098062117636849e-06,
+      "loss": 0.5669,
+      "step": 1570
+    },
+    {
+      "epoch": 1.7572706935123041,
+      "grad_norm": 0.5099507570266724,
+      "learning_rate": 4.096914441748613e-06,
+      "loss": 0.5758,
+      "step": 1571
+    },
+    {
+      "epoch": 1.7583892617449663,
+      "grad_norm": 0.49736326932907104,
+      "learning_rate": 4.095766197073433e-06,
+      "loss": 0.5624,
+      "step": 1572
+    },
+    {
+      "epoch": 1.7595078299776286,
+      "grad_norm": 0.494096577167511,
+      "learning_rate": 4.094617384020287e-06,
+      "loss": 0.5902,
+      "step": 1573
+    },
+    {
+      "epoch": 1.7606263982102908,
+      "grad_norm": 0.5103378295898438,
+      "learning_rate": 4.09346800299836e-06,
+      "loss": 0.5903,
+      "step": 1574
+    },
+    {
+      "epoch": 1.761744966442953,
+      "grad_norm": 0.49033430218696594,
+      "learning_rate": 4.092318054417036e-06,
+      "loss": 0.5703,
+      "step": 1575
+    },
+    {
+      "epoch": 1.7628635346756152,
+      "grad_norm": 0.5065776705741882,
+      "learning_rate": 4.0911675386859015e-06,
+      "loss": 0.5648,
+      "step": 1576
+    },
+    {
+      "epoch": 1.7639821029082774,
+      "grad_norm": 0.5001108050346375,
+      "learning_rate": 4.0900164562147485e-06,
+      "loss": 0.5857,
+      "step": 1577
+    },
+    {
+      "epoch": 1.7651006711409396,
+      "grad_norm": 0.5082598924636841,
+      "learning_rate": 4.0888648074135645e-06,
+      "loss": 0.5488,
+      "step": 1578
+    },
+    {
+      "epoch": 1.7662192393736018,
+      "grad_norm": 0.4972035586833954,
+      "learning_rate": 4.087712592692544e-06,
+      "loss": 0.594,
+      "step": 1579
+    },
+    {
+      "epoch": 1.767337807606264,
+      "grad_norm": 0.4995720088481903,
+      "learning_rate": 4.086559812462082e-06,
+      "loss": 0.5974,
+      "step": 1580
+    },
+    {
+      "epoch": 1.7684563758389262,
+      "grad_norm": 0.5213293433189392,
+      "learning_rate": 4.085406467132774e-06,
+      "loss": 0.5956,
+      "step": 1581
+    },
+    {
+      "epoch": 1.7695749440715884,
+      "grad_norm": 0.4914456903934479,
+      "learning_rate": 4.0842525571154165e-06,
+      "loss": 0.5804,
+      "step": 1582
+    },
+    {
+      "epoch": 1.7706935123042506,
+      "grad_norm": 0.4833807945251465,
+      "learning_rate": 4.083098082821007e-06,
+      "loss": 0.5468,
+      "step": 1583
+    },
+    {
+      "epoch": 1.7718120805369129,
+      "grad_norm": 0.4896129071712494,
+      "learning_rate": 4.081943044660746e-06,
+      "loss": 0.5949,
+      "step": 1584
+    },
+    {
+      "epoch": 1.772930648769575,
+      "grad_norm": 0.4938727021217346,
+      "learning_rate": 4.080787443046034e-06,
+      "loss": 0.576,
+      "step": 1585
+    },
+    {
+      "epoch": 1.7740492170022373,
+      "grad_norm": 0.49900153279304504,
+      "learning_rate": 4.07963127838847e-06,
+      "loss": 0.5788,
+      "step": 1586
+    },
+    {
+      "epoch": 1.7751677852348995,
+      "grad_norm": 0.5096091628074646,
+      "learning_rate": 4.0784745510998556e-06,
+      "loss": 0.6012,
+      "step": 1587
+    },
+    {
+      "epoch": 1.7762863534675615,
+      "grad_norm": 0.4790339469909668,
+      "learning_rate": 4.077317261592194e-06,
+      "loss": 0.572,
+      "step": 1588
+    },
+    {
+      "epoch": 1.7774049217002237,
+      "grad_norm": 0.49272850155830383,
+      "learning_rate": 4.076159410277685e-06,
+      "loss": 0.559,
+      "step": 1589
+    },
+    {
+      "epoch": 1.778523489932886,
+      "grad_norm": 0.46964558959007263,
+      "learning_rate": 4.075000997568732e-06,
+      "loss": 0.5498,
+      "step": 1590
+    },
+    {
+      "epoch": 1.779642058165548,
+      "grad_norm": 0.4938308298587799,
+      "learning_rate": 4.0738420238779365e-06,
+      "loss": 0.5915,
+      "step": 1591
+    },
+    {
+      "epoch": 1.7807606263982103,
+      "grad_norm": 0.5084967017173767,
+      "learning_rate": 4.072682489618101e-06,
+      "loss": 0.5918,
+      "step": 1592
+    },
+    {
+      "epoch": 1.7818791946308725,
+      "grad_norm": 0.49874448776245117,
+      "learning_rate": 4.071522395202226e-06,
+      "loss": 0.5622,
+      "step": 1593
+    },
+    {
+      "epoch": 1.7829977628635347,
+      "grad_norm": 0.49470290541648865,
+      "learning_rate": 4.070361741043511e-06,
+      "loss": 0.5466,
+      "step": 1594
+    },
+    {
+      "epoch": 1.7841163310961967,
+      "grad_norm": 0.5095946192741394,
+      "learning_rate": 4.0692005275553595e-06,
+      "loss": 0.5731,
+      "step": 1595
+    },
+    {
+      "epoch": 1.785234899328859,
+      "grad_norm": 0.5301234722137451,
+      "learning_rate": 4.06803875515137e-06,
+      "loss": 0.5867,
+      "step": 1596
+    },
+    {
+      "epoch": 1.7863534675615211,
+      "grad_norm": 0.514214813709259,
+      "learning_rate": 4.06687642424534e-06,
+      "loss": 0.6005,
+      "step": 1597
+    },
+    {
+      "epoch": 1.7874720357941833,
+      "grad_norm": 0.5104812383651733,
+      "learning_rate": 4.065713535251268e-06,
+      "loss": 0.5935,
+      "step": 1598
+    },
+    {
+      "epoch": 1.7885906040268456,
+      "grad_norm": 0.5091798901557922,
+      "learning_rate": 4.064550088583349e-06,
+      "loss": 0.5944,
+      "step": 1599
+    },
+    {
+      "epoch": 1.7897091722595078,
+      "grad_norm": 0.4886806607246399,
+      "learning_rate": 4.0633860846559794e-06,
+      "loss": 0.551,
+      "step": 1600
+    },
+    {
+      "epoch": 1.79082774049217,
+      "grad_norm": 0.49812769889831543,
+      "learning_rate": 4.062221523883753e-06,
+      "loss": 0.5491,
+      "step": 1601
+    },
+    {
+      "epoch": 1.7919463087248322,
+      "grad_norm": 0.5070154070854187,
+      "learning_rate": 4.06105640668146e-06,
+      "loss": 0.5942,
+      "step": 1602
+    },
+    {
+      "epoch": 1.7930648769574944,
+      "grad_norm": 0.5108977556228638,
+      "learning_rate": 4.059890733464091e-06,
+      "loss": 0.5943,
+      "step": 1603
+    },
+    {
+      "epoch": 1.7941834451901566,
+      "grad_norm": 0.5012465119361877,
+      "learning_rate": 4.058724504646834e-06,
+      "loss": 0.5642,
+      "step": 1604
+    },
+    {
+      "epoch": 1.7953020134228188,
+      "grad_norm": 0.49801158905029297,
+      "learning_rate": 4.057557720645076e-06,
+      "loss": 0.5709,
+      "step": 1605
+    },
+    {
+      "epoch": 1.796420581655481,
+      "grad_norm": 0.4962095320224762,
+      "learning_rate": 4.0563903818743984e-06,
+      "loss": 0.5965,
+      "step": 1606
+    },
+    {
+      "epoch": 1.7975391498881432,
+      "grad_norm": 0.4680939018726349,
+      "learning_rate": 4.055222488750584e-06,
+      "loss": 0.5476,
+      "step": 1607
+    },
+    {
+      "epoch": 1.7986577181208054,
+      "grad_norm": 0.48658618330955505,
+      "learning_rate": 4.054054041689612e-06,
+      "loss": 0.5959,
+      "step": 1608
+    },
+    {
+      "epoch": 1.7997762863534676,
+      "grad_norm": 0.4988051652908325,
+      "learning_rate": 4.052885041107656e-06,
+      "loss": 0.5821,
+      "step": 1609
+    },
+    {
+      "epoch": 1.8008948545861299,
+      "grad_norm": 0.49107739329338074,
+      "learning_rate": 4.051715487421092e-06,
+      "loss": 0.6017,
+      "step": 1610
+    },
+    {
+      "epoch": 1.802013422818792,
+      "grad_norm": 0.5018040537834167,
+      "learning_rate": 4.050545381046488e-06,
+      "loss": 0.5873,
+      "step": 1611
+    },
+    {
+      "epoch": 1.8031319910514543,
+      "grad_norm": 0.49129587411880493,
+      "learning_rate": 4.049374722400613e-06,
+      "loss": 0.5474,
+      "step": 1612
+    },
+    {
+      "epoch": 1.8042505592841165,
+      "grad_norm": 0.49445709586143494,
+      "learning_rate": 4.048203511900428e-06,
+      "loss": 0.5787,
+      "step": 1613
+    },
+    {
+      "epoch": 1.8053691275167785,
+      "grad_norm": 0.4987739026546478,
+      "learning_rate": 4.047031749963095e-06,
+      "loss": 0.5673,
+      "step": 1614
+    },
+    {
+      "epoch": 1.8064876957494407,
+      "grad_norm": 0.4809098243713379,
+      "learning_rate": 4.045859437005971e-06,
+      "loss": 0.5701,
+      "step": 1615
+    },
+    {
+      "epoch": 1.807606263982103,
+      "grad_norm": 0.5122459530830383,
+      "learning_rate": 4.044686573446608e-06,
+      "loss": 0.5996,
+      "step": 1616
+    },
+    {
+      "epoch": 1.808724832214765,
+      "grad_norm": 0.49991941452026367,
+      "learning_rate": 4.0435131597027564e-06,
+      "loss": 0.5981,
+      "step": 1617
+    },
+    {
+      "epoch": 1.8098434004474273,
+      "grad_norm": 0.4867013990879059,
+      "learning_rate": 4.042339196192358e-06,
+      "loss": 0.5762,
+      "step": 1618
+    },
+    {
+      "epoch": 1.8109619686800895,
+      "grad_norm": 0.4812503159046173,
+      "learning_rate": 4.041164683333558e-06,
+      "loss": 0.5484,
+      "step": 1619
+    },
+    {
+      "epoch": 1.8120805369127517,
+      "grad_norm": 0.4890724718570709,
+      "learning_rate": 4.03998962154469e-06,
+      "loss": 0.6001,
+      "step": 1620
+    },
+    {
+      "epoch": 1.8131991051454137,
+      "grad_norm": 0.5026843547821045,
+      "learning_rate": 4.038814011244286e-06,
+      "loss": 0.5838,
+      "step": 1621
+    },
+    {
+      "epoch": 1.814317673378076,
+      "grad_norm": 0.49154821038246155,
+      "learning_rate": 4.037637852851075e-06,
+      "loss": 0.5607,
+      "step": 1622
+    },
+    {
+      "epoch": 1.8154362416107381,
+      "grad_norm": 0.4974023699760437,
+      "learning_rate": 4.036461146783979e-06,
+      "loss": 0.5528,
+      "step": 1623
+    },
+    {
+      "epoch": 1.8165548098434003,
+      "grad_norm": 0.5217268466949463,
+      "learning_rate": 4.035283893462114e-06,
+      "loss": 0.5978,
+      "step": 1624
+    },
+    {
+      "epoch": 1.8176733780760626,
+      "grad_norm": 0.5096696019172668,
+      "learning_rate": 4.034106093304795e-06,
+      "loss": 0.6219,
+      "step": 1625
+    },
+    {
+      "epoch": 1.8187919463087248,
+      "grad_norm": 0.49499067664146423,
+      "learning_rate": 4.032927746731528e-06,
+      "loss": 0.6066,
+      "step": 1626
+    },
+    {
+      "epoch": 1.819910514541387,
+      "grad_norm": 0.48196083307266235,
+      "learning_rate": 4.031748854162014e-06,
+      "loss": 0.562,
+      "step": 1627
+    },
+    {
+      "epoch": 1.8210290827740492,
+      "grad_norm": 0.5044178366661072,
+      "learning_rate": 4.030569416016152e-06,
+      "loss": 0.6103,
+      "step": 1628
+    },
+    {
+      "epoch": 1.8221476510067114,
+      "grad_norm": 0.48558661341667175,
+      "learning_rate": 4.0293894327140315e-06,
+      "loss": 0.5558,
+      "step": 1629
+    },
+    {
+      "epoch": 1.8232662192393736,
+      "grad_norm": 0.492214173078537,
+      "learning_rate": 4.0282089046759365e-06,
+      "loss": 0.5665,
+      "step": 1630
+    },
+    {
+      "epoch": 1.8243847874720358,
+      "grad_norm": 0.4871746599674225,
+      "learning_rate": 4.027027832322348e-06,
+      "loss": 0.5497,
+      "step": 1631
+    },
+    {
+      "epoch": 1.825503355704698,
+      "grad_norm": 0.4865874648094177,
+      "learning_rate": 4.025846216073938e-06,
+      "loss": 0.5558,
+      "step": 1632
+    },
+    {
+      "epoch": 1.8266219239373602,
+      "grad_norm": 0.4919774532318115,
+      "learning_rate": 4.024664056351572e-06,
+      "loss": 0.5804,
+      "step": 1633
+    },
+    {
+      "epoch": 1.8277404921700224,
+      "grad_norm": 0.49305927753448486,
+      "learning_rate": 4.02348135357631e-06,
+      "loss": 0.5869,
+      "step": 1634
+    },
+    {
+      "epoch": 1.8288590604026846,
+      "grad_norm": 0.523074209690094,
+      "learning_rate": 4.022298108169408e-06,
+      "loss": 0.5828,
+      "step": 1635
+    },
+    {
+      "epoch": 1.8299776286353469,
+      "grad_norm": 0.4927634298801422,
+      "learning_rate": 4.021114320552311e-06,
+      "loss": 0.5614,
+      "step": 1636
+    },
+    {
+      "epoch": 1.831096196868009,
+      "grad_norm": 0.49153605103492737,
+      "learning_rate": 4.019929991146659e-06,
+      "loss": 0.5589,
+      "step": 1637
+    },
+    {
+      "epoch": 1.8322147651006713,
+      "grad_norm": 0.5093767046928406,
+      "learning_rate": 4.018745120374286e-06,
+      "loss": 0.5923,
+      "step": 1638
+    },
+    {
+      "epoch": 1.8333333333333335,
+      "grad_norm": 0.5137583017349243,
+      "learning_rate": 4.017559708657216e-06,
+      "loss": 0.6154,
+      "step": 1639
+    },
+    {
+      "epoch": 1.8344519015659957,
+      "grad_norm": 0.5136058926582336,
+      "learning_rate": 4.016373756417669e-06,
+      "loss": 0.5634,
+      "step": 1640
+    },
+    {
+      "epoch": 1.8355704697986577,
+      "grad_norm": 0.4985203146934509,
+      "learning_rate": 4.0151872640780554e-06,
+      "loss": 0.5906,
+      "step": 1641
+    },
+    {
+      "epoch": 1.8366890380313199,
+      "grad_norm": 0.4897896349430084,
+      "learning_rate": 4.014000232060978e-06,
+      "loss": 0.5707,
+      "step": 1642
+    },
+    {
+      "epoch": 1.837807606263982,
+      "grad_norm": 0.49876436591148376,
+      "learning_rate": 4.012812660789233e-06,
+      "loss": 0.5998,
+      "step": 1643
+    },
+    {
+      "epoch": 1.8389261744966443,
+      "grad_norm": 0.5061901807785034,
+      "learning_rate": 4.011624550685808e-06,
+      "loss": 0.602,
+      "step": 1644
+    },
+    {
+      "epoch": 1.8400447427293065,
+      "grad_norm": 0.505077600479126,
+      "learning_rate": 4.010435902173883e-06,
+      "loss": 0.5753,
+      "step": 1645
+    },
+    {
+      "epoch": 1.8411633109619687,
+      "grad_norm": 0.4808630049228668,
+      "learning_rate": 4.0092467156768274e-06,
+      "loss": 0.5737,
+      "step": 1646
+    },
+    {
+      "epoch": 1.8422818791946307,
+      "grad_norm": 0.5141550898551941,
+      "learning_rate": 4.008056991618206e-06,
+      "loss": 0.5963,
+      "step": 1647
+    },
+    {
+      "epoch": 1.843400447427293,
+      "grad_norm": 0.4977302551269531,
+      "learning_rate": 4.006866730421773e-06,
+      "loss": 0.5772,
+      "step": 1648
+    },
+    {
+      "epoch": 1.8445190156599551,
+      "grad_norm": 0.48998597264289856,
+      "learning_rate": 4.005675932511474e-06,
+      "loss": 0.5539,
+      "step": 1649
+    },
+    {
+      "epoch": 1.8456375838926173,
+      "grad_norm": 0.5025562644004822,
+      "learning_rate": 4.004484598311445e-06,
+      "loss": 0.5811,
+      "step": 1650
+    },
+    {
+      "epoch": 1.8467561521252795,
+      "grad_norm": 0.49174991250038147,
+      "learning_rate": 4.003292728246015e-06,
+      "loss": 0.5752,
+      "step": 1651
+    },
+    {
+      "epoch": 1.8478747203579418,
+      "grad_norm": 0.49403971433639526,
+      "learning_rate": 4.0021003227397015e-06,
+      "loss": 0.6112,
+      "step": 1652
+    },
+    {
+      "epoch": 1.848993288590604,
+      "grad_norm": 0.5124534368515015,
+      "learning_rate": 4.000907382217215e-06,
+      "loss": 0.5871,
+      "step": 1653
+    },
+    {
+      "epoch": 1.8501118568232662,
+      "grad_norm": 0.5068091750144958,
+      "learning_rate": 3.9997139071034555e-06,
+      "loss": 0.6084,
+      "step": 1654
+    },
+    {
+      "epoch": 1.8512304250559284,
+      "grad_norm": 0.5051630139350891,
+      "learning_rate": 3.998519897823514e-06,
+      "loss": 0.5964,
+      "step": 1655
+    },
+    {
+      "epoch": 1.8523489932885906,
+      "grad_norm": 0.49604538083076477,
+      "learning_rate": 3.997325354802669e-06,
+      "loss": 0.5655,
+      "step": 1656
+    },
+    {
+      "epoch": 1.8534675615212528,
+      "grad_norm": 0.5094951391220093,
+      "learning_rate": 3.996130278466394e-06,
+      "loss": 0.5824,
+      "step": 1657
+    },
+    {
+      "epoch": 1.854586129753915,
+      "grad_norm": 0.4972209930419922,
+      "learning_rate": 3.994934669240347e-06,
+      "loss": 0.5644,
+      "step": 1658
+    },
+    {
+      "epoch": 1.8557046979865772,
+      "grad_norm": 0.49230918288230896,
+      "learning_rate": 3.993738527550382e-06,
+      "loss": 0.5801,
+      "step": 1659
+    },
+    {
+      "epoch": 1.8568232662192394,
+      "grad_norm": 0.5058436989784241,
+      "learning_rate": 3.9925418538225355e-06,
+      "loss": 0.5917,
+      "step": 1660
+    },
+    {
+      "epoch": 1.8579418344519016,
+      "grad_norm": 0.5173357129096985,
+      "learning_rate": 3.9913446484830396e-06,
+      "loss": 0.5809,
+      "step": 1661
+    },
+    {
+      "epoch": 1.8590604026845639,
+      "grad_norm": 0.48803800344467163,
+      "learning_rate": 3.9901469119583125e-06,
+      "loss": 0.5721,
+      "step": 1662
+    },
+    {
+      "epoch": 1.860178970917226,
+      "grad_norm": 0.5094105005264282,
+      "learning_rate": 3.988948644674962e-06,
+      "loss": 0.5742,
+      "step": 1663
+    },
+    {
+      "epoch": 1.8612975391498883,
+      "grad_norm": 0.49270302057266235,
+      "learning_rate": 3.987749847059788e-06,
+      "loss": 0.5864,
+      "step": 1664
+    },
+    {
+      "epoch": 1.8624161073825505,
+      "grad_norm": 0.49852290749549866,
+      "learning_rate": 3.986550519539773e-06,
+      "loss": 0.5875,
+      "step": 1665
+    },
+    {
+      "epoch": 1.8635346756152127,
+      "grad_norm": 0.4975663423538208,
+      "learning_rate": 3.985350662542096e-06,
+      "loss": 0.5649,
+      "step": 1666
+    },
+    {
+      "epoch": 1.8646532438478747,
+      "grad_norm": 0.5100948214530945,
+      "learning_rate": 3.984150276494116e-06,
+      "loss": 0.564,
+      "step": 1667
+    },
+    {
+      "epoch": 1.8657718120805369,
+      "grad_norm": 0.5079052448272705,
+      "learning_rate": 3.982949361823388e-06,
+      "loss": 0.5662,
+      "step": 1668
+    },
+    {
+      "epoch": 1.866890380313199,
+      "grad_norm": 0.49659186601638794,
+      "learning_rate": 3.981747918957653e-06,
+      "loss": 0.55,
+      "step": 1669
+    },
+    {
+      "epoch": 1.8680089485458613,
+      "grad_norm": 0.48802655935287476,
+      "learning_rate": 3.980545948324835e-06,
+      "loss": 0.5676,
+      "step": 1670
+    },
+    {
+      "epoch": 1.8691275167785235,
+      "grad_norm": 0.4884412884712219,
+      "learning_rate": 3.979343450353056e-06,
+      "loss": 0.5771,
+      "step": 1671
+    },
+    {
+      "epoch": 1.8702460850111857,
+      "grad_norm": 0.49414676427841187,
+      "learning_rate": 3.978140425470615e-06,
+      "loss": 0.5606,
+      "step": 1672
+    },
+    {
+      "epoch": 1.8713646532438477,
+      "grad_norm": 0.4894448220729828,
+      "learning_rate": 3.976936874106008e-06,
+      "loss": 0.5428,
+      "step": 1673
+    },
+    {
+      "epoch": 1.87248322147651,
+      "grad_norm": 0.48966702818870544,
+      "learning_rate": 3.9757327966879134e-06,
+      "loss": 0.5601,
+      "step": 1674
+    },
+    {
+      "epoch": 1.8736017897091721,
+      "grad_norm": 0.513053297996521,
+      "learning_rate": 3.974528193645196e-06,
+      "loss": 0.5993,
+      "step": 1675
+    },
+    {
+      "epoch": 1.8747203579418343,
+      "grad_norm": 0.5094168782234192,
+      "learning_rate": 3.973323065406911e-06,
+      "loss": 0.59,
+      "step": 1676
+    },
+    {
+      "epoch": 1.8758389261744965,
+      "grad_norm": 0.513595461845398,
+      "learning_rate": 3.9721174124023e-06,
+      "loss": 0.5842,
+      "step": 1677
+    },
+    {
+      "epoch": 1.8769574944071588,
+      "grad_norm": 0.5007264614105225,
+      "learning_rate": 3.9709112350607904e-06,
+      "loss": 0.5798,
+      "step": 1678
+    },
+    {
+      "epoch": 1.878076062639821,
+      "grad_norm": 0.48793676495552063,
+      "learning_rate": 3.969704533811997e-06,
+      "loss": 0.5724,
+      "step": 1679
+    },
+    {
+      "epoch": 1.8791946308724832,
+      "grad_norm": 0.5160763263702393,
+      "learning_rate": 3.96849730908572e-06,
+      "loss": 0.5812,
+      "step": 1680
+    },
+    {
+      "epoch": 1.8803131991051454,
+      "grad_norm": 0.5015366077423096,
+      "learning_rate": 3.967289561311949e-06,
+      "loss": 0.562,
+      "step": 1681
+    },
+    {
+      "epoch": 1.8814317673378076,
+      "grad_norm": 0.4753899574279785,
+      "learning_rate": 3.9660812909208575e-06,
+      "loss": 0.548,
+      "step": 1682
+    },
+    {
+      "epoch": 1.8825503355704698,
+      "grad_norm": 0.5044482350349426,
+      "learning_rate": 3.964872498342805e-06,
+      "loss": 0.606,
+      "step": 1683
+    },
+    {
+      "epoch": 1.883668903803132,
+      "grad_norm": 0.4928360879421234,
+      "learning_rate": 3.963663184008338e-06,
+      "loss": 0.5852,
+      "step": 1684
+    },
+    {
+      "epoch": 1.8847874720357942,
+      "grad_norm": 0.4930068254470825,
+      "learning_rate": 3.962453348348189e-06,
+      "loss": 0.5818,
+      "step": 1685
+    },
+    {
+      "epoch": 1.8859060402684564,
+      "grad_norm": 0.5112072229385376,
+      "learning_rate": 3.961242991793276e-06,
+      "loss": 0.6012,
+      "step": 1686
+    },
+    {
+      "epoch": 1.8870246085011186,
+      "grad_norm": 0.5157861113548279,
+      "learning_rate": 3.960032114774702e-06,
+      "loss": 0.6044,
+      "step": 1687
+    },
+    {
+      "epoch": 1.8881431767337808,
+      "grad_norm": 0.4951840937137604,
+      "learning_rate": 3.958820717723754e-06,
+      "loss": 0.5944,
+      "step": 1688
+    },
+    {
+      "epoch": 1.889261744966443,
+      "grad_norm": 0.5044351816177368,
+      "learning_rate": 3.957608801071907e-06,
+      "loss": 0.5814,
+      "step": 1689
+    },
+    {
+      "epoch": 1.8903803131991053,
+      "grad_norm": 0.498749315738678,
+      "learning_rate": 3.956396365250821e-06,
+      "loss": 0.5821,
+      "step": 1690
+    },
+    {
+      "epoch": 1.8914988814317675,
+      "grad_norm": 0.5246557593345642,
+      "learning_rate": 3.955183410692338e-06,
+      "loss": 0.5937,
+      "step": 1691
+    },
+    {
+      "epoch": 1.8926174496644297,
+      "grad_norm": 0.49238988757133484,
+      "learning_rate": 3.953969937828489e-06,
+      "loss": 0.5567,
+      "step": 1692
+    },
+    {
+      "epoch": 1.8937360178970917,
+      "grad_norm": 0.49119988083839417,
+      "learning_rate": 3.952755947091485e-06,
+      "loss": 0.553,
+      "step": 1693
+    },
+    {
+      "epoch": 1.8948545861297539,
+      "grad_norm": 0.5058138370513916,
+      "learning_rate": 3.951541438913723e-06,
+      "loss": 0.5559,
+      "step": 1694
+    },
+    {
+      "epoch": 1.895973154362416,
+      "grad_norm": 0.49434196949005127,
+      "learning_rate": 3.950326413727788e-06,
+      "loss": 0.5723,
+      "step": 1695
+    },
+    {
+      "epoch": 1.8970917225950783,
+      "grad_norm": 0.49539026618003845,
+      "learning_rate": 3.949110871966444e-06,
+      "loss": 0.5426,
+      "step": 1696
+    },
+    {
+      "epoch": 1.8982102908277405,
+      "grad_norm": 0.5126526355743408,
+      "learning_rate": 3.9478948140626414e-06,
+      "loss": 0.5969,
+      "step": 1697
+    },
+    {
+      "epoch": 1.8993288590604027,
+      "grad_norm": 0.5030180811882019,
+      "learning_rate": 3.946678240449515e-06,
+      "loss": 0.5667,
+      "step": 1698
+    },
+    {
+      "epoch": 1.900447427293065,
+      "grad_norm": 0.5048319101333618,
+      "learning_rate": 3.945461151560382e-06,
+      "loss": 0.5755,
+      "step": 1699
+    },
+    {
+      "epoch": 1.901565995525727,
+      "grad_norm": 0.49721550941467285,
+      "learning_rate": 3.944243547828742e-06,
+      "loss": 0.5752,
+      "step": 1700
+    },
+    {
+      "epoch": 1.9026845637583891,
+      "grad_norm": 0.5135868191719055,
+      "learning_rate": 3.943025429688281e-06,
+      "loss": 0.6088,
+      "step": 1701
+    },
+    {
+      "epoch": 1.9038031319910513,
+      "grad_norm": 0.48717236518859863,
+      "learning_rate": 3.941806797572867e-06,
+      "loss": 0.5583,
+      "step": 1702
+    },
+    {
+      "epoch": 1.9049217002237135,
+      "grad_norm": 0.4741756319999695,
+      "learning_rate": 3.940587651916551e-06,
+      "loss": 0.5203,
+      "step": 1703
+    },
+    {
+      "epoch": 1.9060402684563758,
+      "grad_norm": 0.49962809681892395,
+      "learning_rate": 3.939367993153566e-06,
+      "loss": 0.5598,
+      "step": 1704
+    },
+    {
+      "epoch": 1.907158836689038,
+      "grad_norm": 0.48792213201522827,
+      "learning_rate": 3.938147821718328e-06,
+      "loss": 0.5504,
+      "step": 1705
+    },
+    {
+      "epoch": 1.9082774049217002,
+      "grad_norm": 0.5173740386962891,
+      "learning_rate": 3.936927138045438e-06,
+      "loss": 0.5623,
+      "step": 1706
+    },
+    {
+      "epoch": 1.9093959731543624,
+      "grad_norm": 0.5256320834159851,
+      "learning_rate": 3.935705942569675e-06,
+      "loss": 0.5873,
+      "step": 1707
+    },
+    {
+      "epoch": 1.9105145413870246,
+      "grad_norm": 0.5090080499649048,
+      "learning_rate": 3.934484235726006e-06,
+      "loss": 0.6075,
+      "step": 1708
+    },
+    {
+      "epoch": 1.9116331096196868,
+      "grad_norm": 0.4803988039493561,
+      "learning_rate": 3.933262017949574e-06,
+      "loss": 0.5692,
+      "step": 1709
+    },
+    {
+      "epoch": 1.912751677852349,
+      "grad_norm": 0.4928509294986725,
+      "learning_rate": 3.932039289675709e-06,
+      "loss": 0.565,
+      "step": 1710
+    },
+    {
+      "epoch": 1.9138702460850112,
+      "grad_norm": 0.5102552771568298,
+      "learning_rate": 3.930816051339918e-06,
+      "loss": 0.5732,
+      "step": 1711
+    },
+    {
+      "epoch": 1.9149888143176734,
+      "grad_norm": 0.49371102452278137,
+      "learning_rate": 3.929592303377896e-06,
+      "loss": 0.5702,
+      "step": 1712
+    },
+    {
+      "epoch": 1.9161073825503356,
+      "grad_norm": 0.48887091875076294,
+      "learning_rate": 3.928368046225515e-06,
+      "loss": 0.5743,
+      "step": 1713
+    },
+    {
+      "epoch": 1.9172259507829978,
+      "grad_norm": 0.5040315985679626,
+      "learning_rate": 3.927143280318828e-06,
+      "loss": 0.5998,
+      "step": 1714
+    },
+    {
+      "epoch": 1.91834451901566,
+      "grad_norm": 0.5192286372184753,
+      "learning_rate": 3.92591800609407e-06,
+      "loss": 0.5694,
+      "step": 1715
+    },
+    {
+      "epoch": 1.9194630872483223,
+      "grad_norm": 0.5107104778289795,
+      "learning_rate": 3.924692223987661e-06,
+      "loss": 0.6087,
+      "step": 1716
+    },
+    {
+      "epoch": 1.9205816554809845,
+      "grad_norm": 0.5028025507926941,
+      "learning_rate": 3.923465934436195e-06,
+      "loss": 0.5696,
+      "step": 1717
+    },
+    {
+      "epoch": 1.9217002237136467,
+      "grad_norm": 0.5084637403488159,
+      "learning_rate": 3.922239137876452e-06,
+      "loss": 0.6032,
+      "step": 1718
+    },
+    {
+      "epoch": 1.9228187919463087,
+      "grad_norm": 0.5004264116287231,
+      "learning_rate": 3.921011834745392e-06,
+      "loss": 0.5559,
+      "step": 1719
+    },
+    {
+      "epoch": 1.9239373601789709,
+      "grad_norm": 0.522179365158081,
+      "learning_rate": 3.919784025480151e-06,
+      "loss": 0.612,
+      "step": 1720
+    },
+    {
+      "epoch": 1.925055928411633,
+      "grad_norm": 0.508663535118103,
+      "learning_rate": 3.918555710518051e-06,
+      "loss": 0.5731,
+      "step": 1721
+    },
+    {
+      "epoch": 1.9261744966442953,
+      "grad_norm": 0.5001826286315918,
+      "learning_rate": 3.917326890296591e-06,
+      "loss": 0.5906,
+      "step": 1722
+    },
+    {
+      "epoch": 1.9272930648769575,
+      "grad_norm": 0.4933156967163086,
+      "learning_rate": 3.91609756525345e-06,
+      "loss": 0.5816,
+      "step": 1723
+    },
+    {
+      "epoch": 1.9284116331096197,
+      "grad_norm": 0.48916542530059814,
+      "learning_rate": 3.914867735826489e-06,
+      "loss": 0.6014,
+      "step": 1724
+    },
+    {
+      "epoch": 1.929530201342282,
+      "grad_norm": 0.5224503874778748,
+      "learning_rate": 3.913637402453745e-06,
+      "loss": 0.5787,
+      "step": 1725
+    },
+    {
+      "epoch": 1.930648769574944,
+      "grad_norm": 0.4952391982078552,
+      "learning_rate": 3.912406565573438e-06,
+      "loss": 0.5606,
+      "step": 1726
+    },
+    {
+      "epoch": 1.9317673378076061,
+      "grad_norm": 0.5111702680587769,
+      "learning_rate": 3.911175225623965e-06,
+      "loss": 0.5668,
+      "step": 1727
+    },
+    {
+      "epoch": 1.9328859060402683,
+      "grad_norm": 0.4997490346431732,
+      "learning_rate": 3.909943383043904e-06,
+      "loss": 0.5695,
+      "step": 1728
+    },
+    {
+      "epoch": 1.9340044742729305,
+      "grad_norm": 0.5036013126373291,
+      "learning_rate": 3.908711038272011e-06,
+      "loss": 0.5928,
+      "step": 1729
+    },
+    {
+      "epoch": 1.9351230425055927,
+      "grad_norm": 0.5162931084632874,
+      "learning_rate": 3.907478191747221e-06,
+      "loss": 0.5749,
+      "step": 1730
+    },
+    {
+      "epoch": 1.936241610738255,
+      "grad_norm": 0.48660942912101746,
+      "learning_rate": 3.906244843908647e-06,
+      "loss": 0.5539,
+      "step": 1731
+    },
+    {
+      "epoch": 1.9373601789709172,
+      "grad_norm": 0.4941866993904114,
+      "learning_rate": 3.905010995195582e-06,
+      "loss": 0.5854,
+      "step": 1732
+    },
+    {
+      "epoch": 1.9384787472035794,
+      "grad_norm": 0.48834383487701416,
+      "learning_rate": 3.903776646047496e-06,
+      "loss": 0.6049,
+      "step": 1733
+    },
+    {
+      "epoch": 1.9395973154362416,
+      "grad_norm": 0.5004755258560181,
+      "learning_rate": 3.902541796904038e-06,
+      "loss": 0.5544,
+      "step": 1734
+    },
+    {
+      "epoch": 1.9407158836689038,
+      "grad_norm": 0.508080005645752,
+      "learning_rate": 3.901306448205035e-06,
+      "loss": 0.5804,
+      "step": 1735
+    },
+    {
+      "epoch": 1.941834451901566,
+      "grad_norm": 0.49556127190589905,
+      "learning_rate": 3.9000706003904934e-06,
+      "loss": 0.5871,
+      "step": 1736
+    },
+    {
+      "epoch": 1.9429530201342282,
+      "grad_norm": 0.49995389580726624,
+      "learning_rate": 3.898834253900594e-06,
+      "loss": 0.59,
+      "step": 1737
+    },
+    {
+      "epoch": 1.9440715883668904,
+      "grad_norm": 0.4921482503414154,
+      "learning_rate": 3.8975974091756975e-06,
+      "loss": 0.5611,
+      "step": 1738
+    },
+    {
+      "epoch": 1.9451901565995526,
+      "grad_norm": 0.5095739364624023,
+      "learning_rate": 3.896360066656342e-06,
+      "loss": 0.5636,
+      "step": 1739
+    },
+    {
+      "epoch": 1.9463087248322148,
+      "grad_norm": 0.4894288182258606,
+      "learning_rate": 3.895122226783243e-06,
+      "loss": 0.5896,
+      "step": 1740
+    },
+    {
+      "epoch": 1.947427293064877,
+      "grad_norm": 0.516846776008606,
+      "learning_rate": 3.893883889997292e-06,
+      "loss": 0.5938,
+      "step": 1741
+    },
+    {
+      "epoch": 1.9485458612975393,
+      "grad_norm": 0.4975781738758087,
+      "learning_rate": 3.892645056739559e-06,
+      "loss": 0.5635,
+      "step": 1742
+    },
+    {
+      "epoch": 1.9496644295302015,
+      "grad_norm": 0.49723535776138306,
+      "learning_rate": 3.891405727451289e-06,
+      "loss": 0.588,
+      "step": 1743
+    },
+    {
+      "epoch": 1.9507829977628637,
+      "grad_norm": 0.5051411390304565,
+      "learning_rate": 3.8901659025739044e-06,
+      "loss": 0.5574,
+      "step": 1744
+    },
+    {
+      "epoch": 1.951901565995526,
+      "grad_norm": 0.4824368953704834,
+      "learning_rate": 3.888925582549006e-06,
+      "loss": 0.5679,
+      "step": 1745
+    },
+    {
+      "epoch": 1.9530201342281879,
+      "grad_norm": 0.49928563833236694,
+      "learning_rate": 3.887684767818369e-06,
+      "loss": 0.5664,
+      "step": 1746
+    },
+    {
+      "epoch": 1.95413870246085,
+      "grad_norm": 0.5037205219268799,
+      "learning_rate": 3.886443458823944e-06,
+      "loss": 0.5685,
+      "step": 1747
+    },
+    {
+      "epoch": 1.9552572706935123,
+      "grad_norm": 0.49731746315956116,
+      "learning_rate": 3.88520165600786e-06,
+      "loss": 0.563,
+      "step": 1748
+    },
+    {
+      "epoch": 1.9563758389261745,
+      "grad_norm": 0.5214453935623169,
+      "learning_rate": 3.883959359812421e-06,
+      "loss": 0.5882,
+      "step": 1749
+    },
+    {
+      "epoch": 1.9574944071588367,
+      "grad_norm": 0.49750804901123047,
+      "learning_rate": 3.8827165706801055e-06,
+      "loss": 0.5872,
+      "step": 1750
+    },
+    {
+      "epoch": 1.958612975391499,
+      "grad_norm": 0.5092785358428955,
+      "learning_rate": 3.881473289053569e-06,
+      "loss": 0.5948,
+      "step": 1751
+    },
+    {
+      "epoch": 1.959731543624161,
+      "grad_norm": 0.5069229602813721,
+      "learning_rate": 3.880229515375642e-06,
+      "loss": 0.5836,
+      "step": 1752
+    },
+    {
+      "epoch": 1.9608501118568231,
+      "grad_norm": 0.48659929633140564,
+      "learning_rate": 3.878985250089331e-06,
+      "loss": 0.5723,
+      "step": 1753
+    },
+    {
+      "epoch": 1.9619686800894853,
+      "grad_norm": 0.4850320518016815,
+      "learning_rate": 3.8777404936378145e-06,
+      "loss": 0.567,
+      "step": 1754
+    },
+    {
+      "epoch": 1.9630872483221475,
+      "grad_norm": 0.49184632301330566,
+      "learning_rate": 3.87649524646445e-06,
+      "loss": 0.5594,
+      "step": 1755
+    },
+    {
+      "epoch": 1.9642058165548097,
+      "grad_norm": 0.5033354163169861,
+      "learning_rate": 3.875249509012769e-06,
+      "loss": 0.5774,
+      "step": 1756
+    },
+    {
+      "epoch": 1.965324384787472,
+      "grad_norm": 0.4867788851261139,
+      "learning_rate": 3.874003281726474e-06,
+      "loss": 0.5656,
+      "step": 1757
+    },
+    {
+      "epoch": 1.9664429530201342,
+      "grad_norm": 0.5024017095565796,
+      "learning_rate": 3.872756565049447e-06,
+      "loss": 0.542,
+      "step": 1758
+    },
+    {
+      "epoch": 1.9675615212527964,
+      "grad_norm": 0.4939708709716797,
+      "learning_rate": 3.871509359425741e-06,
+      "loss": 0.5554,
+      "step": 1759
+    },
+    {
+      "epoch": 1.9686800894854586,
+      "grad_norm": 0.48120421171188354,
+      "learning_rate": 3.870261665299583e-06,
+      "loss": 0.5724,
+      "step": 1760
+    },
+    {
+      "epoch": 1.9697986577181208,
+      "grad_norm": 0.49949952960014343,
+      "learning_rate": 3.869013483115378e-06,
+      "loss": 0.5732,
+      "step": 1761
+    },
+    {
+      "epoch": 1.970917225950783,
+      "grad_norm": 0.5079313516616821,
+      "learning_rate": 3.867764813317699e-06,
+      "loss": 0.5969,
+      "step": 1762
+    },
+    {
+      "epoch": 1.9720357941834452,
+      "grad_norm": 0.5027442574501038,
+      "learning_rate": 3.866515656351297e-06,
+      "loss": 0.5742,
+      "step": 1763
+    },
+    {
+      "epoch": 1.9731543624161074,
+      "grad_norm": 0.4810044765472412,
+      "learning_rate": 3.865266012661095e-06,
+      "loss": 0.5704,
+      "step": 1764
+    },
+    {
+      "epoch": 1.9742729306487696,
+      "grad_norm": 0.49268779158592224,
+      "learning_rate": 3.86401588269219e-06,
+      "loss": 0.5459,
+      "step": 1765
+    },
+    {
+      "epoch": 1.9753914988814318,
+      "grad_norm": 0.4957699179649353,
+      "learning_rate": 3.86276526688985e-06,
+      "loss": 0.5722,
+      "step": 1766
+    },
+    {
+      "epoch": 1.976510067114094,
+      "grad_norm": 0.4951817989349365,
+      "learning_rate": 3.8615141656995194e-06,
+      "loss": 0.5564,
+      "step": 1767
+    },
+    {
+      "epoch": 1.9776286353467563,
+      "grad_norm": 0.5150087475776672,
+      "learning_rate": 3.860262579566813e-06,
+      "loss": 0.5721,
+      "step": 1768
+    },
+    {
+      "epoch": 1.9787472035794185,
+      "grad_norm": 0.5071924328804016,
+      "learning_rate": 3.859010508937521e-06,
+      "loss": 0.5685,
+      "step": 1769
+    },
+    {
+      "epoch": 1.9798657718120807,
+      "grad_norm": 0.5006820559501648,
+      "learning_rate": 3.857757954257601e-06,
+      "loss": 0.5701,
+      "step": 1770
+    },
+    {
+      "epoch": 1.9809843400447429,
+      "grad_norm": 0.5228837132453918,
+      "learning_rate": 3.856504915973188e-06,
+      "loss": 0.5691,
+      "step": 1771
+    },
+    {
+      "epoch": 1.9821029082774049,
+      "grad_norm": 0.4966356158256531,
+      "learning_rate": 3.855251394530589e-06,
+      "loss": 0.5598,
+      "step": 1772
+    },
+    {
+      "epoch": 1.983221476510067,
+      "grad_norm": 0.5056618452072144,
+      "learning_rate": 3.85399739037628e-06,
+      "loss": 0.5656,
+      "step": 1773
+    },
+    {
+      "epoch": 1.9843400447427293,
+      "grad_norm": 0.49823620915412903,
+      "learning_rate": 3.85274290395691e-06,
+      "loss": 0.5685,
+      "step": 1774
+    },
+    {
+      "epoch": 1.9854586129753915,
+      "grad_norm": 0.5027350187301636,
+      "learning_rate": 3.851487935719302e-06,
+      "loss": 0.5697,
+      "step": 1775
+    },
+    {
+      "epoch": 1.9865771812080537,
+      "grad_norm": 0.5098000764846802,
+      "learning_rate": 3.8502324861104475e-06,
+      "loss": 0.5658,
+      "step": 1776
+    },
+    {
+      "epoch": 1.987695749440716,
+      "grad_norm": 0.5115095376968384,
+      "learning_rate": 3.848976555577513e-06,
+      "loss": 0.5695,
+      "step": 1777
+    },
+    {
+      "epoch": 1.9888143176733781,
+      "grad_norm": 0.485042005777359,
+      "learning_rate": 3.847720144567832e-06,
+      "loss": 0.5505,
+      "step": 1778
+    },
+    {
+      "epoch": 1.9899328859060401,
+      "grad_norm": 0.5096463561058044,
+      "learning_rate": 3.846463253528913e-06,
+      "loss": 0.576,
+      "step": 1779
+    },
+    {
+      "epoch": 1.9910514541387023,
+      "grad_norm": 0.508958101272583,
+      "learning_rate": 3.845205882908432e-06,
+      "loss": 0.58,
+      "step": 1780
+    },
+    {
+      "epoch": 1.9921700223713645,
+      "grad_norm": 0.5001410245895386,
+      "learning_rate": 3.8439480331542394e-06,
+      "loss": 0.5778,
+      "step": 1781
+    },
+    {
+      "epoch": 1.9932885906040267,
+      "grad_norm": 0.49567386507987976,
+      "learning_rate": 3.842689704714354e-06,
+      "loss": 0.5688,
+      "step": 1782
+    },
+    {
+      "epoch": 1.994407158836689,
+      "grad_norm": 0.5215677618980408,
+      "learning_rate": 3.841430898036966e-06,
+      "loss": 0.5879,
+      "step": 1783
+    },
+    {
+      "epoch": 1.9955257270693512,
+      "grad_norm": 0.5176389217376709,
+      "learning_rate": 3.840171613570435e-06,
+      "loss": 0.5788,
+      "step": 1784
+    },
+    {
+      "epoch": 1.9966442953020134,
+      "grad_norm": 0.5101540088653564,
+      "learning_rate": 3.838911851763291e-06,
+      "loss": 0.563,
+      "step": 1785
+    },
+    {
+      "epoch": 1.9977628635346756,
+      "grad_norm": 0.5037233233451843,
+      "learning_rate": 3.837651613064234e-06,
+      "loss": 0.582,
+      "step": 1786
+    },
+    {
+      "epoch": 1.9988814317673378,
+      "grad_norm": 0.5072547197341919,
+      "learning_rate": 3.836390897922136e-06,
+      "loss": 0.5856,
+      "step": 1787
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.49261927604675293,
+      "learning_rate": 3.835129706786036e-06,
+      "loss": 0.5879,
+      "step": 1788
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 5364,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 6,
+  "save_steps": 894,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.005375691167105e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}