diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,17647 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.0,
+  "eval_steps": 500,
+  "global_step": 2515,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0019880715705765406,
+      "grad_norm": 5.719694352988623,
+      "learning_rate": 3.174603174603175e-07,
+      "loss": 1.0761,
+      "step": 1
+    },
+    {
+      "epoch": 0.003976143141153081,
+      "grad_norm": 5.753446360073084,
+      "learning_rate": 6.34920634920635e-07,
+      "loss": 1.0759,
+      "step": 2
+    },
+    {
+      "epoch": 0.005964214711729622,
+      "grad_norm": 5.6362180717832855,
+      "learning_rate": 9.523809523809525e-07,
+      "loss": 1.0634,
+      "step": 3
+    },
+    {
+      "epoch": 0.007952286282306162,
+      "grad_norm": 5.735606400955685,
+      "learning_rate": 1.26984126984127e-06,
+      "loss": 1.079,
+      "step": 4
+    },
+    {
+      "epoch": 0.009940357852882704,
+      "grad_norm": 5.575817061249966,
+      "learning_rate": 1.5873015873015873e-06,
+      "loss": 1.0746,
+      "step": 5
+    },
+    {
+      "epoch": 0.011928429423459244,
+      "grad_norm": 5.260360155484042,
+      "learning_rate": 1.904761904761905e-06,
+      "loss": 1.0615,
+      "step": 6
+    },
+    {
+      "epoch": 0.013916500994035786,
+      "grad_norm": 4.183791741925817,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 1.023,
+      "step": 7
+    },
+    {
+      "epoch": 0.015904572564612324,
+      "grad_norm": 3.890604251583998,
+      "learning_rate": 2.53968253968254e-06,
+      "loss": 1.0225,
+      "step": 8
+    },
+    {
+      "epoch": 0.017892644135188866,
+      "grad_norm": 2.4418688472172203,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 0.9791,
+      "step": 9
+    },
+    {
+      "epoch": 0.019880715705765408,
+      "grad_norm": 2.3267529009813304,
+      "learning_rate": 3.1746031746031746e-06,
+      "loss": 0.9752,
+      "step": 10
+    },
+    {
+      "epoch": 0.02186878727634195,
+      "grad_norm": 2.249898884427966,
+      "learning_rate": 3.492063492063492e-06,
+      "loss": 0.9821,
+      "step": 11
+    },
+    {
+      "epoch": 0.02385685884691849,
+      "grad_norm": 3.177129872043783,
+      "learning_rate": 3.80952380952381e-06,
+      "loss": 0.9412,
+      "step": 12
+    },
+    {
+      "epoch": 0.02584493041749503,
+      "grad_norm": 3.418098934813388,
+      "learning_rate": 4.126984126984127e-06,
+      "loss": 0.9534,
+      "step": 13
+    },
+    {
+      "epoch": 0.027833001988071572,
+      "grad_norm": 3.170560383167798,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 0.9408,
+      "step": 14
+    },
+    {
+      "epoch": 0.02982107355864811,
+      "grad_norm": 2.824593276204021,
+      "learning_rate": 4.761904761904762e-06,
+      "loss": 0.9202,
+      "step": 15
+    },
+    {
+      "epoch": 0.03180914512922465,
+      "grad_norm": 1.980942339361093,
+      "learning_rate": 5.07936507936508e-06,
+      "loss": 0.8779,
+      "step": 16
+    },
+    {
+      "epoch": 0.033797216699801194,
+      "grad_norm": 2.0329912524472835,
+      "learning_rate": 5.396825396825397e-06,
+      "loss": 0.8759,
+      "step": 17
+    },
+    {
+      "epoch": 0.03578528827037773,
+      "grad_norm": 1.7783267821238207,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 0.8651,
+      "step": 18
+    },
+    {
+      "epoch": 0.03777335984095427,
+      "grad_norm": 1.2741893396855972,
+      "learning_rate": 6.031746031746032e-06,
+      "loss": 0.8495,
+      "step": 19
+    },
+    {
+      "epoch": 0.039761431411530816,
+      "grad_norm": 0.9557871843897173,
+      "learning_rate": 6.349206349206349e-06,
+      "loss": 0.8366,
+      "step": 20
+    },
+    {
+      "epoch": 0.041749502982107355,
+      "grad_norm": 1.0086159115273798,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.8291,
+      "step": 21
+    },
+    {
+      "epoch": 0.0437375745526839,
+      "grad_norm": 1.1612726770284232,
+      "learning_rate": 6.984126984126984e-06,
+      "loss": 0.8187,
+      "step": 22
+    },
+    {
+      "epoch": 0.04572564612326044,
+      "grad_norm": 1.0570750814322025,
+      "learning_rate": 7.301587301587301e-06,
+      "loss": 0.8107,
+      "step": 23
+    },
+    {
+      "epoch": 0.04771371769383698,
+      "grad_norm": 0.7751796252916353,
+      "learning_rate": 7.61904761904762e-06,
+      "loss": 0.7934,
+      "step": 24
+    },
+    {
+      "epoch": 0.04970178926441352,
+      "grad_norm": 0.7945616282295676,
+      "learning_rate": 7.936507936507936e-06,
+      "loss": 0.7779,
+      "step": 25
+    },
+    {
+      "epoch": 0.05168986083499006,
+      "grad_norm": 0.858160028742513,
+      "learning_rate": 8.253968253968254e-06,
+      "loss": 0.7796,
+      "step": 26
+    },
+    {
+      "epoch": 0.0536779324055666,
+      "grad_norm": 0.7579216923981651,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 0.7735,
+      "step": 27
+    },
+    {
+      "epoch": 0.055666003976143144,
+      "grad_norm": 0.6924307407288729,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 0.766,
+      "step": 28
+    },
+    {
+      "epoch": 0.05765407554671968,
+      "grad_norm": 0.6685114725284176,
+      "learning_rate": 9.206349206349207e-06,
+      "loss": 0.7672,
+      "step": 29
+    },
+    {
+      "epoch": 0.05964214711729622,
+      "grad_norm": 0.6913397934010306,
+      "learning_rate": 9.523809523809525e-06,
+      "loss": 0.7494,
+      "step": 30
+    },
+    {
+      "epoch": 0.061630218687872766,
+      "grad_norm": 0.6134605758081872,
+      "learning_rate": 9.841269841269842e-06,
+      "loss": 0.7552,
+      "step": 31
+    },
+    {
+      "epoch": 0.0636182902584493,
+      "grad_norm": 0.543250098390486,
+      "learning_rate": 1.015873015873016e-05,
+      "loss": 0.7505,
+      "step": 32
+    },
+    {
+      "epoch": 0.06560636182902585,
+      "grad_norm": 0.5141193730679742,
+      "learning_rate": 1.0476190476190477e-05,
+      "loss": 0.7446,
+      "step": 33
+    },
+    {
+      "epoch": 0.06759443339960239,
+      "grad_norm": 0.5115646872354824,
+      "learning_rate": 1.0793650793650794e-05,
+      "loss": 0.745,
+      "step": 34
+    },
+    {
+      "epoch": 0.06958250497017893,
+      "grad_norm": 0.5623419612031815,
+      "learning_rate": 1.1111111111111113e-05,
+      "loss": 0.7342,
+      "step": 35
+    },
+    {
+      "epoch": 0.07157057654075547,
+      "grad_norm": 0.5502558641447046,
+      "learning_rate": 1.1428571428571429e-05,
+      "loss": 0.7322,
+      "step": 36
+    },
+    {
+      "epoch": 0.073558648111332,
+      "grad_norm": 0.5359887902606768,
+      "learning_rate": 1.1746031746031748e-05,
+      "loss": 0.73,
+      "step": 37
+    },
+    {
+      "epoch": 0.07554671968190854,
+      "grad_norm": 0.44251623117375427,
+      "learning_rate": 1.2063492063492064e-05,
+      "loss": 0.7282,
+      "step": 38
+    },
+    {
+      "epoch": 0.0775347912524851,
+      "grad_norm": 0.4507677964518194,
+      "learning_rate": 1.2380952380952383e-05,
+      "loss": 0.7244,
+      "step": 39
+    },
+    {
+      "epoch": 0.07952286282306163,
+      "grad_norm": 0.4512848693711087,
+      "learning_rate": 1.2698412698412699e-05,
+      "loss": 0.7278,
+      "step": 40
+    },
+    {
+      "epoch": 0.08151093439363817,
+      "grad_norm": 0.41505660856918547,
+      "learning_rate": 1.3015873015873018e-05,
+      "loss": 0.7254,
+      "step": 41
+    },
+    {
+      "epoch": 0.08349900596421471,
+      "grad_norm": 0.48813041957576425,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.7194,
+      "step": 42
+    },
+    {
+      "epoch": 0.08548707753479125,
+      "grad_norm": 0.3877838036386138,
+      "learning_rate": 1.3650793650793652e-05,
+      "loss": 0.7176,
+      "step": 43
+    },
+    {
+      "epoch": 0.0874751491053678,
+      "grad_norm": 0.4181924462251027,
+      "learning_rate": 1.3968253968253968e-05,
+      "loss": 0.7156,
+      "step": 44
+    },
+    {
+      "epoch": 0.08946322067594434,
+      "grad_norm": 0.3939861879487352,
+      "learning_rate": 1.4285714285714287e-05,
+      "loss": 0.7056,
+      "step": 45
+    },
+    {
+      "epoch": 0.09145129224652088,
+      "grad_norm": 0.32819440379311476,
+      "learning_rate": 1.4603174603174603e-05,
+      "loss": 0.7053,
+      "step": 46
+    },
+    {
+      "epoch": 0.09343936381709742,
+      "grad_norm": 0.4345116128871954,
+      "learning_rate": 1.4920634920634922e-05,
+      "loss": 0.7097,
+      "step": 47
+    },
+    {
+      "epoch": 0.09542743538767395,
+      "grad_norm": 0.357575137500084,
+      "learning_rate": 1.523809523809524e-05,
+      "loss": 0.7002,
+      "step": 48
+    },
+    {
+      "epoch": 0.09741550695825049,
+      "grad_norm": 0.36960029460147453,
+      "learning_rate": 1.555555555555556e-05,
+      "loss": 0.6974,
+      "step": 49
+    },
+    {
+      "epoch": 0.09940357852882704,
+      "grad_norm": 0.3665609922915068,
+      "learning_rate": 1.5873015873015872e-05,
+      "loss": 0.6999,
+      "step": 50
+    },
+    {
+      "epoch": 0.10139165009940358,
+      "grad_norm": 0.3080867101340467,
+      "learning_rate": 1.6190476190476193e-05,
+      "loss": 0.7007,
+      "step": 51
+    },
+    {
+      "epoch": 0.10337972166998012,
+      "grad_norm": 0.29433385790793876,
+      "learning_rate": 1.6507936507936507e-05,
+      "loss": 0.6958,
+      "step": 52
+    },
+    {
+      "epoch": 0.10536779324055666,
+      "grad_norm": 0.2932180630106322,
+      "learning_rate": 1.6825396825396828e-05,
+      "loss": 0.6963,
+      "step": 53
+    },
+    {
+      "epoch": 0.1073558648111332,
+      "grad_norm": 0.390572352111665,
+      "learning_rate": 1.7142857142857142e-05,
+      "loss": 0.692,
+      "step": 54
+    },
+    {
+      "epoch": 0.10934393638170974,
+      "grad_norm": 0.46019460280581603,
+      "learning_rate": 1.7460317460317463e-05,
+      "loss": 0.6899,
+      "step": 55
+    },
+    {
+      "epoch": 0.11133200795228629,
+      "grad_norm": 0.4456425633626217,
+      "learning_rate": 1.7777777777777777e-05,
+      "loss": 0.6934,
+      "step": 56
+    },
+    {
+      "epoch": 0.11332007952286283,
+      "grad_norm": 0.5779834684485384,
+      "learning_rate": 1.8095238095238097e-05,
+      "loss": 0.7005,
+      "step": 57
+    },
+    {
+      "epoch": 0.11530815109343936,
+      "grad_norm": 0.43345848661699465,
+      "learning_rate": 1.8412698412698415e-05,
+      "loss": 0.688,
+      "step": 58
+    },
+    {
+      "epoch": 0.1172962226640159,
+      "grad_norm": 0.4702542027909562,
+      "learning_rate": 1.8730158730158732e-05,
+      "loss": 0.6838,
+      "step": 59
+    },
+    {
+      "epoch": 0.11928429423459244,
+      "grad_norm": 0.4532557455295433,
+      "learning_rate": 1.904761904761905e-05,
+      "loss": 0.6866,
+      "step": 60
+    },
+    {
+      "epoch": 0.12127236580516898,
+      "grad_norm": 0.6378809717045135,
+      "learning_rate": 1.9365079365079367e-05,
+      "loss": 0.6854,
+      "step": 61
+    },
+    {
+      "epoch": 0.12326043737574553,
+      "grad_norm": 0.8186502076427445,
+      "learning_rate": 1.9682539682539684e-05,
+      "loss": 0.6831,
+      "step": 62
+    },
+    {
+      "epoch": 0.12524850894632206,
+      "grad_norm": 0.9490391751989258,
+      "learning_rate": 2e-05,
+      "loss": 0.6798,
+      "step": 63
+    },
+    {
+      "epoch": 0.1272365805168986,
+      "grad_norm": 0.8318620223308042,
+      "learning_rate": 2.031746031746032e-05,
+      "loss": 0.6842,
+      "step": 64
+    },
+    {
+      "epoch": 0.12922465208747516,
+      "grad_norm": 0.6266730729415922,
+      "learning_rate": 2.063492063492064e-05,
+      "loss": 0.676,
+      "step": 65
+    },
+    {
+      "epoch": 0.1312127236580517,
+      "grad_norm": 0.4366217658351886,
+      "learning_rate": 2.0952380952380954e-05,
+      "loss": 0.6792,
+      "step": 66
+    },
+    {
+      "epoch": 0.13320079522862824,
+      "grad_norm": 0.4533555372067667,
+      "learning_rate": 2.126984126984127e-05,
+      "loss": 0.6797,
+      "step": 67
+    },
+    {
+      "epoch": 0.13518886679920478,
+      "grad_norm": 0.6802950259274722,
+      "learning_rate": 2.158730158730159e-05,
+      "loss": 0.6773,
+      "step": 68
+    },
+    {
+      "epoch": 0.13717693836978131,
+      "grad_norm": 1.1241036512188962,
+      "learning_rate": 2.190476190476191e-05,
+      "loss": 0.6783,
+      "step": 69
+    },
+    {
+      "epoch": 0.13916500994035785,
+      "grad_norm": 1.125722397572469,
+      "learning_rate": 2.2222222222222227e-05,
+      "loss": 0.6715,
+      "step": 70
+    },
+    {
+      "epoch": 0.1411530815109344,
+      "grad_norm": 0.8427474910536563,
+      "learning_rate": 2.253968253968254e-05,
+      "loss": 0.671,
+      "step": 71
+    },
+    {
+      "epoch": 0.14314115308151093,
+      "grad_norm": 1.0095817810923369,
+      "learning_rate": 2.2857142857142858e-05,
+      "loss": 0.673,
+      "step": 72
+    },
+    {
+      "epoch": 0.14512922465208747,
+      "grad_norm": 1.2860444414014913,
+      "learning_rate": 2.317460317460318e-05,
+      "loss": 0.6781,
+      "step": 73
+    },
+    {
+      "epoch": 0.147117296222664,
+      "grad_norm": 0.88467179821398,
+      "learning_rate": 2.3492063492063496e-05,
+      "loss": 0.6701,
+      "step": 74
+    },
+    {
+      "epoch": 0.14910536779324055,
+      "grad_norm": 1.1930360448025386,
+      "learning_rate": 2.380952380952381e-05,
+      "loss": 0.6714,
+      "step": 75
+    },
+    {
+      "epoch": 0.15109343936381708,
+      "grad_norm": 1.2462987345720242,
+      "learning_rate": 2.4126984126984128e-05,
+      "loss": 0.6674,
+      "step": 76
+    },
+    {
+      "epoch": 0.15308151093439365,
+      "grad_norm": 1.0610109095972218,
+      "learning_rate": 2.444444444444445e-05,
+      "loss": 0.6631,
+      "step": 77
+    },
+    {
+      "epoch": 0.1550695825049702,
+      "grad_norm": 1.13945144581175,
+      "learning_rate": 2.4761904761904766e-05,
+      "loss": 0.6686,
+      "step": 78
+    },
+    {
+      "epoch": 0.15705765407554673,
+      "grad_norm": 0.8784096423124421,
+      "learning_rate": 2.507936507936508e-05,
+      "loss": 0.666,
+      "step": 79
+    },
+    {
+      "epoch": 0.15904572564612326,
+      "grad_norm": 1.7453043654997522,
+      "learning_rate": 2.5396825396825397e-05,
+      "loss": 0.6682,
+      "step": 80
+    },
+    {
+      "epoch": 0.1610337972166998,
+      "grad_norm": 0.6565327581012248,
+      "learning_rate": 2.5714285714285718e-05,
+      "loss": 0.6619,
+      "step": 81
+    },
+    {
+      "epoch": 0.16302186878727634,
+      "grad_norm": 1.457355963249708,
+      "learning_rate": 2.6031746031746035e-05,
+      "loss": 0.6618,
+      "step": 82
+    },
+    {
+      "epoch": 0.16500994035785288,
+      "grad_norm": 1.3748440778156223,
+      "learning_rate": 2.6349206349206353e-05,
+      "loss": 0.662,
+      "step": 83
+    },
+    {
+      "epoch": 0.16699801192842942,
+      "grad_norm": 0.6576930469096256,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.6645,
+      "step": 84
+    },
+    {
+      "epoch": 0.16898608349900596,
+      "grad_norm": 1.216101814201686,
+      "learning_rate": 2.6984126984126987e-05,
+      "loss": 0.6632,
+      "step": 85
+    },
+    {
+      "epoch": 0.1709741550695825,
+      "grad_norm": 1.6651254013048873,
+      "learning_rate": 2.7301587301587305e-05,
+      "loss": 0.6608,
+      "step": 86
+    },
+    {
+      "epoch": 0.17296222664015903,
+      "grad_norm": 0.736668310308581,
+      "learning_rate": 2.7619047619047622e-05,
+      "loss": 0.6614,
+      "step": 87
+    },
+    {
+      "epoch": 0.1749502982107356,
+      "grad_norm": 2.657224735711684,
+      "learning_rate": 2.7936507936507936e-05,
+      "loss": 0.6699,
+      "step": 88
+    },
+    {
+      "epoch": 0.17693836978131214,
+      "grad_norm": 1.8281737180328776,
+      "learning_rate": 2.8253968253968257e-05,
+      "loss": 0.6629,
+      "step": 89
+    },
+    {
+      "epoch": 0.17892644135188868,
+      "grad_norm": 2.458633290210547,
+      "learning_rate": 2.8571428571428574e-05,
+      "loss": 0.6602,
+      "step": 90
+    },
+    {
+      "epoch": 0.18091451292246521,
+      "grad_norm": 2.712443155137568,
+      "learning_rate": 2.888888888888889e-05,
+      "loss": 0.6749,
+      "step": 91
+    },
+    {
+      "epoch": 0.18290258449304175,
+      "grad_norm": 0.8261764251352429,
+      "learning_rate": 2.9206349206349206e-05,
+      "loss": 0.6612,
+      "step": 92
+    },
+    {
+      "epoch": 0.1848906560636183,
+      "grad_norm": 2.054547558417178,
+      "learning_rate": 2.9523809523809526e-05,
+      "loss": 0.667,
+      "step": 93
+    },
+    {
+      "epoch": 0.18687872763419483,
+      "grad_norm": 1.236783488406299,
+      "learning_rate": 2.9841269841269844e-05,
+      "loss": 0.6624,
+      "step": 94
+    },
+    {
+      "epoch": 0.18886679920477137,
+      "grad_norm": 2.162948847530382,
+      "learning_rate": 3.015873015873016e-05,
+      "loss": 0.6663,
+      "step": 95
+    },
+    {
+      "epoch": 0.1908548707753479,
+      "grad_norm": 1.930702254563974,
+      "learning_rate": 3.047619047619048e-05,
+      "loss": 0.6827,
+      "step": 96
+    },
+    {
+      "epoch": 0.19284294234592445,
+      "grad_norm": 1.609723044776149,
+      "learning_rate": 3.0793650793650796e-05,
+      "loss": 0.669,
+      "step": 97
+    },
+    {
+      "epoch": 0.19483101391650098,
+      "grad_norm": 1.5177019425537601,
+      "learning_rate": 3.111111111111112e-05,
+      "loss": 0.6689,
+      "step": 98
+    },
+    {
+      "epoch": 0.19681908548707752,
+      "grad_norm": 1.5351591826271913,
+      "learning_rate": 3.142857142857143e-05,
+      "loss": 0.6664,
+      "step": 99
+    },
+    {
+      "epoch": 0.1988071570576541,
+      "grad_norm": 1.0566494701757996,
+      "learning_rate": 3.1746031746031745e-05,
+      "loss": 0.6527,
+      "step": 100
+    },
+    {
+      "epoch": 0.20079522862823063,
+      "grad_norm": 1.4731697755973054,
+      "learning_rate": 3.2063492063492065e-05,
+      "loss": 0.6605,
+      "step": 101
+    },
+    {
+      "epoch": 0.20278330019880716,
+      "grad_norm": 1.1128607790678517,
+      "learning_rate": 3.2380952380952386e-05,
+      "loss": 0.6611,
+      "step": 102
+    },
+    {
+      "epoch": 0.2047713717693837,
+      "grad_norm": 1.478193180756222,
+      "learning_rate": 3.26984126984127e-05,
+      "loss": 0.6566,
+      "step": 103
+    },
+    {
+      "epoch": 0.20675944333996024,
+      "grad_norm": 1.3087065995108067,
+      "learning_rate": 3.3015873015873014e-05,
+      "loss": 0.66,
+      "step": 104
+    },
+    {
+      "epoch": 0.20874751491053678,
+      "grad_norm": 1.4156409974895987,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.6638,
+      "step": 105
+    },
+    {
+      "epoch": 0.21073558648111332,
+      "grad_norm": 1.30062464042937,
+      "learning_rate": 3.3650793650793656e-05,
+      "loss": 0.6514,
+      "step": 106
+    },
+    {
+      "epoch": 0.21272365805168986,
+      "grad_norm": 1.4113886514836769,
+      "learning_rate": 3.396825396825397e-05,
+      "loss": 0.6569,
+      "step": 107
+    },
+    {
+      "epoch": 0.2147117296222664,
+      "grad_norm": 1.0156596270150422,
+      "learning_rate": 3.4285714285714284e-05,
+      "loss": 0.6504,
+      "step": 108
+    },
+    {
+      "epoch": 0.21669980119284293,
+      "grad_norm": 1.5825915741647385,
+      "learning_rate": 3.4603174603174604e-05,
+      "loss": 0.6533,
+      "step": 109
+    },
+    {
+      "epoch": 0.21868787276341947,
+      "grad_norm": 1.1296868604446741,
+      "learning_rate": 3.4920634920634925e-05,
+      "loss": 0.6534,
+      "step": 110
+    },
+    {
+      "epoch": 0.220675944333996,
+      "grad_norm": 1.3923998966758822,
+      "learning_rate": 3.523809523809524e-05,
+      "loss": 0.65,
+      "step": 111
+    },
+    {
+      "epoch": 0.22266401590457258,
+      "grad_norm": 1.2640733985651982,
+      "learning_rate": 3.555555555555555e-05,
+      "loss": 0.6519,
+      "step": 112
+    },
+    {
+      "epoch": 0.22465208747514911,
+      "grad_norm": 1.4802049067534304,
+      "learning_rate": 3.5873015873015874e-05,
+      "loss": 0.6508,
+      "step": 113
+    },
+    {
+      "epoch": 0.22664015904572565,
+      "grad_norm": 1.5035055412914218,
+      "learning_rate": 3.6190476190476195e-05,
+      "loss": 0.6473,
+      "step": 114
+    },
+    {
+      "epoch": 0.2286282306163022,
+      "grad_norm": 1.1937240148726864,
+      "learning_rate": 3.650793650793651e-05,
+      "loss": 0.6512,
+      "step": 115
+    },
+    {
+      "epoch": 0.23061630218687873,
+      "grad_norm": 1.2457856561461578,
+      "learning_rate": 3.682539682539683e-05,
+      "loss": 0.6447,
+      "step": 116
+    },
+    {
+      "epoch": 0.23260437375745527,
+      "grad_norm": 1.3273129343443546,
+      "learning_rate": 3.714285714285715e-05,
+      "loss": 0.6464,
+      "step": 117
+    },
+    {
+      "epoch": 0.2345924453280318,
+      "grad_norm": 1.3281865305863692,
+      "learning_rate": 3.7460317460317464e-05,
+      "loss": 0.6477,
+      "step": 118
+    },
+    {
+      "epoch": 0.23658051689860835,
+      "grad_norm": 1.0056815583069727,
+      "learning_rate": 3.777777777777778e-05,
+      "loss": 0.645,
+      "step": 119
+    },
+    {
+      "epoch": 0.23856858846918488,
+      "grad_norm": 1.382121816169979,
+      "learning_rate": 3.80952380952381e-05,
+      "loss": 0.6546,
+      "step": 120
+    },
+    {
+      "epoch": 0.24055666003976142,
+      "grad_norm": 1.3100694949960303,
+      "learning_rate": 3.841269841269842e-05,
+      "loss": 0.6488,
+      "step": 121
+    },
+    {
+      "epoch": 0.24254473161033796,
+      "grad_norm": 1.4947029004278205,
+      "learning_rate": 3.8730158730158734e-05,
+      "loss": 0.6477,
+      "step": 122
+    },
+    {
+      "epoch": 0.24453280318091453,
+      "grad_norm": 1.4644178209776242,
+      "learning_rate": 3.904761904761905e-05,
+      "loss": 0.652,
+      "step": 123
+    },
+    {
+      "epoch": 0.24652087475149106,
+      "grad_norm": 1.3769771435448332,
+      "learning_rate": 3.936507936507937e-05,
+      "loss": 0.6489,
+      "step": 124
+    },
+    {
+      "epoch": 0.2485089463220676,
+      "grad_norm": 1.0653896387340898,
+      "learning_rate": 3.968253968253969e-05,
+      "loss": 0.6459,
+      "step": 125
+    },
+    {
+      "epoch": 0.2504970178926441,
+      "grad_norm": 1.4120675480758071,
+      "learning_rate": 4e-05,
+      "loss": 0.6521,
+      "step": 126
+    },
+    {
+      "epoch": 0.2524850894632207,
+      "grad_norm": 1.2331141631565885,
+      "learning_rate": 4.031746031746032e-05,
+      "loss": 0.6412,
+      "step": 127
+    },
+    {
+      "epoch": 0.2544731610337972,
+      "grad_norm": 1.4510580599851584,
+      "learning_rate": 4.063492063492064e-05,
+      "loss": 0.6424,
+      "step": 128
+    },
+    {
+      "epoch": 0.25646123260437376,
+      "grad_norm": 1.1641813315566232,
+      "learning_rate": 4.095238095238095e-05,
+      "loss": 0.6456,
+      "step": 129
+    },
+    {
+      "epoch": 0.2584493041749503,
+      "grad_norm": 1.504580390843022,
+      "learning_rate": 4.126984126984128e-05,
+      "loss": 0.6408,
+      "step": 130
+    },
+    {
+      "epoch": 0.26043737574552683,
+      "grad_norm": 1.2257434898909374,
+      "learning_rate": 4.1587301587301594e-05,
+      "loss": 0.6429,
+      "step": 131
+    },
+    {
+      "epoch": 0.2624254473161034,
+      "grad_norm": 0.9912463823975768,
+      "learning_rate": 4.190476190476191e-05,
+      "loss": 0.6431,
+      "step": 132
+    },
+    {
+      "epoch": 0.2644135188866799,
+      "grad_norm": 1.5247568836492176,
+      "learning_rate": 4.222222222222223e-05,
+      "loss": 0.6555,
+      "step": 133
+    },
+    {
+      "epoch": 0.2664015904572565,
+      "grad_norm": 1.5773163450396874,
+      "learning_rate": 4.253968253968254e-05,
+      "loss": 0.6545,
+      "step": 134
+    },
+    {
+      "epoch": 0.268389662027833,
+      "grad_norm": 1.291828731349488,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 0.6429,
+      "step": 135
+    },
+    {
+      "epoch": 0.27037773359840955,
+      "grad_norm": 1.0985522555843004,
+      "learning_rate": 4.317460317460318e-05,
+      "loss": 0.6408,
+      "step": 136
+    },
+    {
+      "epoch": 0.27236580516898606,
+      "grad_norm": 1.5828560475019822,
+      "learning_rate": 4.349206349206349e-05,
+      "loss": 0.6594,
+      "step": 137
+    },
+    {
+      "epoch": 0.27435387673956263,
+      "grad_norm": 1.3407587991159484,
+      "learning_rate": 4.380952380952382e-05,
+      "loss": 0.6382,
+      "step": 138
+    },
+    {
+      "epoch": 0.27634194831013914,
+      "grad_norm": 1.1882787897778562,
+      "learning_rate": 4.412698412698413e-05,
+      "loss": 0.6435,
+      "step": 139
+    },
+    {
+      "epoch": 0.2783300198807157,
+      "grad_norm": 1.4743146172658406,
+      "learning_rate": 4.444444444444445e-05,
+      "loss": 0.6426,
+      "step": 140
+    },
+    {
+      "epoch": 0.2803180914512923,
+      "grad_norm": 1.3198867552344011,
+      "learning_rate": 4.476190476190477e-05,
+      "loss": 0.6418,
+      "step": 141
+    },
+    {
+      "epoch": 0.2823061630218688,
+      "grad_norm": 1.087940001986759,
+      "learning_rate": 4.507936507936508e-05,
+      "loss": 0.6438,
+      "step": 142
+    },
+    {
+      "epoch": 0.28429423459244535,
+      "grad_norm": 1.7460817891906895,
+      "learning_rate": 4.53968253968254e-05,
+      "loss": 0.6438,
+      "step": 143
+    },
+    {
+      "epoch": 0.28628230616302186,
+      "grad_norm": 1.072098243868636,
+      "learning_rate": 4.5714285714285716e-05,
+      "loss": 0.6449,
+      "step": 144
+    },
+    {
+      "epoch": 0.2882703777335984,
+      "grad_norm": 2.1835116598411504,
+      "learning_rate": 4.603174603174603e-05,
+      "loss": 0.6489,
+      "step": 145
+    },
+    {
+      "epoch": 0.29025844930417494,
+      "grad_norm": 1.879705553582977,
+      "learning_rate": 4.634920634920636e-05,
+      "loss": 0.6514,
+      "step": 146
+    },
+    {
+      "epoch": 0.2922465208747515,
+      "grad_norm": 1.4090093524508818,
+      "learning_rate": 4.666666666666667e-05,
+      "loss": 0.6409,
+      "step": 147
+    },
+    {
+      "epoch": 0.294234592445328,
+      "grad_norm": 1.8436589830840873,
+      "learning_rate": 4.698412698412699e-05,
+      "loss": 0.6442,
+      "step": 148
+    },
+    {
+      "epoch": 0.2962226640159046,
+      "grad_norm": 1.0212379094945134,
+      "learning_rate": 4.7301587301587306e-05,
+      "loss": 0.6402,
+      "step": 149
+    },
+    {
+      "epoch": 0.2982107355864811,
+      "grad_norm": 1.7700790575852108,
+      "learning_rate": 4.761904761904762e-05,
+      "loss": 0.6415,
+      "step": 150
+    },
+    {
+      "epoch": 0.30019880715705766,
+      "grad_norm": 1.6067327390517923,
+      "learning_rate": 4.793650793650794e-05,
+      "loss": 0.6438,
+      "step": 151
+    },
+    {
+      "epoch": 0.30218687872763417,
+      "grad_norm": 1.1795271072398732,
+      "learning_rate": 4.8253968253968255e-05,
+      "loss": 0.635,
+      "step": 152
+    },
+    {
+      "epoch": 0.30417495029821073,
+      "grad_norm": 1.5562700844547241,
+      "learning_rate": 4.857142857142857e-05,
+      "loss": 0.6395,
+      "step": 153
+    },
+    {
+      "epoch": 0.3061630218687873,
+      "grad_norm": 1.1852659442882558,
+      "learning_rate": 4.88888888888889e-05,
+      "loss": 0.6353,
+      "step": 154
+    },
+    {
+      "epoch": 0.3081510934393638,
+      "grad_norm": 1.303291076586374,
+      "learning_rate": 4.920634920634921e-05,
+      "loss": 0.639,
+      "step": 155
+    },
+    {
+      "epoch": 0.3101391650099404,
+      "grad_norm": 1.3279480936113928,
+      "learning_rate": 4.952380952380953e-05,
+      "loss": 0.6408,
+      "step": 156
+    },
+    {
+      "epoch": 0.3121272365805169,
+      "grad_norm": 1.2480990916236967,
+      "learning_rate": 4.9841269841269845e-05,
+      "loss": 0.6395,
+      "step": 157
+    },
+    {
+      "epoch": 0.31411530815109345,
+      "grad_norm": 1.7232506056868664,
+      "learning_rate": 5.015873015873016e-05,
+      "loss": 0.6429,
+      "step": 158
+    },
+    {
+      "epoch": 0.31610337972166996,
+      "grad_norm": 1.1165626843230316,
+      "learning_rate": 5.047619047619048e-05,
+      "loss": 0.6377,
+      "step": 159
+    },
+    {
+      "epoch": 0.31809145129224653,
+      "grad_norm": 1.365984266306236,
+      "learning_rate": 5.0793650793650794e-05,
+      "loss": 0.6385,
+      "step": 160
+    },
+    {
+      "epoch": 0.32007952286282304,
+      "grad_norm": 1.161379924883256,
+      "learning_rate": 5.111111111111111e-05,
+      "loss": 0.6342,
+      "step": 161
+    },
+    {
+      "epoch": 0.3220675944333996,
+      "grad_norm": 1.5501649766832735,
+      "learning_rate": 5.1428571428571436e-05,
+      "loss": 0.6397,
+      "step": 162
+    },
+    {
+      "epoch": 0.3240556660039761,
+      "grad_norm": 1.236059822703464,
+      "learning_rate": 5.174603174603175e-05,
+      "loss": 0.6345,
+      "step": 163
+    },
+    {
+      "epoch": 0.3260437375745527,
+      "grad_norm": 1.1632007138023943,
+      "learning_rate": 5.206349206349207e-05,
+      "loss": 0.6331,
+      "step": 164
+    },
+    {
+      "epoch": 0.32803180914512925,
+      "grad_norm": 1.8957350418517902,
+      "learning_rate": 5.2380952380952384e-05,
+      "loss": 0.6424,
+      "step": 165
+    },
+    {
+      "epoch": 0.33001988071570576,
+      "grad_norm": 1.2762565748269343,
+      "learning_rate": 5.2698412698412705e-05,
+      "loss": 0.6388,
+      "step": 166
+    },
+    {
+      "epoch": 0.3320079522862823,
+      "grad_norm": 0.9926513996495598,
+      "learning_rate": 5.301587301587302e-05,
+      "loss": 0.6334,
+      "step": 167
+    },
+    {
+      "epoch": 0.33399602385685884,
+      "grad_norm": 1.5072734890729735,
+      "learning_rate": 5.333333333333333e-05,
+      "loss": 0.6402,
+      "step": 168
+    },
+    {
+      "epoch": 0.3359840954274354,
+      "grad_norm": 1.6848000191372772,
+      "learning_rate": 5.3650793650793654e-05,
+      "loss": 0.6314,
+      "step": 169
+    },
+    {
+      "epoch": 0.3379721669980119,
+      "grad_norm": 1.4016229963298041,
+      "learning_rate": 5.3968253968253975e-05,
+      "loss": 0.6278,
+      "step": 170
+    },
+    {
+      "epoch": 0.3399602385685885,
+      "grad_norm": 1.0923456202776507,
+      "learning_rate": 5.4285714285714295e-05,
+      "loss": 0.6297,
+      "step": 171
+    },
+    {
+      "epoch": 0.341948310139165,
+      "grad_norm": 2.5232602422771997,
+      "learning_rate": 5.460317460317461e-05,
+      "loss": 0.6382,
+      "step": 172
+    },
+    {
+      "epoch": 0.34393638170974156,
+      "grad_norm": 1.3834050979851695,
+      "learning_rate": 5.4920634920634923e-05,
+      "loss": 0.634,
+      "step": 173
+    },
+    {
+      "epoch": 0.34592445328031807,
+      "grad_norm": 3.0435619097934263,
+      "learning_rate": 5.5238095238095244e-05,
+      "loss": 0.6523,
+      "step": 174
+    },
+    {
+      "epoch": 0.34791252485089463,
+      "grad_norm": 2.9421056263010574,
+      "learning_rate": 5.555555555555556e-05,
+      "loss": 0.6548,
+      "step": 175
+    },
+    {
+      "epoch": 0.3499005964214712,
+      "grad_norm": 1.4643678503113071,
+      "learning_rate": 5.587301587301587e-05,
+      "loss": 0.6332,
+      "step": 176
+    },
+    {
+      "epoch": 0.3518886679920477,
+      "grad_norm": 1.9166228640874332,
+      "learning_rate": 5.619047619047619e-05,
+      "loss": 0.6409,
+      "step": 177
+    },
+    {
+      "epoch": 0.3538767395626243,
+      "grad_norm": 1.4000017299232366,
+      "learning_rate": 5.6507936507936514e-05,
+      "loss": 0.638,
+      "step": 178
+    },
+    {
+      "epoch": 0.3558648111332008,
+      "grad_norm": 1.6183232051146397,
+      "learning_rate": 5.6825396825396835e-05,
+      "loss": 0.6336,
+      "step": 179
+    },
+    {
+      "epoch": 0.35785288270377735,
+      "grad_norm": 1.5556653610365612,
+      "learning_rate": 5.714285714285715e-05,
+      "loss": 0.6382,
+      "step": 180
+    },
+    {
+      "epoch": 0.35984095427435386,
+      "grad_norm": 1.2516421633034058,
+      "learning_rate": 5.746031746031746e-05,
+      "loss": 0.6289,
+      "step": 181
+    },
+    {
+      "epoch": 0.36182902584493043,
+      "grad_norm": 1.3620554525101845,
+      "learning_rate": 5.777777777777778e-05,
+      "loss": 0.637,
+      "step": 182
+    },
+    {
+      "epoch": 0.36381709741550694,
+      "grad_norm": 0.8661709785769057,
+      "learning_rate": 5.80952380952381e-05,
+      "loss": 0.6379,
+      "step": 183
+    },
+    {
+      "epoch": 0.3658051689860835,
+      "grad_norm": 1.4436211568633606,
+      "learning_rate": 5.841269841269841e-05,
+      "loss": 0.6393,
+      "step": 184
+    },
+    {
+      "epoch": 0.36779324055666,
+      "grad_norm": 1.12234286687745,
+      "learning_rate": 5.873015873015873e-05,
+      "loss": 0.6326,
+      "step": 185
+    },
+    {
+      "epoch": 0.3697813121272366,
+      "grad_norm": 1.5736829330194766,
+      "learning_rate": 5.904761904761905e-05,
+      "loss": 0.6331,
+      "step": 186
+    },
+    {
+      "epoch": 0.3717693836978131,
+      "grad_norm": 1.3310376751788484,
+      "learning_rate": 5.9365079365079374e-05,
+      "loss": 0.6398,
+      "step": 187
+    },
+    {
+      "epoch": 0.37375745526838966,
+      "grad_norm": 1.2257869128697072,
+      "learning_rate": 5.968253968253969e-05,
+      "loss": 0.6335,
+      "step": 188
+    },
+    {
+      "epoch": 0.3757455268389662,
+      "grad_norm": 1.6139532041614482,
+      "learning_rate": 6.000000000000001e-05,
+      "loss": 0.6392,
+      "step": 189
+    },
+    {
+      "epoch": 0.37773359840954274,
+      "grad_norm": 1.3350592444809761,
+      "learning_rate": 6.031746031746032e-05,
+      "loss": 0.6382,
+      "step": 190
+    },
+    {
+      "epoch": 0.3797216699801193,
+      "grad_norm": 1.2377247673288536,
+      "learning_rate": 6.0634920634920636e-05,
+      "loss": 0.6246,
+      "step": 191
+    },
+    {
+      "epoch": 0.3817097415506958,
+      "grad_norm": 1.49455356135162,
+      "learning_rate": 6.095238095238096e-05,
+      "loss": 0.6318,
+      "step": 192
+    },
+    {
+      "epoch": 0.3836978131212724,
+      "grad_norm": 1.596684995637217,
+      "learning_rate": 6.126984126984128e-05,
+      "loss": 0.6287,
+      "step": 193
+    },
+    {
+      "epoch": 0.3856858846918489,
+      "grad_norm": 0.6665298437010337,
+      "learning_rate": 6.158730158730159e-05,
+      "loss": 0.6263,
+      "step": 194
+    },
+    {
+      "epoch": 0.38767395626242546,
+      "grad_norm": 1.756691203631161,
+      "learning_rate": 6.19047619047619e-05,
+      "loss": 0.6267,
+      "step": 195
+    },
+    {
+      "epoch": 0.38966202783300197,
+      "grad_norm": 1.0331188464436891,
+      "learning_rate": 6.222222222222223e-05,
+      "loss": 0.6212,
+      "step": 196
+    },
+    {
+      "epoch": 0.39165009940357853,
+      "grad_norm": 1.122470045661076,
+      "learning_rate": 6.253968253968255e-05,
+      "loss": 0.6284,
+      "step": 197
+    },
+    {
+      "epoch": 0.39363817097415504,
+      "grad_norm": 1.7664037452706398,
+      "learning_rate": 6.285714285714286e-05,
+      "loss": 0.642,
+      "step": 198
+    },
+    {
+      "epoch": 0.3956262425447316,
+      "grad_norm": 1.486929540611025,
+      "learning_rate": 6.317460317460318e-05,
+      "loss": 0.6246,
+      "step": 199
+    },
+    {
+      "epoch": 0.3976143141153082,
+      "grad_norm": 1.2175342983856243,
+      "learning_rate": 6.349206349206349e-05,
+      "loss": 0.6313,
+      "step": 200
+    },
+    {
+      "epoch": 0.3996023856858847,
+      "grad_norm": 1.4121892732145778,
+      "learning_rate": 6.380952380952382e-05,
+      "loss": 0.6329,
+      "step": 201
+    },
+    {
+      "epoch": 0.40159045725646125,
+      "grad_norm": 1.1436798208988694,
+      "learning_rate": 6.412698412698413e-05,
+      "loss": 0.6233,
+      "step": 202
+    },
+    {
+      "epoch": 0.40357852882703776,
+      "grad_norm": 1.8368150848847797,
+      "learning_rate": 6.444444444444446e-05,
+      "loss": 0.6224,
+      "step": 203
+    },
+    {
+      "epoch": 0.40556660039761433,
+      "grad_norm": 1.065714896345243,
+      "learning_rate": 6.476190476190477e-05,
+      "loss": 0.6279,
+      "step": 204
+    },
+    {
+      "epoch": 0.40755467196819084,
+      "grad_norm": 1.4516811834829857,
+      "learning_rate": 6.507936507936509e-05,
+      "loss": 0.635,
+      "step": 205
+    },
+    {
+      "epoch": 0.4095427435387674,
+      "grad_norm": 1.4806307032968837,
+      "learning_rate": 6.53968253968254e-05,
+      "loss": 0.6236,
+      "step": 206
+    },
+    {
+      "epoch": 0.4115308151093439,
+      "grad_norm": 1.8498374502351982,
+      "learning_rate": 6.571428571428571e-05,
+      "loss": 0.6285,
+      "step": 207
+    },
+    {
+      "epoch": 0.4135188866799205,
+      "grad_norm": 0.8403035362769382,
+      "learning_rate": 6.603174603174603e-05,
+      "loss": 0.6231,
+      "step": 208
+    },
+    {
+      "epoch": 0.415506958250497,
+      "grad_norm": 2.091792915751743,
+      "learning_rate": 6.634920634920636e-05,
+      "loss": 0.6309,
+      "step": 209
+    },
+    {
+      "epoch": 0.41749502982107356,
+      "grad_norm": 1.26117745196078,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 0.6325,
+      "step": 210
+    },
+    {
+      "epoch": 0.4194831013916501,
+      "grad_norm": 1.783033140481242,
+      "learning_rate": 6.6984126984127e-05,
+      "loss": 0.6309,
+      "step": 211
+    },
+    {
+      "epoch": 0.42147117296222664,
+      "grad_norm": 1.8515301788648455,
+      "learning_rate": 6.730158730158731e-05,
+      "loss": 0.6382,
+      "step": 212
+    },
+    {
+      "epoch": 0.4234592445328032,
+      "grad_norm": 1.1778109141989737,
+      "learning_rate": 6.761904761904763e-05,
+      "loss": 0.6289,
+      "step": 213
+    },
+    {
+      "epoch": 0.4254473161033797,
+      "grad_norm": 2.54816769327095,
+      "learning_rate": 6.793650793650794e-05,
+      "loss": 0.6371,
+      "step": 214
+    },
+    {
+      "epoch": 0.4274353876739563,
+      "grad_norm": 2.048704735899834,
+      "learning_rate": 6.825396825396825e-05,
+      "loss": 0.6398,
+      "step": 215
+    },
+    {
+      "epoch": 0.4294234592445328,
+      "grad_norm": 2.1019781611448853,
+      "learning_rate": 6.857142857142857e-05,
+      "loss": 0.6295,
+      "step": 216
+    },
+    {
+      "epoch": 0.43141153081510936,
+      "grad_norm": 1.7281786345970265,
+      "learning_rate": 6.88888888888889e-05,
+      "loss": 0.6275,
+      "step": 217
+    },
+    {
+      "epoch": 0.43339960238568587,
+      "grad_norm": 2.14951909400337,
+      "learning_rate": 6.920634920634921e-05,
+      "loss": 0.6331,
+      "step": 218
+    },
+    {
+      "epoch": 0.43538767395626243,
+      "grad_norm": 1.7356401726987714,
+      "learning_rate": 6.952380952380954e-05,
+      "loss": 0.6333,
+      "step": 219
+    },
+    {
+      "epoch": 0.43737574552683894,
+      "grad_norm": 1.9690563790102926,
+      "learning_rate": 6.984126984126985e-05,
+      "loss": 0.6267,
+      "step": 220
+    },
+    {
+      "epoch": 0.4393638170974155,
+      "grad_norm": 1.3726181777240785,
+      "learning_rate": 7.015873015873016e-05,
+      "loss": 0.6292,
+      "step": 221
+    },
+    {
+      "epoch": 0.441351888667992,
+      "grad_norm": 1.8939096029386184,
+      "learning_rate": 7.047619047619048e-05,
+      "loss": 0.6188,
+      "step": 222
+    },
+    {
+      "epoch": 0.4433399602385686,
+      "grad_norm": 1.2762772953205688,
+      "learning_rate": 7.079365079365079e-05,
+      "loss": 0.6273,
+      "step": 223
+    },
+    {
+      "epoch": 0.44532803180914515,
+      "grad_norm": 1.776596462339463,
+      "learning_rate": 7.11111111111111e-05,
+      "loss": 0.6239,
+      "step": 224
+    },
+    {
+      "epoch": 0.44731610337972166,
+      "grad_norm": 1.3274942963495349,
+      "learning_rate": 7.142857142857143e-05,
+      "loss": 0.6241,
+      "step": 225
+    },
+    {
+      "epoch": 0.44930417495029823,
+      "grad_norm": 1.8591216328666749,
+      "learning_rate": 7.174603174603175e-05,
+      "loss": 0.6249,
+      "step": 226
+    },
+    {
+      "epoch": 0.45129224652087474,
+      "grad_norm": 1.632104839191781,
+      "learning_rate": 7.206349206349208e-05,
+      "loss": 0.622,
+      "step": 227
+    },
+    {
+      "epoch": 0.4532803180914513,
+      "grad_norm": 1.5321880702968802,
+      "learning_rate": 7.238095238095239e-05,
+      "loss": 0.624,
+      "step": 228
+    },
+    {
+      "epoch": 0.4552683896620278,
+      "grad_norm": 1.3222149905940208,
+      "learning_rate": 7.26984126984127e-05,
+      "loss": 0.6176,
+      "step": 229
+    },
+    {
+      "epoch": 0.4572564612326044,
+      "grad_norm": 1.313836019162892,
+      "learning_rate": 7.301587301587302e-05,
+      "loss": 0.6143,
+      "step": 230
+    },
+    {
+      "epoch": 0.4592445328031809,
+      "grad_norm": 1.131038423221744,
+      "learning_rate": 7.333333333333333e-05,
+      "loss": 0.6188,
+      "step": 231
+    },
+    {
+      "epoch": 0.46123260437375746,
+      "grad_norm": 1.283678851750674,
+      "learning_rate": 7.365079365079366e-05,
+      "loss": 0.619,
+      "step": 232
+    },
+    {
+      "epoch": 0.46322067594433397,
+      "grad_norm": 1.6515171578755334,
+      "learning_rate": 7.396825396825397e-05,
+      "loss": 0.6262,
+      "step": 233
+    },
+    {
+      "epoch": 0.46520874751491054,
+      "grad_norm": 1.2107665827950345,
+      "learning_rate": 7.42857142857143e-05,
+      "loss": 0.6134,
+      "step": 234
+    },
+    {
+      "epoch": 0.4671968190854871,
+      "grad_norm": 1.3081158148843528,
+      "learning_rate": 7.460317460317461e-05,
+      "loss": 0.6207,
+      "step": 235
+    },
+    {
+      "epoch": 0.4691848906560636,
+      "grad_norm": 1.1721076028263735,
+      "learning_rate": 7.492063492063493e-05,
+      "loss": 0.625,
+      "step": 236
+    },
+    {
+      "epoch": 0.4711729622266402,
+      "grad_norm": 1.5047073563346702,
+      "learning_rate": 7.523809523809524e-05,
+      "loss": 0.62,
+      "step": 237
+    },
+    {
+      "epoch": 0.4731610337972167,
+      "grad_norm": 0.8406971392950173,
+      "learning_rate": 7.555555555555556e-05,
+      "loss": 0.621,
+      "step": 238
+    },
+    {
+      "epoch": 0.47514910536779326,
+      "grad_norm": 1.2339916132908735,
+      "learning_rate": 7.587301587301587e-05,
+      "loss": 0.6271,
+      "step": 239
+    },
+    {
+      "epoch": 0.47713717693836977,
+      "grad_norm": 1.3926448624868792,
+      "learning_rate": 7.61904761904762e-05,
+      "loss": 0.6218,
+      "step": 240
+    },
+    {
+      "epoch": 0.47912524850894633,
+      "grad_norm": 2.204044181851112,
+      "learning_rate": 7.650793650793651e-05,
+      "loss": 0.6193,
+      "step": 241
+    },
+    {
+      "epoch": 0.48111332007952284,
+      "grad_norm": 0.8533613284104712,
+      "learning_rate": 7.682539682539684e-05,
+      "loss": 0.6136,
+      "step": 242
+    },
+    {
+      "epoch": 0.4831013916500994,
+      "grad_norm": 1.9892818324819723,
+      "learning_rate": 7.714285714285715e-05,
+      "loss": 0.6392,
+      "step": 243
+    },
+    {
+      "epoch": 0.4850894632206759,
+      "grad_norm": 1.3210748896263227,
+      "learning_rate": 7.746031746031747e-05,
+      "loss": 0.6274,
+      "step": 244
+    },
+    {
+      "epoch": 0.4870775347912525,
+      "grad_norm": 1.257534879967076,
+      "learning_rate": 7.777777777777778e-05,
+      "loss": 0.6267,
+      "step": 245
+    },
+    {
+      "epoch": 0.48906560636182905,
+      "grad_norm": 1.5681584428773365,
+      "learning_rate": 7.80952380952381e-05,
+      "loss": 0.6334,
+      "step": 246
+    },
+    {
+      "epoch": 0.49105367793240556,
+      "grad_norm": 1.217282156325022,
+      "learning_rate": 7.841269841269841e-05,
+      "loss": 0.6188,
+      "step": 247
+    },
+    {
+      "epoch": 0.49304174950298213,
+      "grad_norm": 1.668210588102185,
+      "learning_rate": 7.873015873015874e-05,
+      "loss": 0.6289,
+      "step": 248
+    },
+    {
+      "epoch": 0.49502982107355864,
+      "grad_norm": 1.1732055301577373,
+      "learning_rate": 7.904761904761905e-05,
+      "loss": 0.6277,
+      "step": 249
+    },
+    {
+      "epoch": 0.4970178926441352,
+      "grad_norm": 1.1391292798020574,
+      "learning_rate": 7.936507936507938e-05,
+      "loss": 0.6252,
+      "step": 250
+    },
+    {
+      "epoch": 0.4990059642147117,
+      "grad_norm": 1.3420755566613047,
+      "learning_rate": 7.968253968253969e-05,
+      "loss": 0.6229,
+      "step": 251
+    },
+    {
+      "epoch": 0.5009940357852882,
+      "grad_norm": 1.7212519000230158,
+      "learning_rate": 8e-05,
+      "loss": 0.6283,
+      "step": 252
+    },
+    {
+      "epoch": 0.5029821073558648,
+      "grad_norm": 0.8460706799764476,
+      "learning_rate": 7.999996145566446e-05,
+      "loss": 0.6277,
+      "step": 253
+    },
+    {
+      "epoch": 0.5049701789264414,
+      "grad_norm": 1.2028891530381527,
+      "learning_rate": 7.999984582273212e-05,
+      "loss": 0.6298,
+      "step": 254
+    },
+    {
+      "epoch": 0.5069582504970179,
+      "grad_norm": 1.6321071690462996,
+      "learning_rate": 7.999965310142581e-05,
+      "loss": 0.6218,
+      "step": 255
+    },
+    {
+      "epoch": 0.5089463220675944,
+      "grad_norm": 0.8020880986907298,
+      "learning_rate": 7.999938329211695e-05,
+      "loss": 0.6145,
+      "step": 256
+    },
+    {
+      "epoch": 0.510934393638171,
+      "grad_norm": 1.5123244136663403,
+      "learning_rate": 7.999903639532553e-05,
+      "loss": 0.6233,
+      "step": 257
+    },
+    {
+      "epoch": 0.5129224652087475,
+      "grad_norm": 0.8619895667367529,
+      "learning_rate": 7.99986124117201e-05,
+      "loss": 0.6259,
+      "step": 258
+    },
+    {
+      "epoch": 0.5149105367793241,
+      "grad_norm": 1.674504615597631,
+      "learning_rate": 7.999811134211777e-05,
+      "loss": 0.623,
+      "step": 259
+    },
+    {
+      "epoch": 0.5168986083499006,
+      "grad_norm": 1.244705549383792,
+      "learning_rate": 7.99975331874842e-05,
+      "loss": 0.6259,
+      "step": 260
+    },
+    {
+      "epoch": 0.5188866799204771,
+      "grad_norm": 1.4164036526291948,
+      "learning_rate": 7.999687794893362e-05,
+      "loss": 0.6254,
+      "step": 261
+    },
+    {
+      "epoch": 0.5208747514910537,
+      "grad_norm": 1.1420458314329858,
+      "learning_rate": 7.999614562772881e-05,
+      "loss": 0.6224,
+      "step": 262
+    },
+    {
+      "epoch": 0.5228628230616302,
+      "grad_norm": 0.9971122950245115,
+      "learning_rate": 7.999533622528113e-05,
+      "loss": 0.6167,
+      "step": 263
+    },
+    {
+      "epoch": 0.5248508946322068,
+      "grad_norm": 1.53841350114554,
+      "learning_rate": 7.999444974315046e-05,
+      "loss": 0.6203,
+      "step": 264
+    },
+    {
+      "epoch": 0.5268389662027833,
+      "grad_norm": 1.355905799554643,
+      "learning_rate": 7.999348618304524e-05,
+      "loss": 0.6209,
+      "step": 265
+    },
+    {
+      "epoch": 0.5288270377733598,
+      "grad_norm": 1.1382955214819985,
+      "learning_rate": 7.999244554682249e-05,
+      "loss": 0.6137,
+      "step": 266
+    },
+    {
+      "epoch": 0.5308151093439364,
+      "grad_norm": 1.2022228391943435,
+      "learning_rate": 7.99913278364877e-05,
+      "loss": 0.6145,
+      "step": 267
+    },
+    {
+      "epoch": 0.532803180914513,
+      "grad_norm": 1.086139055972047,
+      "learning_rate": 7.999013305419497e-05,
+      "loss": 0.6145,
+      "step": 268
+    },
+    {
+      "epoch": 0.5347912524850894,
+      "grad_norm": 1.322274294817971,
+      "learning_rate": 7.998886120224688e-05,
+      "loss": 0.6234,
+      "step": 269
+    },
+    {
+      "epoch": 0.536779324055666,
+      "grad_norm": 0.9234044414783897,
+      "learning_rate": 7.998751228309457e-05,
+      "loss": 0.6144,
+      "step": 270
+    },
+    {
+      "epoch": 0.5387673956262425,
+      "grad_norm": 0.9326761013754535,
+      "learning_rate": 7.998608629933773e-05,
+      "loss": 0.61,
+      "step": 271
+    },
+    {
+      "epoch": 0.5407554671968191,
+      "grad_norm": 1.5279987086559388,
+      "learning_rate": 7.99845832537245e-05,
+      "loss": 0.6174,
+      "step": 272
+    },
+    {
+      "epoch": 0.5427435387673957,
+      "grad_norm": 0.9928341222876796,
+      "learning_rate": 7.99830031491516e-05,
+      "loss": 0.616,
+      "step": 273
+    },
+    {
+      "epoch": 0.5447316103379721,
+      "grad_norm": 1.3570459611370076,
+      "learning_rate": 7.998134598866422e-05,
+      "loss": 0.6187,
+      "step": 274
+    },
+    {
+      "epoch": 0.5467196819085487,
+      "grad_norm": 1.4557589984647394,
+      "learning_rate": 7.997961177545607e-05,
+      "loss": 0.6135,
+      "step": 275
+    },
+    {
+      "epoch": 0.5487077534791253,
+      "grad_norm": 0.7773435304804408,
+      "learning_rate": 7.997780051286937e-05,
+      "loss": 0.612,
+      "step": 276
+    },
+    {
+      "epoch": 0.5506958250497018,
+      "grad_norm": 1.223588643454511,
+      "learning_rate": 7.997591220439479e-05,
+      "loss": 0.6148,
+      "step": 277
+    },
+    {
+      "epoch": 0.5526838966202783,
+      "grad_norm": 1.4724994222951164,
+      "learning_rate": 7.997394685367153e-05,
+      "loss": 0.6149,
+      "step": 278
+    },
+    {
+      "epoch": 0.5546719681908548,
+      "grad_norm": 0.8204200684143781,
+      "learning_rate": 7.997190446448724e-05,
+      "loss": 0.6115,
+      "step": 279
+    },
+    {
+      "epoch": 0.5566600397614314,
+      "grad_norm": 1.3559612721067071,
+      "learning_rate": 7.996978504077805e-05,
+      "loss": 0.6089,
+      "step": 280
+    },
+    {
+      "epoch": 0.558648111332008,
+      "grad_norm": 1.0310273576587536,
+      "learning_rate": 7.996758858662855e-05,
+      "loss": 0.6174,
+      "step": 281
+    },
+    {
+      "epoch": 0.5606361829025845,
+      "grad_norm": 0.9946930215525382,
+      "learning_rate": 7.996531510627176e-05,
+      "loss": 0.6153,
+      "step": 282
+    },
+    {
+      "epoch": 0.562624254473161,
+      "grad_norm": 1.1224346378529806,
+      "learning_rate": 7.996296460408921e-05,
+      "loss": 0.6119,
+      "step": 283
+    },
+    {
+      "epoch": 0.5646123260437376,
+      "grad_norm": 1.3765896724427593,
+      "learning_rate": 7.996053708461078e-05,
+      "loss": 0.6156,
+      "step": 284
+    },
+    {
+      "epoch": 0.5666003976143141,
+      "grad_norm": 1.0089569688803084,
+      "learning_rate": 7.995803255251488e-05,
+      "loss": 0.6088,
+      "step": 285
+    },
+    {
+      "epoch": 0.5685884691848907,
+      "grad_norm": 0.950972323050223,
+      "learning_rate": 7.995545101262825e-05,
+      "loss": 0.6104,
+      "step": 286
+    },
+    {
+      "epoch": 0.5705765407554672,
+      "grad_norm": 1.4332474212744974,
+      "learning_rate": 7.995279246992608e-05,
+      "loss": 0.6166,
+      "step": 287
+    },
+    {
+      "epoch": 0.5725646123260437,
+      "grad_norm": 1.1557864927220203,
+      "learning_rate": 7.995005692953196e-05,
+      "loss": 0.607,
+      "step": 288
+    },
+    {
+      "epoch": 0.5745526838966203,
+      "grad_norm": 0.8094455060578747,
+      "learning_rate": 7.994724439671787e-05,
+      "loss": 0.6079,
+      "step": 289
+    },
+    {
+      "epoch": 0.5765407554671969,
+      "grad_norm": 1.058813447760782,
+      "learning_rate": 7.994435487690418e-05,
+      "loss": 0.6159,
+      "step": 290
+    },
+    {
+      "epoch": 0.5785288270377733,
+      "grad_norm": 1.7842490215031124,
+      "learning_rate": 7.994138837565962e-05,
+      "loss": 0.6064,
+      "step": 291
+    },
+    {
+      "epoch": 0.5805168986083499,
+      "grad_norm": 0.7937318931864666,
+      "learning_rate": 7.993834489870127e-05,
+      "loss": 0.61,
+      "step": 292
+    },
+    {
+      "epoch": 0.5825049701789264,
+      "grad_norm": 2.0448392448161266,
+      "learning_rate": 7.993522445189458e-05,
+      "loss": 0.6187,
+      "step": 293
+    },
+    {
+      "epoch": 0.584493041749503,
+      "grad_norm": 1.1616482588221761,
+      "learning_rate": 7.993202704125332e-05,
+      "loss": 0.6168,
+      "step": 294
+    },
+    {
+      "epoch": 0.5864811133200796,
+      "grad_norm": 1.7634514220529318,
+      "learning_rate": 7.992875267293958e-05,
+      "loss": 0.6202,
+      "step": 295
+    },
+    {
+      "epoch": 0.588469184890656,
+      "grad_norm": 1.3557395854086942,
+      "learning_rate": 7.99254013532638e-05,
+      "loss": 0.6287,
+      "step": 296
+    },
+    {
+      "epoch": 0.5904572564612326,
+      "grad_norm": 1.221924990086436,
+      "learning_rate": 7.99219730886847e-05,
+      "loss": 0.6177,
+      "step": 297
+    },
+    {
+      "epoch": 0.5924453280318092,
+      "grad_norm": 1.1570566780155553,
+      "learning_rate": 7.991846788580928e-05,
+      "loss": 0.6202,
+      "step": 298
+    },
+    {
+      "epoch": 0.5944333996023857,
+      "grad_norm": 0.8682851210033183,
+      "learning_rate": 7.991488575139283e-05,
+      "loss": 0.6108,
+      "step": 299
+    },
+    {
+      "epoch": 0.5964214711729622,
+      "grad_norm": 0.9892322185926827,
+      "learning_rate": 7.99112266923389e-05,
+      "loss": 0.6093,
+      "step": 300
+    },
+    {
+      "epoch": 0.5984095427435387,
+      "grad_norm": 1.3225369418198032,
+      "learning_rate": 7.990749071569928e-05,
+      "loss": 0.6116,
+      "step": 301
+    },
+    {
+      "epoch": 0.6003976143141153,
+      "grad_norm": 0.7090274559838418,
+      "learning_rate": 7.990367782867403e-05,
+      "loss": 0.6068,
+      "step": 302
+    },
+    {
+      "epoch": 0.6023856858846919,
+      "grad_norm": 1.1894865112831852,
+      "learning_rate": 7.989978803861138e-05,
+      "loss": 0.6106,
+      "step": 303
+    },
+    {
+      "epoch": 0.6043737574552683,
+      "grad_norm": 0.6833264633772178,
+      "learning_rate": 7.98958213530078e-05,
+      "loss": 0.6058,
+      "step": 304
+    },
+    {
+      "epoch": 0.6063618290258449,
+      "grad_norm": 0.9084651046263741,
+      "learning_rate": 7.989177777950798e-05,
+      "loss": 0.6014,
+      "step": 305
+    },
+    {
+      "epoch": 0.6083499005964215,
+      "grad_norm": 0.9853511757871085,
+      "learning_rate": 7.988765732590476e-05,
+      "loss": 0.6043,
+      "step": 306
+    },
+    {
+      "epoch": 0.610337972166998,
+      "grad_norm": 1.1489434663121463,
+      "learning_rate": 7.988346000013912e-05,
+      "loss": 0.6042,
+      "step": 307
+    },
+    {
+      "epoch": 0.6123260437375746,
+      "grad_norm": 0.9477747519287059,
+      "learning_rate": 7.987918581030024e-05,
+      "loss": 0.6039,
+      "step": 308
+    },
+    {
+      "epoch": 0.614314115308151,
+      "grad_norm": 1.1393103642754758,
+      "learning_rate": 7.987483476462539e-05,
+      "loss": 0.6028,
+      "step": 309
+    },
+    {
+      "epoch": 0.6163021868787276,
+      "grad_norm": 1.2313579785241444,
+      "learning_rate": 7.98704068715e-05,
+      "loss": 0.6103,
+      "step": 310
+    },
+    {
+      "epoch": 0.6182902584493042,
+      "grad_norm": 0.9717710635949429,
+      "learning_rate": 7.986590213945756e-05,
+      "loss": 0.606,
+      "step": 311
+    },
+    {
+      "epoch": 0.6202783300198808,
+      "grad_norm": 1.0672532374525363,
+      "learning_rate": 7.98613205771797e-05,
+      "loss": 0.6003,
+      "step": 312
+    },
+    {
+      "epoch": 0.6222664015904572,
+      "grad_norm": 1.3583742092426443,
+      "learning_rate": 7.985666219349604e-05,
+      "loss": 0.6022,
+      "step": 313
+    },
+    {
+      "epoch": 0.6242544731610338,
+      "grad_norm": 0.5837502834461586,
+      "learning_rate": 7.985192699738432e-05,
+      "loss": 0.5966,
+      "step": 314
+    },
+    {
+      "epoch": 0.6262425447316103,
+      "grad_norm": 0.6446776148368711,
+      "learning_rate": 7.984711499797029e-05,
+      "loss": 0.6038,
+      "step": 315
+    },
+    {
+      "epoch": 0.6282306163021869,
+      "grad_norm": 1.2386506977675524,
+      "learning_rate": 7.98422262045277e-05,
+      "loss": 0.6077,
+      "step": 316
+    },
+    {
+      "epoch": 0.6302186878727635,
+      "grad_norm": 1.1751514354670538,
+      "learning_rate": 7.983726062647833e-05,
+      "loss": 0.6042,
+      "step": 317
+    },
+    {
+      "epoch": 0.6322067594433399,
+      "grad_norm": 0.8754529089309102,
+      "learning_rate": 7.983221827339192e-05,
+      "loss": 0.6005,
+      "step": 318
+    },
+    {
+      "epoch": 0.6341948310139165,
+      "grad_norm": 1.070805762546688,
+      "learning_rate": 7.982709915498618e-05,
+      "loss": 0.603,
+      "step": 319
+    },
+    {
+      "epoch": 0.6361829025844931,
+      "grad_norm": 1.2753097622493266,
+      "learning_rate": 7.982190328112674e-05,
+      "loss": 0.601,
+      "step": 320
+    },
+    {
+      "epoch": 0.6381709741550696,
+      "grad_norm": 0.755670446629981,
+      "learning_rate": 7.981663066182721e-05,
+      "loss": 0.6011,
+      "step": 321
+    },
+    {
+      "epoch": 0.6401590457256461,
+      "grad_norm": 0.7148388951079416,
+      "learning_rate": 7.981128130724907e-05,
+      "loss": 0.6,
+      "step": 322
+    },
+    {
+      "epoch": 0.6421471172962226,
+      "grad_norm": 0.9122439016384183,
+      "learning_rate": 7.980585522770166e-05,
+      "loss": 0.6017,
+      "step": 323
+    },
+    {
+      "epoch": 0.6441351888667992,
+      "grad_norm": 1.3055890086405564,
+      "learning_rate": 7.980035243364221e-05,
+      "loss": 0.6086,
+      "step": 324
+    },
+    {
+      "epoch": 0.6461232604373758,
+      "grad_norm": 0.8349565637932869,
+      "learning_rate": 7.97947729356758e-05,
+      "loss": 0.5988,
+      "step": 325
+    },
+    {
+      "epoch": 0.6481113320079522,
+      "grad_norm": 0.9251569550381776,
+      "learning_rate": 7.978911674455538e-05,
+      "loss": 0.6086,
+      "step": 326
+    },
+    {
+      "epoch": 0.6500994035785288,
+      "grad_norm": 0.9646584755316521,
+      "learning_rate": 7.978338387118158e-05,
+      "loss": 0.602,
+      "step": 327
+    },
+    {
+      "epoch": 0.6520874751491054,
+      "grad_norm": 0.941633551491786,
+      "learning_rate": 7.977757432660295e-05,
+      "loss": 0.6055,
+      "step": 328
+    },
+    {
+      "epoch": 0.6540755467196819,
+      "grad_norm": 0.9205186480063723,
+      "learning_rate": 7.97716881220157e-05,
+      "loss": 0.6039,
+      "step": 329
+    },
+    {
+      "epoch": 0.6560636182902585,
+      "grad_norm": 0.9330713864830623,
+      "learning_rate": 7.976572526876386e-05,
+      "loss": 0.6007,
+      "step": 330
+    },
+    {
+      "epoch": 0.658051689860835,
+      "grad_norm": 1.259958088941969,
+      "learning_rate": 7.975968577833911e-05,
+      "loss": 0.6045,
+      "step": 331
+    },
+    {
+      "epoch": 0.6600397614314115,
+      "grad_norm": 0.8298038712817966,
+      "learning_rate": 7.975356966238087e-05,
+      "loss": 0.5998,
+      "step": 332
+    },
+    {
+      "epoch": 0.6620278330019881,
+      "grad_norm": 0.7181692734835541,
+      "learning_rate": 7.974737693267621e-05,
+      "loss": 0.6029,
+      "step": 333
+    },
+    {
+      "epoch": 0.6640159045725647,
+      "grad_norm": 0.7747035845063713,
+      "learning_rate": 7.97411076011599e-05,
+      "loss": 0.595,
+      "step": 334
+    },
+    {
+      "epoch": 0.6660039761431411,
+      "grad_norm": 0.8763620703275192,
+      "learning_rate": 7.973476167991426e-05,
+      "loss": 0.5947,
+      "step": 335
+    },
+    {
+      "epoch": 0.6679920477137177,
+      "grad_norm": 1.0767794678895612,
+      "learning_rate": 7.972833918116926e-05,
+      "loss": 0.596,
+      "step": 336
+    },
+    {
+      "epoch": 0.6699801192842942,
+      "grad_norm": 1.2632858490991365,
+      "learning_rate": 7.972184011730246e-05,
+      "loss": 0.5967,
+      "step": 337
+    },
+    {
+      "epoch": 0.6719681908548708,
+      "grad_norm": 0.8583135331342279,
+      "learning_rate": 7.971526450083897e-05,
+      "loss": 0.5942,
+      "step": 338
+    },
+    {
+      "epoch": 0.6739562624254473,
+      "grad_norm": 1.080095604857901,
+      "learning_rate": 7.970861234445142e-05,
+      "loss": 0.5962,
+      "step": 339
+    },
+    {
+      "epoch": 0.6759443339960238,
+      "grad_norm": 1.2462305949675103,
+      "learning_rate": 7.970188366095996e-05,
+      "loss": 0.5994,
+      "step": 340
+    },
+    {
+      "epoch": 0.6779324055666004,
+      "grad_norm": 0.8353002045769494,
+      "learning_rate": 7.969507846333221e-05,
+      "loss": 0.5925,
+      "step": 341
+    },
+    {
+      "epoch": 0.679920477137177,
+      "grad_norm": 0.6862279336154131,
+      "learning_rate": 7.968819676468327e-05,
+      "loss": 0.5942,
+      "step": 342
+    },
+    {
+      "epoch": 0.6819085487077535,
+      "grad_norm": 0.7767309327289988,
+      "learning_rate": 7.968123857827568e-05,
+      "loss": 0.5961,
+      "step": 343
+    },
+    {
+      "epoch": 0.68389662027833,
+      "grad_norm": 0.8656265135412583,
+      "learning_rate": 7.967420391751935e-05,
+      "loss": 0.6002,
+      "step": 344
+    },
+    {
+      "epoch": 0.6858846918489065,
+      "grad_norm": 0.8852817636204119,
+      "learning_rate": 7.96670927959716e-05,
+      "loss": 0.5973,
+      "step": 345
+    },
+    {
+      "epoch": 0.6878727634194831,
+      "grad_norm": 0.8201129197174135,
+      "learning_rate": 7.965990522733713e-05,
+      "loss": 0.5877,
+      "step": 346
+    },
+    {
+      "epoch": 0.6898608349900597,
+      "grad_norm": 0.7216882280944779,
+      "learning_rate": 7.96526412254679e-05,
+      "loss": 0.5933,
+      "step": 347
+    },
+    {
+      "epoch": 0.6918489065606361,
+      "grad_norm": 0.9822298632247115,
+      "learning_rate": 7.964530080436324e-05,
+      "loss": 0.5922,
+      "step": 348
+    },
+    {
+      "epoch": 0.6938369781312127,
+      "grad_norm": 1.2671518884810031,
+      "learning_rate": 7.963788397816974e-05,
+      "loss": 0.5913,
+      "step": 349
+    },
+    {
+      "epoch": 0.6958250497017893,
+      "grad_norm": 0.7838716714178371,
+      "learning_rate": 7.96303907611812e-05,
+      "loss": 0.591,
+      "step": 350
+    },
+    {
+      "epoch": 0.6978131212723658,
+      "grad_norm": 1.1891668128997623,
+      "learning_rate": 7.962282116783873e-05,
+      "loss": 0.5956,
+      "step": 351
+    },
+    {
+      "epoch": 0.6998011928429424,
+      "grad_norm": 1.1010146966411596,
+      "learning_rate": 7.96151752127305e-05,
+      "loss": 0.5896,
+      "step": 352
+    },
+    {
+      "epoch": 0.7017892644135189,
+      "grad_norm": 1.188668203426625,
+      "learning_rate": 7.9607452910592e-05,
+      "loss": 0.5934,
+      "step": 353
+    },
+    {
+      "epoch": 0.7037773359840954,
+      "grad_norm": 0.8731139591987406,
+      "learning_rate": 7.959965427630574e-05,
+      "loss": 0.5908,
+      "step": 354
+    },
+    {
+      "epoch": 0.705765407554672,
+      "grad_norm": 1.0694314172043198,
+      "learning_rate": 7.959177932490138e-05,
+      "loss": 0.5886,
+      "step": 355
+    },
+    {
+      "epoch": 0.7077534791252486,
+      "grad_norm": 1.2085859343148335,
+      "learning_rate": 7.958382807155566e-05,
+      "loss": 0.5894,
+      "step": 356
+    },
+    {
+      "epoch": 0.709741550695825,
+      "grad_norm": 0.8270753340299087,
+      "learning_rate": 7.957580053159237e-05,
+      "loss": 0.5891,
+      "step": 357
+    },
+    {
+      "epoch": 0.7117296222664016,
+      "grad_norm": 0.7137872070663243,
+      "learning_rate": 7.956769672048232e-05,
+      "loss": 0.5834,
+      "step": 358
+    },
+    {
+      "epoch": 0.7137176938369781,
+      "grad_norm": 0.8113240095237567,
+      "learning_rate": 7.955951665384333e-05,
+      "loss": 0.5882,
+      "step": 359
+    },
+    {
+      "epoch": 0.7157057654075547,
+      "grad_norm": 1.1469104720759737,
+      "learning_rate": 7.955126034744012e-05,
+      "loss": 0.5896,
+      "step": 360
+    },
+    {
+      "epoch": 0.7176938369781312,
+      "grad_norm": 1.180373367476227,
+      "learning_rate": 7.954292781718443e-05,
+      "loss": 0.5928,
+      "step": 361
+    },
+    {
+      "epoch": 0.7196819085487077,
+      "grad_norm": 0.9444939465878108,
+      "learning_rate": 7.953451907913482e-05,
+      "loss": 0.5918,
+      "step": 362
+    },
+    {
+      "epoch": 0.7216699801192843,
+      "grad_norm": 1.0006658840657565,
+      "learning_rate": 7.952603414949674e-05,
+      "loss": 0.5952,
+      "step": 363
+    },
+    {
+      "epoch": 0.7236580516898609,
+      "grad_norm": 1.2325093196009407,
+      "learning_rate": 7.951747304462255e-05,
+      "loss": 0.5951,
+      "step": 364
+    },
+    {
+      "epoch": 0.7256461232604374,
+      "grad_norm": 0.7749758827049822,
+      "learning_rate": 7.950883578101128e-05,
+      "loss": 0.589,
+      "step": 365
+    },
+    {
+      "epoch": 0.7276341948310139,
+      "grad_norm": 0.6979516318617547,
+      "learning_rate": 7.950012237530886e-05,
+      "loss": 0.5833,
+      "step": 366
+    },
+    {
+      "epoch": 0.7296222664015904,
+      "grad_norm": 0.7160911526302164,
+      "learning_rate": 7.949133284430788e-05,
+      "loss": 0.5896,
+      "step": 367
+    },
+    {
+      "epoch": 0.731610337972167,
+      "grad_norm": 0.9105506423506521,
+      "learning_rate": 7.948246720494769e-05,
+      "loss": 0.5923,
+      "step": 368
+    },
+    {
+      "epoch": 0.7335984095427436,
+      "grad_norm": 1.0461010915532079,
+      "learning_rate": 7.947352547431429e-05,
+      "loss": 0.5873,
+      "step": 369
+    },
+    {
+      "epoch": 0.73558648111332,
+      "grad_norm": 0.8558435340216544,
+      "learning_rate": 7.946450766964035e-05,
+      "loss": 0.5931,
+      "step": 370
+    },
+    {
+      "epoch": 0.7375745526838966,
+      "grad_norm": 0.7994654876766595,
+      "learning_rate": 7.945541380830512e-05,
+      "loss": 0.5923,
+      "step": 371
+    },
+    {
+      "epoch": 0.7395626242544732,
+      "grad_norm": 0.893131431334677,
+      "learning_rate": 7.944624390783444e-05,
+      "loss": 0.599,
+      "step": 372
+    },
+    {
+      "epoch": 0.7415506958250497,
+      "grad_norm": 1.2442016855861993,
+      "learning_rate": 7.94369979859007e-05,
+      "loss": 0.5952,
+      "step": 373
+    },
+    {
+      "epoch": 0.7435387673956262,
+      "grad_norm": 0.9079647599598836,
+      "learning_rate": 7.94276760603228e-05,
+      "loss": 0.5892,
+      "step": 374
+    },
+    {
+      "epoch": 0.7455268389662028,
+      "grad_norm": 0.9812110515231576,
+      "learning_rate": 7.941827814906611e-05,
+      "loss": 0.5898,
+      "step": 375
+    },
+    {
+      "epoch": 0.7475149105367793,
+      "grad_norm": 0.9969278983559571,
+      "learning_rate": 7.940880427024243e-05,
+      "loss": 0.587,
+      "step": 376
+    },
+    {
+      "epoch": 0.7495029821073559,
+      "grad_norm": 0.8529726891532479,
+      "learning_rate": 7.939925444211e-05,
+      "loss": 0.5843,
+      "step": 377
+    },
+    {
+      "epoch": 0.7514910536779325,
+      "grad_norm": 0.761433775409296,
+      "learning_rate": 7.93896286830734e-05,
+      "loss": 0.5914,
+      "step": 378
+    },
+    {
+      "epoch": 0.7534791252485089,
+      "grad_norm": 0.8206538735583792,
+      "learning_rate": 7.937992701168353e-05,
+      "loss": 0.5867,
+      "step": 379
+    },
+    {
+      "epoch": 0.7554671968190855,
+      "grad_norm": 0.9583546824707148,
+      "learning_rate": 7.937014944663766e-05,
+      "loss": 0.5882,
+      "step": 380
+    },
+    {
+      "epoch": 0.757455268389662,
+      "grad_norm": 1.2370646283419247,
+      "learning_rate": 7.936029600677924e-05,
+      "loss": 0.5888,
+      "step": 381
+    },
+    {
+      "epoch": 0.7594433399602386,
+      "grad_norm": 0.8519053228451124,
+      "learning_rate": 7.9350366711098e-05,
+      "loss": 0.5886,
+      "step": 382
+    },
+    {
+      "epoch": 0.7614314115308151,
+      "grad_norm": 0.7277634202579358,
+      "learning_rate": 7.934036157872983e-05,
+      "loss": 0.5889,
+      "step": 383
+    },
+    {
+      "epoch": 0.7634194831013916,
+      "grad_norm": 0.8707040598292405,
+      "learning_rate": 7.933028062895683e-05,
+      "loss": 0.5871,
+      "step": 384
+    },
+    {
+      "epoch": 0.7654075546719682,
+      "grad_norm": 1.0438936247793003,
+      "learning_rate": 7.932012388120714e-05,
+      "loss": 0.5905,
+      "step": 385
+    },
+    {
+      "epoch": 0.7673956262425448,
+      "grad_norm": 1.0450358135630395,
+      "learning_rate": 7.9309891355055e-05,
+      "loss": 0.5896,
+      "step": 386
+    },
+    {
+      "epoch": 0.7693836978131213,
+      "grad_norm": 1.2549615714436533,
+      "learning_rate": 7.929958307022074e-05,
+      "loss": 0.5844,
+      "step": 387
+    },
+    {
+      "epoch": 0.7713717693836978,
+      "grad_norm": 0.6622790149096027,
+      "learning_rate": 7.928919904657065e-05,
+      "loss": 0.5848,
+      "step": 388
+    },
+    {
+      "epoch": 0.7733598409542743,
+      "grad_norm": 0.6076325225296729,
+      "learning_rate": 7.9278739304117e-05,
+      "loss": 0.585,
+      "step": 389
+    },
+    {
+      "epoch": 0.7753479125248509,
+      "grad_norm": 0.8778111359677369,
+      "learning_rate": 7.926820386301795e-05,
+      "loss": 0.5915,
+      "step": 390
+    },
+    {
+      "epoch": 0.7773359840954275,
+      "grad_norm": 1.092006397904628,
+      "learning_rate": 7.925759274357763e-05,
+      "loss": 0.5897,
+      "step": 391
+    },
+    {
+      "epoch": 0.7793240556660039,
+      "grad_norm": 1.1071022336112284,
+      "learning_rate": 7.924690596624592e-05,
+      "loss": 0.5849,
+      "step": 392
+    },
+    {
+      "epoch": 0.7813121272365805,
+      "grad_norm": 0.8813657744060249,
+      "learning_rate": 7.923614355161859e-05,
+      "loss": 0.5844,
+      "step": 393
+    },
+    {
+      "epoch": 0.7833001988071571,
+      "grad_norm": 1.144804916630537,
+      "learning_rate": 7.922530552043712e-05,
+      "loss": 0.5888,
+      "step": 394
+    },
+    {
+      "epoch": 0.7852882703777336,
+      "grad_norm": 0.9647130509140422,
+      "learning_rate": 7.921439189358877e-05,
+      "loss": 0.5871,
+      "step": 395
+    },
+    {
+      "epoch": 0.7872763419483101,
+      "grad_norm": 0.8100748237359459,
+      "learning_rate": 7.920340269210643e-05,
+      "loss": 0.5895,
+      "step": 396
+    },
+    {
+      "epoch": 0.7892644135188867,
+      "grad_norm": 0.8424797920827156,
+      "learning_rate": 7.919233793716872e-05,
+      "loss": 0.5812,
+      "step": 397
+    },
+    {
+      "epoch": 0.7912524850894632,
+      "grad_norm": 0.981537746558217,
+      "learning_rate": 7.918119765009979e-05,
+      "loss": 0.5852,
+      "step": 398
+    },
+    {
+      "epoch": 0.7932405566600398,
+      "grad_norm": 1.1186079042253614,
+      "learning_rate": 7.916998185236939e-05,
+      "loss": 0.5844,
+      "step": 399
+    },
+    {
+      "epoch": 0.7952286282306164,
+      "grad_norm": 0.8679154615113002,
+      "learning_rate": 7.915869056559279e-05,
+      "loss": 0.5862,
+      "step": 400
+    },
+    {
+      "epoch": 0.7972166998011928,
+      "grad_norm": 0.7686569644536589,
+      "learning_rate": 7.914732381153079e-05,
+      "loss": 0.5849,
+      "step": 401
+    },
+    {
+      "epoch": 0.7992047713717694,
+      "grad_norm": 0.796490817813745,
+      "learning_rate": 7.913588161208952e-05,
+      "loss": 0.5858,
+      "step": 402
+    },
+    {
+      "epoch": 0.8011928429423459,
+      "grad_norm": 0.8035281403452298,
+      "learning_rate": 7.912436398932063e-05,
+      "loss": 0.5847,
+      "step": 403
+    },
+    {
+      "epoch": 0.8031809145129225,
+      "grad_norm": 1.0308204022548164,
+      "learning_rate": 7.911277096542106e-05,
+      "loss": 0.583,
+      "step": 404
+    },
+    {
+      "epoch": 0.805168986083499,
+      "grad_norm": 1.3305335250593018,
+      "learning_rate": 7.910110256273307e-05,
+      "loss": 0.5864,
+      "step": 405
+    },
+    {
+      "epoch": 0.8071570576540755,
+      "grad_norm": 0.6852998373300851,
+      "learning_rate": 7.908935880374423e-05,
+      "loss": 0.5804,
+      "step": 406
+    },
+    {
+      "epoch": 0.8091451292246521,
+      "grad_norm": 0.6572110197570675,
+      "learning_rate": 7.907753971108728e-05,
+      "loss": 0.58,
+      "step": 407
+    },
+    {
+      "epoch": 0.8111332007952287,
+      "grad_norm": 0.8879964008698213,
+      "learning_rate": 7.906564530754018e-05,
+      "loss": 0.5834,
+      "step": 408
+    },
+    {
+      "epoch": 0.8131212723658051,
+      "grad_norm": 1.0942411047383964,
+      "learning_rate": 7.905367561602603e-05,
+      "loss": 0.5824,
+      "step": 409
+    },
+    {
+      "epoch": 0.8151093439363817,
+      "grad_norm": 0.9734482127356657,
+      "learning_rate": 7.904163065961302e-05,
+      "loss": 0.5836,
+      "step": 410
+    },
+    {
+      "epoch": 0.8170974155069582,
+      "grad_norm": 0.9526293022396148,
+      "learning_rate": 7.90295104615144e-05,
+      "loss": 0.5873,
+      "step": 411
+    },
+    {
+      "epoch": 0.8190854870775348,
+      "grad_norm": 0.9111513732433741,
+      "learning_rate": 7.90173150450884e-05,
+      "loss": 0.5911,
+      "step": 412
+    },
+    {
+      "epoch": 0.8210735586481114,
+      "grad_norm": 1.000262045401118,
+      "learning_rate": 7.900504443383825e-05,
+      "loss": 0.5877,
+      "step": 413
+    },
+    {
+      "epoch": 0.8230616302186878,
+      "grad_norm": 1.0840607430740403,
+      "learning_rate": 7.899269865141206e-05,
+      "loss": 0.581,
+      "step": 414
+    },
+    {
+      "epoch": 0.8250497017892644,
+      "grad_norm": 1.0280039612546272,
+      "learning_rate": 7.898027772160284e-05,
+      "loss": 0.5774,
+      "step": 415
+    },
+    {
+      "epoch": 0.827037773359841,
+      "grad_norm": 1.052852112873632,
+      "learning_rate": 7.896778166834841e-05,
+      "loss": 0.5797,
+      "step": 416
+    },
+    {
+      "epoch": 0.8290258449304175,
+      "grad_norm": 0.9992150239594432,
+      "learning_rate": 7.895521051573139e-05,
+      "loss": 0.5839,
+      "step": 417
+    },
+    {
+      "epoch": 0.831013916500994,
+      "grad_norm": 0.8503482948129996,
+      "learning_rate": 7.894256428797909e-05,
+      "loss": 0.5784,
+      "step": 418
+    },
+    {
+      "epoch": 0.8330019880715706,
+      "grad_norm": 0.7307905930470503,
+      "learning_rate": 7.892984300946354e-05,
+      "loss": 0.5892,
+      "step": 419
+    },
+    {
+      "epoch": 0.8349900596421471,
+      "grad_norm": 0.7304703844232084,
+      "learning_rate": 7.891704670470142e-05,
+      "loss": 0.582,
+      "step": 420
+    },
+    {
+      "epoch": 0.8369781312127237,
+      "grad_norm": 0.8191311030518267,
+      "learning_rate": 7.890417539835395e-05,
+      "loss": 0.5793,
+      "step": 421
+    },
+    {
+      "epoch": 0.8389662027833003,
+      "grad_norm": 0.9761301233522803,
+      "learning_rate": 7.889122911522695e-05,
+      "loss": 0.5811,
+      "step": 422
+    },
+    {
+      "epoch": 0.8409542743538767,
+      "grad_norm": 1.1735275864088888,
+      "learning_rate": 7.887820788027072e-05,
+      "loss": 0.5835,
+      "step": 423
+    },
+    {
+      "epoch": 0.8429423459244533,
+      "grad_norm": 0.6856347281014243,
+      "learning_rate": 7.886511171858e-05,
+      "loss": 0.5856,
+      "step": 424
+    },
+    {
+      "epoch": 0.8449304174950298,
+      "grad_norm": 0.49809432528479436,
+      "learning_rate": 7.885194065539392e-05,
+      "loss": 0.5871,
+      "step": 425
+    },
+    {
+      "epoch": 0.8469184890656064,
+      "grad_norm": 0.6980729511674448,
+      "learning_rate": 7.883869471609598e-05,
+      "loss": 0.5854,
+      "step": 426
+    },
+    {
+      "epoch": 0.8489065606361829,
+      "grad_norm": 0.8088615313801084,
+      "learning_rate": 7.882537392621397e-05,
+      "loss": 0.5788,
+      "step": 427
+    },
+    {
+      "epoch": 0.8508946322067594,
+      "grad_norm": 0.6430145546564471,
+      "learning_rate": 7.881197831141994e-05,
+      "loss": 0.5834,
+      "step": 428
+    },
+    {
+      "epoch": 0.852882703777336,
+      "grad_norm": 0.5992622107260112,
+      "learning_rate": 7.879850789753016e-05,
+      "loss": 0.5819,
+      "step": 429
+    },
+    {
+      "epoch": 0.8548707753479126,
+      "grad_norm": 0.6800607847366943,
+      "learning_rate": 7.878496271050502e-05,
+      "loss": 0.5828,
+      "step": 430
+    },
+    {
+      "epoch": 0.856858846918489,
+      "grad_norm": 0.7499052373458801,
+      "learning_rate": 7.877134277644905e-05,
+      "loss": 0.5807,
+      "step": 431
+    },
+    {
+      "epoch": 0.8588469184890656,
+      "grad_norm": 0.9057195016780537,
+      "learning_rate": 7.875764812161079e-05,
+      "loss": 0.5694,
+      "step": 432
+    },
+    {
+      "epoch": 0.8608349900596421,
+      "grad_norm": 1.0520429016398172,
+      "learning_rate": 7.874387877238283e-05,
+      "loss": 0.5793,
+      "step": 433
+    },
+    {
+      "epoch": 0.8628230616302187,
+      "grad_norm": 1.0753300240675914,
+      "learning_rate": 7.873003475530169e-05,
+      "loss": 0.5835,
+      "step": 434
+    },
+    {
+      "epoch": 0.8648111332007953,
+      "grad_norm": 1.1933706500073453,
+      "learning_rate": 7.871611609704777e-05,
+      "loss": 0.5795,
+      "step": 435
+    },
+    {
+      "epoch": 0.8667992047713717,
+      "grad_norm": 0.9800602567078205,
+      "learning_rate": 7.870212282444536e-05,
+      "loss": 0.5777,
+      "step": 436
+    },
+    {
+      "epoch": 0.8687872763419483,
+      "grad_norm": 0.985646291001119,
+      "learning_rate": 7.868805496446254e-05,
+      "loss": 0.5825,
+      "step": 437
+    },
+    {
+      "epoch": 0.8707753479125249,
+      "grad_norm": 0.9173028079338591,
+      "learning_rate": 7.86739125442111e-05,
+      "loss": 0.5826,
+      "step": 438
+    },
+    {
+      "epoch": 0.8727634194831014,
+      "grad_norm": 0.8343252949496813,
+      "learning_rate": 7.865969559094657e-05,
+      "loss": 0.5765,
+      "step": 439
+    },
+    {
+      "epoch": 0.8747514910536779,
+      "grad_norm": 0.9011659813788296,
+      "learning_rate": 7.864540413206809e-05,
+      "loss": 0.5807,
+      "step": 440
+    },
+    {
+      "epoch": 0.8767395626242545,
+      "grad_norm": 1.1481938792997213,
+      "learning_rate": 7.863103819511841e-05,
+      "loss": 0.5756,
+      "step": 441
+    },
+    {
+      "epoch": 0.878727634194831,
+      "grad_norm": 0.8692225637264639,
+      "learning_rate": 7.861659780778379e-05,
+      "loss": 0.5797,
+      "step": 442
+    },
+    {
+      "epoch": 0.8807157057654076,
+      "grad_norm": 0.7792876237320553,
+      "learning_rate": 7.860208299789399e-05,
+      "loss": 0.5789,
+      "step": 443
+    },
+    {
+      "epoch": 0.882703777335984,
+      "grad_norm": 0.9806726470334519,
+      "learning_rate": 7.85874937934222e-05,
+      "loss": 0.572,
+      "step": 444
+    },
+    {
+      "epoch": 0.8846918489065606,
+      "grad_norm": 1.1128801316838084,
+      "learning_rate": 7.857283022248499e-05,
+      "loss": 0.5771,
+      "step": 445
+    },
+    {
+      "epoch": 0.8866799204771372,
+      "grad_norm": 0.7356169677900156,
+      "learning_rate": 7.855809231334222e-05,
+      "loss": 0.5706,
+      "step": 446
+    },
+    {
+      "epoch": 0.8886679920477137,
+      "grad_norm": 0.5716244438538673,
+      "learning_rate": 7.854328009439704e-05,
+      "loss": 0.5788,
+      "step": 447
+    },
+    {
+      "epoch": 0.8906560636182903,
+      "grad_norm": 0.7756825473319496,
+      "learning_rate": 7.852839359419583e-05,
+      "loss": 0.5726,
+      "step": 448
+    },
+    {
+      "epoch": 0.8926441351888668,
+      "grad_norm": 0.9597859189328001,
+      "learning_rate": 7.851343284142806e-05,
+      "loss": 0.5793,
+      "step": 449
+    },
+    {
+      "epoch": 0.8946322067594433,
+      "grad_norm": 0.9840514136250063,
+      "learning_rate": 7.849839786492638e-05,
+      "loss": 0.5816,
+      "step": 450
+    },
+    {
+      "epoch": 0.8966202783300199,
+      "grad_norm": 1.2387344205054815,
+      "learning_rate": 7.848328869366644e-05,
+      "loss": 0.5777,
+      "step": 451
+    },
+    {
+      "epoch": 0.8986083499005965,
+      "grad_norm": 0.9408504441388281,
+      "learning_rate": 7.846810535676687e-05,
+      "loss": 0.5686,
+      "step": 452
+    },
+    {
+      "epoch": 0.9005964214711729,
+      "grad_norm": 0.8661605445202164,
+      "learning_rate": 7.845284788348928e-05,
+      "loss": 0.5756,
+      "step": 453
+    },
+    {
+      "epoch": 0.9025844930417495,
+      "grad_norm": 0.8836259932861911,
+      "learning_rate": 7.843751630323811e-05,
+      "loss": 0.5754,
+      "step": 454
+    },
+    {
+      "epoch": 0.904572564612326,
+      "grad_norm": 0.7484011259578548,
+      "learning_rate": 7.842211064556065e-05,
+      "loss": 0.5747,
+      "step": 455
+    },
+    {
+      "epoch": 0.9065606361829026,
+      "grad_norm": 0.5916896881610786,
+      "learning_rate": 7.840663094014692e-05,
+      "loss": 0.5753,
+      "step": 456
+    },
+    {
+      "epoch": 0.9085487077534792,
+      "grad_norm": 0.6446171878161961,
+      "learning_rate": 7.839107721682971e-05,
+      "loss": 0.5758,
+      "step": 457
+    },
+    {
+      "epoch": 0.9105367793240556,
+      "grad_norm": 0.7331317560231023,
+      "learning_rate": 7.837544950558437e-05,
+      "loss": 0.5767,
+      "step": 458
+    },
+    {
+      "epoch": 0.9125248508946322,
+      "grad_norm": 0.7169698033515578,
+      "learning_rate": 7.835974783652892e-05,
+      "loss": 0.5774,
+      "step": 459
+    },
+    {
+      "epoch": 0.9145129224652088,
+      "grad_norm": 0.6975707545239594,
+      "learning_rate": 7.834397223992387e-05,
+      "loss": 0.5742,
+      "step": 460
+    },
+    {
+      "epoch": 0.9165009940357853,
+      "grad_norm": 0.7873223472670946,
+      "learning_rate": 7.83281227461722e-05,
+      "loss": 0.5752,
+      "step": 461
+    },
+    {
+      "epoch": 0.9184890656063618,
+      "grad_norm": 0.8677658380277969,
+      "learning_rate": 7.831219938581934e-05,
+      "loss": 0.5799,
+      "step": 462
+    },
+    {
+      "epoch": 0.9204771371769384,
+      "grad_norm": 0.8564280021238664,
+      "learning_rate": 7.829620218955303e-05,
+      "loss": 0.5764,
+      "step": 463
+    },
+    {
+      "epoch": 0.9224652087475149,
+      "grad_norm": 0.8770540530088526,
+      "learning_rate": 7.828013118820338e-05,
+      "loss": 0.5825,
+      "step": 464
+    },
+    {
+      "epoch": 0.9244532803180915,
+      "grad_norm": 0.8775401382866653,
+      "learning_rate": 7.826398641274266e-05,
+      "loss": 0.5743,
+      "step": 465
+    },
+    {
+      "epoch": 0.9264413518886679,
+      "grad_norm": 0.8396168474705734,
+      "learning_rate": 7.824776789428536e-05,
+      "loss": 0.5798,
+      "step": 466
+    },
+    {
+      "epoch": 0.9284294234592445,
+      "grad_norm": 0.6451882476229347,
+      "learning_rate": 7.823147566408808e-05,
+      "loss": 0.5739,
+      "step": 467
+    },
+    {
+      "epoch": 0.9304174950298211,
+      "grad_norm": 0.5473862955414479,
+      "learning_rate": 7.821510975354948e-05,
+      "loss": 0.572,
+      "step": 468
+    },
+    {
+      "epoch": 0.9324055666003976,
+      "grad_norm": 0.7334994329755088,
+      "learning_rate": 7.819867019421023e-05,
+      "loss": 0.5782,
+      "step": 469
+    },
+    {
+      "epoch": 0.9343936381709742,
+      "grad_norm": 0.8493149045492733,
+      "learning_rate": 7.818215701775289e-05,
+      "loss": 0.5691,
+      "step": 470
+    },
+    {
+      "epoch": 0.9363817097415507,
+      "grad_norm": 0.8916093585824147,
+      "learning_rate": 7.816557025600196e-05,
+      "loss": 0.5737,
+      "step": 471
+    },
+    {
+      "epoch": 0.9383697813121272,
+      "grad_norm": 1.0524874978245338,
+      "learning_rate": 7.814890994092371e-05,
+      "loss": 0.5783,
+      "step": 472
+    },
+    {
+      "epoch": 0.9403578528827038,
+      "grad_norm": 1.1448842494252496,
+      "learning_rate": 7.813217610462622e-05,
+      "loss": 0.5814,
+      "step": 473
+    },
+    {
+      "epoch": 0.9423459244532804,
+      "grad_norm": 0.8875324561425191,
+      "learning_rate": 7.811536877935916e-05,
+      "loss": 0.5692,
+      "step": 474
+    },
+    {
+      "epoch": 0.9443339960238568,
+      "grad_norm": 0.7862741860973999,
+      "learning_rate": 7.809848799751392e-05,
+      "loss": 0.579,
+      "step": 475
+    },
+    {
+      "epoch": 0.9463220675944334,
+      "grad_norm": 0.7404744964609349,
+      "learning_rate": 7.808153379162344e-05,
+      "loss": 0.5724,
+      "step": 476
+    },
+    {
+      "epoch": 0.94831013916501,
+      "grad_norm": 0.6210045203434688,
+      "learning_rate": 7.806450619436212e-05,
+      "loss": 0.5778,
+      "step": 477
+    },
+    {
+      "epoch": 0.9502982107355865,
+      "grad_norm": 0.5869709000454426,
+      "learning_rate": 7.804740523854584e-05,
+      "loss": 0.5707,
+      "step": 478
+    },
+    {
+      "epoch": 0.952286282306163,
+      "grad_norm": 0.690983672156381,
+      "learning_rate": 7.803023095713185e-05,
+      "loss": 0.5673,
+      "step": 479
+    },
+    {
+      "epoch": 0.9542743538767395,
+      "grad_norm": 0.7769763991531466,
+      "learning_rate": 7.801298338321873e-05,
+      "loss": 0.5716,
+      "step": 480
+    },
+    {
+      "epoch": 0.9562624254473161,
+      "grad_norm": 0.7773730859463129,
+      "learning_rate": 7.799566255004627e-05,
+      "loss": 0.5769,
+      "step": 481
+    },
+    {
+      "epoch": 0.9582504970178927,
+      "grad_norm": 0.6345198639510748,
+      "learning_rate": 7.797826849099549e-05,
+      "loss": 0.5707,
+      "step": 482
+    },
+    {
+      "epoch": 0.9602385685884692,
+      "grad_norm": 0.6588045885859571,
+      "learning_rate": 7.796080123958849e-05,
+      "loss": 0.5717,
+      "step": 483
+    },
+    {
+      "epoch": 0.9622266401590457,
+      "grad_norm": 0.8039607210702434,
+      "learning_rate": 7.794326082948847e-05,
+      "loss": 0.5779,
+      "step": 484
+    },
+    {
+      "epoch": 0.9642147117296223,
+      "grad_norm": 0.7648474063784596,
+      "learning_rate": 7.79256472944996e-05,
+      "loss": 0.5747,
+      "step": 485
+    },
+    {
+      "epoch": 0.9662027833001988,
+      "grad_norm": 0.9891199051114163,
+      "learning_rate": 7.790796066856695e-05,
+      "loss": 0.5761,
+      "step": 486
+    },
+    {
+      "epoch": 0.9681908548707754,
+      "grad_norm": 1.4830791374433934,
+      "learning_rate": 7.789020098577654e-05,
+      "loss": 0.578,
+      "step": 487
+    },
+    {
+      "epoch": 0.9701789264413518,
+      "grad_norm": 0.4001636007612993,
+      "learning_rate": 7.787236828035508e-05,
+      "loss": 0.565,
+      "step": 488
+    },
+    {
+      "epoch": 0.9721669980119284,
+      "grad_norm": 1.391894017796059,
+      "learning_rate": 7.785446258667006e-05,
+      "loss": 0.5728,
+      "step": 489
+    },
+    {
+      "epoch": 0.974155069582505,
+      "grad_norm": 0.8474494313233529,
+      "learning_rate": 7.783648393922968e-05,
+      "loss": 0.5755,
+      "step": 490
+    },
+    {
+      "epoch": 0.9761431411530815,
+      "grad_norm": 0.9368946226000615,
+      "learning_rate": 7.781843237268265e-05,
+      "loss": 0.5772,
+      "step": 491
+    },
+    {
+      "epoch": 0.9781312127236581,
+      "grad_norm": 1.0296054222965858,
+      "learning_rate": 7.780030792181826e-05,
+      "loss": 0.5754,
+      "step": 492
+    },
+    {
+      "epoch": 0.9801192842942346,
+      "grad_norm": 0.8078527279993658,
+      "learning_rate": 7.778211062156625e-05,
+      "loss": 0.5734,
+      "step": 493
+    },
+    {
+      "epoch": 0.9821073558648111,
+      "grad_norm": 0.8032865048136967,
+      "learning_rate": 7.776384050699677e-05,
+      "loss": 0.5722,
+      "step": 494
+    },
+    {
+      "epoch": 0.9840954274353877,
+      "grad_norm": 0.7330125115532979,
+      "learning_rate": 7.77454976133203e-05,
+      "loss": 0.5727,
+      "step": 495
+    },
+    {
+      "epoch": 0.9860834990059643,
+      "grad_norm": 0.5748768568400435,
+      "learning_rate": 7.772708197588755e-05,
+      "loss": 0.5814,
+      "step": 496
+    },
+    {
+      "epoch": 0.9880715705765407,
+      "grad_norm": 0.5663216889025084,
+      "learning_rate": 7.770859363018947e-05,
+      "loss": 0.575,
+      "step": 497
+    },
+    {
+      "epoch": 0.9900596421471173,
+      "grad_norm": 0.5130324016109733,
+      "learning_rate": 7.76900326118571e-05,
+      "loss": 0.5724,
+      "step": 498
+    },
+    {
+      "epoch": 0.9920477137176938,
+      "grad_norm": 0.6118536140979886,
+      "learning_rate": 7.767139895666154e-05,
+      "loss": 0.5741,
+      "step": 499
+    },
+    {
+      "epoch": 0.9940357852882704,
+      "grad_norm": 0.7903621371505978,
+      "learning_rate": 7.765269270051389e-05,
+      "loss": 0.5699,
+      "step": 500
+    },
+    {
+      "epoch": 0.9960238568588469,
+      "grad_norm": 0.9431939811737229,
+      "learning_rate": 7.763391387946515e-05,
+      "loss": 0.5752,
+      "step": 501
+    },
+    {
+      "epoch": 0.9980119284294234,
+      "grad_norm": 1.060145724961572,
+      "learning_rate": 7.76150625297062e-05,
+      "loss": 0.5766,
+      "step": 502
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.949801921673401,
+      "learning_rate": 7.759613868756766e-05,
+      "loss": 0.5741,
+      "step": 503
+    },
+    {
+      "epoch": 1.0019880715705765,
+      "grad_norm": 0.9518658745713702,
+      "learning_rate": 7.757714238951987e-05,
+      "loss": 0.5507,
+      "step": 504
+    },
+    {
+      "epoch": 1.0039761431411531,
+      "grad_norm": 0.9224616680549119,
+      "learning_rate": 7.755807367217283e-05,
+      "loss": 0.5472,
+      "step": 505
+    },
+    {
+      "epoch": 1.0059642147117296,
+      "grad_norm": 0.9467697886335441,
+      "learning_rate": 7.75389325722761e-05,
+      "loss": 0.5397,
+      "step": 506
+    },
+    {
+      "epoch": 1.0079522862823063,
+      "grad_norm": 1.0032787528038696,
+      "learning_rate": 7.751971912671872e-05,
+      "loss": 0.5485,
+      "step": 507
+    },
+    {
+      "epoch": 1.0099403578528827,
+      "grad_norm": 1.0617180843326368,
+      "learning_rate": 7.750043337252915e-05,
+      "loss": 0.5486,
+      "step": 508
+    },
+    {
+      "epoch": 1.0119284294234592,
+      "grad_norm": 0.8345796460493181,
+      "learning_rate": 7.748107534687522e-05,
+      "loss": 0.5412,
+      "step": 509
+    },
+    {
+      "epoch": 1.0139165009940359,
+      "grad_norm": 0.5834246870142797,
+      "learning_rate": 7.746164508706406e-05,
+      "loss": 0.5452,
+      "step": 510
+    },
+    {
+      "epoch": 1.0159045725646123,
+      "grad_norm": 0.6308771152369433,
+      "learning_rate": 7.744214263054197e-05,
+      "loss": 0.5402,
+      "step": 511
+    },
+    {
+      "epoch": 1.0178926441351888,
+      "grad_norm": 0.8463052045590627,
+      "learning_rate": 7.742256801489446e-05,
+      "loss": 0.5479,
+      "step": 512
+    },
+    {
+      "epoch": 1.0198807157057654,
+      "grad_norm": 0.7841972760187628,
+      "learning_rate": 7.740292127784599e-05,
+      "loss": 0.5373,
+      "step": 513
+    },
+    {
+      "epoch": 1.021868787276342,
+      "grad_norm": 0.7132205161493853,
+      "learning_rate": 7.738320245726013e-05,
+      "loss": 0.5393,
+      "step": 514
+    },
+    {
+      "epoch": 1.0238568588469186,
+      "grad_norm": 0.7241903378381845,
+      "learning_rate": 7.736341159113932e-05,
+      "loss": 0.5468,
+      "step": 515
+    },
+    {
+      "epoch": 1.025844930417495,
+      "grad_norm": 0.6745690903906133,
+      "learning_rate": 7.734354871762481e-05,
+      "loss": 0.5443,
+      "step": 516
+    },
+    {
+      "epoch": 1.0278330019880715,
+      "grad_norm": 0.6972843853390581,
+      "learning_rate": 7.73236138749967e-05,
+      "loss": 0.544,
+      "step": 517
+    },
+    {
+      "epoch": 1.0298210735586482,
+      "grad_norm": 0.7989595482697651,
+      "learning_rate": 7.730360710167373e-05,
+      "loss": 0.5392,
+      "step": 518
+    },
+    {
+      "epoch": 1.0318091451292246,
+      "grad_norm": 0.9023867022544568,
+      "learning_rate": 7.728352843621332e-05,
+      "loss": 0.5423,
+      "step": 519
+    },
+    {
+      "epoch": 1.0337972166998013,
+      "grad_norm": 0.924903270385559,
+      "learning_rate": 7.726337791731139e-05,
+      "loss": 0.5406,
+      "step": 520
+    },
+    {
+      "epoch": 1.0357852882703777,
+      "grad_norm": 0.7157014642375783,
+      "learning_rate": 7.724315558380234e-05,
+      "loss": 0.5505,
+      "step": 521
+    },
+    {
+      "epoch": 1.0377733598409542,
+      "grad_norm": 0.5321338664276257,
+      "learning_rate": 7.722286147465902e-05,
+      "loss": 0.5407,
+      "step": 522
+    },
+    {
+      "epoch": 1.0397614314115309,
+      "grad_norm": 0.62645163866993,
+      "learning_rate": 7.720249562899257e-05,
+      "loss": 0.5454,
+      "step": 523
+    },
+    {
+      "epoch": 1.0417495029821073,
+      "grad_norm": 0.7501732694460348,
+      "learning_rate": 7.71820580860524e-05,
+      "loss": 0.5454,
+      "step": 524
+    },
+    {
+      "epoch": 1.0437375745526838,
+      "grad_norm": 0.9392351876663441,
+      "learning_rate": 7.716154888522606e-05,
+      "loss": 0.548,
+      "step": 525
+    },
+    {
+      "epoch": 1.0457256461232605,
+      "grad_norm": 1.1654586518995889,
+      "learning_rate": 7.714096806603925e-05,
+      "loss": 0.5436,
+      "step": 526
+    },
+    {
+      "epoch": 1.047713717693837,
+      "grad_norm": 0.7430639246487083,
+      "learning_rate": 7.712031566815564e-05,
+      "loss": 0.5435,
+      "step": 527
+    },
+    {
+      "epoch": 1.0497017892644136,
+      "grad_norm": 0.507332577185216,
+      "learning_rate": 7.709959173137691e-05,
+      "loss": 0.5459,
+      "step": 528
+    },
+    {
+      "epoch": 1.05168986083499,
+      "grad_norm": 0.47161280639381553,
+      "learning_rate": 7.707879629564258e-05,
+      "loss": 0.537,
+      "step": 529
+    },
+    {
+      "epoch": 1.0536779324055665,
+      "grad_norm": 0.49032557450749087,
+      "learning_rate": 7.705792940102992e-05,
+      "loss": 0.5354,
+      "step": 530
+    },
+    {
+      "epoch": 1.0556660039761432,
+      "grad_norm": 0.5367736352839478,
+      "learning_rate": 7.7036991087754e-05,
+      "loss": 0.5424,
+      "step": 531
+    },
+    {
+      "epoch": 1.0576540755467196,
+      "grad_norm": 0.5197894411370604,
+      "learning_rate": 7.701598139616748e-05,
+      "loss": 0.5452,
+      "step": 532
+    },
+    {
+      "epoch": 1.0596421471172963,
+      "grad_norm": 0.4882004191720187,
+      "learning_rate": 7.699490036676057e-05,
+      "loss": 0.5438,
+      "step": 533
+    },
+    {
+      "epoch": 1.0616302186878728,
+      "grad_norm": 0.5138919640591607,
+      "learning_rate": 7.697374804016102e-05,
+      "loss": 0.5419,
+      "step": 534
+    },
+    {
+      "epoch": 1.0636182902584492,
+      "grad_norm": 0.5587132412348711,
+      "learning_rate": 7.695252445713392e-05,
+      "loss": 0.5436,
+      "step": 535
+    },
+    {
+      "epoch": 1.065606361829026,
+      "grad_norm": 0.6000844383193432,
+      "learning_rate": 7.693122965858173e-05,
+      "loss": 0.5398,
+      "step": 536
+    },
+    {
+      "epoch": 1.0675944333996024,
+      "grad_norm": 0.7480777517729358,
+      "learning_rate": 7.690986368554414e-05,
+      "loss": 0.5411,
+      "step": 537
+    },
+    {
+      "epoch": 1.069582504970179,
+      "grad_norm": 0.9742584925928898,
+      "learning_rate": 7.688842657919799e-05,
+      "loss": 0.5398,
+      "step": 538
+    },
+    {
+      "epoch": 1.0715705765407555,
+      "grad_norm": 1.1824603415781014,
+      "learning_rate": 7.686691838085726e-05,
+      "loss": 0.5472,
+      "step": 539
+    },
+    {
+      "epoch": 1.073558648111332,
+      "grad_norm": 0.7779661582778407,
+      "learning_rate": 7.68453391319729e-05,
+      "loss": 0.5398,
+      "step": 540
+    },
+    {
+      "epoch": 1.0755467196819086,
+      "grad_norm": 0.660572845944423,
+      "learning_rate": 7.682368887413279e-05,
+      "loss": 0.546,
+      "step": 541
+    },
+    {
+      "epoch": 1.077534791252485,
+      "grad_norm": 0.647498839725535,
+      "learning_rate": 7.68019676490617e-05,
+      "loss": 0.5399,
+      "step": 542
+    },
+    {
+      "epoch": 1.0795228628230615,
+      "grad_norm": 0.5616817097968095,
+      "learning_rate": 7.678017549862111e-05,
+      "loss": 0.5414,
+      "step": 543
+    },
+    {
+      "epoch": 1.0815109343936382,
+      "grad_norm": 0.5077349122774221,
+      "learning_rate": 7.675831246480923e-05,
+      "loss": 0.537,
+      "step": 544
+    },
+    {
+      "epoch": 1.0834990059642147,
+      "grad_norm": 0.6460547464386392,
+      "learning_rate": 7.673637858976086e-05,
+      "loss": 0.5374,
+      "step": 545
+    },
+    {
+      "epoch": 1.0854870775347913,
+      "grad_norm": 0.7566165062457662,
+      "learning_rate": 7.671437391574735e-05,
+      "loss": 0.5454,
+      "step": 546
+    },
+    {
+      "epoch": 1.0874751491053678,
+      "grad_norm": 0.8765229126878328,
+      "learning_rate": 7.669229848517645e-05,
+      "loss": 0.5362,
+      "step": 547
+    },
+    {
+      "epoch": 1.0894632206759443,
+      "grad_norm": 0.9759995942905503,
+      "learning_rate": 7.667015234059231e-05,
+      "loss": 0.5489,
+      "step": 548
+    },
+    {
+      "epoch": 1.091451292246521,
+      "grad_norm": 1.0219449344079592,
+      "learning_rate": 7.664793552467537e-05,
+      "loss": 0.5501,
+      "step": 549
+    },
+    {
+      "epoch": 1.0934393638170974,
+      "grad_norm": 1.049500455165683,
+      "learning_rate": 7.662564808024221e-05,
+      "loss": 0.541,
+      "step": 550
+    },
+    {
+      "epoch": 1.0954274353876738,
+      "grad_norm": 1.040451711661253,
+      "learning_rate": 7.660329005024562e-05,
+      "loss": 0.5466,
+      "step": 551
+    },
+    {
+      "epoch": 1.0974155069582505,
+      "grad_norm": 0.9805330653702803,
+      "learning_rate": 7.658086147777433e-05,
+      "loss": 0.5447,
+      "step": 552
+    },
+    {
+      "epoch": 1.099403578528827,
+      "grad_norm": 0.7595556727657715,
+      "learning_rate": 7.655836240605306e-05,
+      "loss": 0.5425,
+      "step": 553
+    },
+    {
+      "epoch": 1.1013916500994037,
+      "grad_norm": 0.5629886257792873,
+      "learning_rate": 7.653579287844242e-05,
+      "loss": 0.5404,
+      "step": 554
+    },
+    {
+      "epoch": 1.10337972166998,
+      "grad_norm": 0.6066304118171499,
+      "learning_rate": 7.651315293843879e-05,
+      "loss": 0.5383,
+      "step": 555
+    },
+    {
+      "epoch": 1.1053677932405566,
+      "grad_norm": 0.7504168630502673,
+      "learning_rate": 7.649044262967421e-05,
+      "loss": 0.5401,
+      "step": 556
+    },
+    {
+      "epoch": 1.1073558648111332,
+      "grad_norm": 0.8672288992174086,
+      "learning_rate": 7.646766199591639e-05,
+      "loss": 0.5408,
+      "step": 557
+    },
+    {
+      "epoch": 1.1093439363817097,
+      "grad_norm": 0.7124336695406922,
+      "learning_rate": 7.644481108106853e-05,
+      "loss": 0.5379,
+      "step": 558
+    },
+    {
+      "epoch": 1.1113320079522864,
+      "grad_norm": 0.4827659980230668,
+      "learning_rate": 7.642188992916933e-05,
+      "loss": 0.5406,
+      "step": 559
+    },
+    {
+      "epoch": 1.1133200795228628,
+      "grad_norm": 0.4284195546204533,
+      "learning_rate": 7.639889858439279e-05,
+      "loss": 0.5408,
+      "step": 560
+    },
+    {
+      "epoch": 1.1153081510934393,
+      "grad_norm": 0.5213339666965665,
+      "learning_rate": 7.637583709104822e-05,
+      "loss": 0.5404,
+      "step": 561
+    },
+    {
+      "epoch": 1.117296222664016,
+      "grad_norm": 0.4100543308356987,
+      "learning_rate": 7.635270549358012e-05,
+      "loss": 0.5398,
+      "step": 562
+    },
+    {
+      "epoch": 1.1192842942345924,
+      "grad_norm": 0.4894280047781026,
+      "learning_rate": 7.63295038365681e-05,
+      "loss": 0.5339,
+      "step": 563
+    },
+    {
+      "epoch": 1.121272365805169,
+      "grad_norm": 0.6660381872451082,
+      "learning_rate": 7.630623216472677e-05,
+      "loss": 0.5382,
+      "step": 564
+    },
+    {
+      "epoch": 1.1232604373757455,
+      "grad_norm": 0.7384209893024911,
+      "learning_rate": 7.62828905229057e-05,
+      "loss": 0.5462,
+      "step": 565
+    },
+    {
+      "epoch": 1.125248508946322,
+      "grad_norm": 0.8439530140670745,
+      "learning_rate": 7.62594789560893e-05,
+      "loss": 0.54,
+      "step": 566
+    },
+    {
+      "epoch": 1.1272365805168987,
+      "grad_norm": 0.9437603370716825,
+      "learning_rate": 7.623599750939669e-05,
+      "loss": 0.5428,
+      "step": 567
+    },
+    {
+      "epoch": 1.1292246520874751,
+      "grad_norm": 0.9838775457498312,
+      "learning_rate": 7.621244622808176e-05,
+      "loss": 0.5417,
+      "step": 568
+    },
+    {
+      "epoch": 1.1312127236580518,
+      "grad_norm": 0.9050007349320468,
+      "learning_rate": 7.61888251575329e-05,
+      "loss": 0.5331,
+      "step": 569
+    },
+    {
+      "epoch": 1.1332007952286283,
+      "grad_norm": 0.6124525965371876,
+      "learning_rate": 7.616513434327306e-05,
+      "loss": 0.5443,
+      "step": 570
+    },
+    {
+      "epoch": 1.1351888667992047,
+      "grad_norm": 0.4152960954671294,
+      "learning_rate": 7.614137383095956e-05,
+      "loss": 0.5399,
+      "step": 571
+    },
+    {
+      "epoch": 1.1371769383697814,
+      "grad_norm": 0.5708141529701758,
+      "learning_rate": 7.611754366638405e-05,
+      "loss": 0.5429,
+      "step": 572
+    },
+    {
+      "epoch": 1.1391650099403579,
+      "grad_norm": 0.5984148199496552,
+      "learning_rate": 7.609364389547244e-05,
+      "loss": 0.5391,
+      "step": 573
+    },
+    {
+      "epoch": 1.1411530815109343,
+      "grad_norm": 0.4907009934797348,
+      "learning_rate": 7.606967456428477e-05,
+      "loss": 0.5387,
+      "step": 574
+    },
+    {
+      "epoch": 1.143141153081511,
+      "grad_norm": 0.4557259695946618,
+      "learning_rate": 7.60456357190151e-05,
+      "loss": 0.5362,
+      "step": 575
+    },
+    {
+      "epoch": 1.1451292246520874,
+      "grad_norm": 0.6032487760061391,
+      "learning_rate": 7.602152740599156e-05,
+      "loss": 0.5386,
+      "step": 576
+    },
+    {
+      "epoch": 1.147117296222664,
+      "grad_norm": 0.9574111490642987,
+      "learning_rate": 7.599734967167604e-05,
+      "loss": 0.5415,
+      "step": 577
+    },
+    {
+      "epoch": 1.1491053677932406,
+      "grad_norm": 1.2259900507384356,
+      "learning_rate": 7.59731025626643e-05,
+      "loss": 0.5439,
+      "step": 578
+    },
+    {
+      "epoch": 1.151093439363817,
+      "grad_norm": 0.5769700449630275,
+      "learning_rate": 7.594878612568577e-05,
+      "loss": 0.5366,
+      "step": 579
+    },
+    {
+      "epoch": 1.1530815109343937,
+      "grad_norm": 0.5807398397185227,
+      "learning_rate": 7.59244004076035e-05,
+      "loss": 0.5361,
+      "step": 580
+    },
+    {
+      "epoch": 1.1550695825049702,
+      "grad_norm": 0.9087659346410144,
+      "learning_rate": 7.589994545541406e-05,
+      "loss": 0.5465,
+      "step": 581
+    },
+    {
+      "epoch": 1.1570576540755466,
+      "grad_norm": 1.0866888064181934,
+      "learning_rate": 7.587542131624741e-05,
+      "loss": 0.5393,
+      "step": 582
+    },
+    {
+      "epoch": 1.1590457256461233,
+      "grad_norm": 0.8604614363303615,
+      "learning_rate": 7.585082803736692e-05,
+      "loss": 0.5394,
+      "step": 583
+    },
+    {
+      "epoch": 1.1610337972166997,
+      "grad_norm": 0.767814178995087,
+      "learning_rate": 7.582616566616917e-05,
+      "loss": 0.5358,
+      "step": 584
+    },
+    {
+      "epoch": 1.1630218687872764,
+      "grad_norm": 0.7506490371579211,
+      "learning_rate": 7.580143425018387e-05,
+      "loss": 0.5445,
+      "step": 585
+    },
+    {
+      "epoch": 1.1650099403578529,
+      "grad_norm": 0.743704401144008,
+      "learning_rate": 7.577663383707384e-05,
+      "loss": 0.5381,
+      "step": 586
+    },
+    {
+      "epoch": 1.1669980119284293,
+      "grad_norm": 0.586644447631041,
+      "learning_rate": 7.575176447463483e-05,
+      "loss": 0.5378,
+      "step": 587
+    },
+    {
+      "epoch": 1.168986083499006,
+      "grad_norm": 0.5352393673773019,
+      "learning_rate": 7.572682621079551e-05,
+      "loss": 0.5368,
+      "step": 588
+    },
+    {
+      "epoch": 1.1709741550695825,
+      "grad_norm": 0.5344762767308605,
+      "learning_rate": 7.570181909361733e-05,
+      "loss": 0.5394,
+      "step": 589
+    },
+    {
+      "epoch": 1.1729622266401591,
+      "grad_norm": 0.5395614621565161,
+      "learning_rate": 7.567674317129441e-05,
+      "loss": 0.5387,
+      "step": 590
+    },
+    {
+      "epoch": 1.1749502982107356,
+      "grad_norm": 0.575796280722566,
+      "learning_rate": 7.56515984921535e-05,
+      "loss": 0.5354,
+      "step": 591
+    },
+    {
+      "epoch": 1.176938369781312,
+      "grad_norm": 0.5915049989827494,
+      "learning_rate": 7.562638510465382e-05,
+      "loss": 0.5425,
+      "step": 592
+    },
+    {
+      "epoch": 1.1789264413518887,
+      "grad_norm": 0.5412929415802115,
+      "learning_rate": 7.560110305738708e-05,
+      "loss": 0.5414,
+      "step": 593
+    },
+    {
+      "epoch": 1.1809145129224652,
+      "grad_norm": 0.48254434256782,
+      "learning_rate": 7.557575239907721e-05,
+      "loss": 0.5372,
+      "step": 594
+    },
+    {
+      "epoch": 1.1829025844930419,
+      "grad_norm": 0.5181528971463945,
+      "learning_rate": 7.555033317858048e-05,
+      "loss": 0.5397,
+      "step": 595
+    },
+    {
+      "epoch": 1.1848906560636183,
+      "grad_norm": 0.5263184354853958,
+      "learning_rate": 7.552484544488519e-05,
+      "loss": 0.5391,
+      "step": 596
+    },
+    {
+      "epoch": 1.1868787276341948,
+      "grad_norm": 0.4115690316834255,
+      "learning_rate": 7.549928924711175e-05,
+      "loss": 0.5317,
+      "step": 597
+    },
+    {
+      "epoch": 1.1888667992047715,
+      "grad_norm": 0.43537299584180583,
+      "learning_rate": 7.547366463451249e-05,
+      "loss": 0.5344,
+      "step": 598
+    },
+    {
+      "epoch": 1.190854870775348,
+      "grad_norm": 0.4995055021742822,
+      "learning_rate": 7.544797165647162e-05,
+      "loss": 0.5344,
+      "step": 599
+    },
+    {
+      "epoch": 1.1928429423459244,
+      "grad_norm": 0.5090345933385654,
+      "learning_rate": 7.542221036250503e-05,
+      "loss": 0.5306,
+      "step": 600
+    },
+    {
+      "epoch": 1.194831013916501,
+      "grad_norm": 1.1457859586008614,
+      "learning_rate": 7.539638080226035e-05,
+      "loss": 0.5395,
+      "step": 601
+    },
+    {
+      "epoch": 1.1968190854870775,
+      "grad_norm": 0.7643479173221415,
+      "learning_rate": 7.537048302551672e-05,
+      "loss": 0.5452,
+      "step": 602
+    },
+    {
+      "epoch": 1.1988071570576542,
+      "grad_norm": 2.7209885317474662,
+      "learning_rate": 7.534451708218481e-05,
+      "loss": 0.5903,
+      "step": 603
+    },
+    {
+      "epoch": 1.2007952286282306,
+      "grad_norm": 15.140365769874222,
+      "learning_rate": 7.531848302230657e-05,
+      "loss": 0.9038,
+      "step": 604
+    },
+    {
+      "epoch": 1.202783300198807,
+      "grad_norm": 1.0923219501841632,
+      "learning_rate": 7.529238089605531e-05,
+      "loss": 0.6039,
+      "step": 605
+    },
+    {
+      "epoch": 1.2047713717693838,
+      "grad_norm": 1.6894455257731635,
+      "learning_rate": 7.526621075373549e-05,
+      "loss": 0.5814,
+      "step": 606
+    },
+    {
+      "epoch": 1.2067594433399602,
+      "grad_norm": 0.9739200902278735,
+      "learning_rate": 7.523997264578262e-05,
+      "loss": 0.5805,
+      "step": 607
+    },
+    {
+      "epoch": 1.2087475149105367,
+      "grad_norm": 1.4388276351964948,
+      "learning_rate": 7.521366662276324e-05,
+      "loss": 0.5814,
+      "step": 608
+    },
+    {
+      "epoch": 1.2107355864811133,
+      "grad_norm": 0.9122412624447593,
+      "learning_rate": 7.518729273537474e-05,
+      "loss": 0.5831,
+      "step": 609
+    },
+    {
+      "epoch": 1.2127236580516898,
+      "grad_norm": 1.11468588542839,
+      "learning_rate": 7.516085103444535e-05,
+      "loss": 0.5831,
+      "step": 610
+    },
+    {
+      "epoch": 1.2147117296222665,
+      "grad_norm": 0.7927291919940254,
+      "learning_rate": 7.513434157093395e-05,
+      "loss": 0.5744,
+      "step": 611
+    },
+    {
+      "epoch": 1.216699801192843,
+      "grad_norm": 0.9231028210170642,
+      "learning_rate": 7.510776439593e-05,
+      "loss": 0.5649,
+      "step": 612
+    },
+    {
+      "epoch": 1.2186878727634194,
+      "grad_norm": 0.7854830072173685,
+      "learning_rate": 7.508111956065351e-05,
+      "loss": 0.5706,
+      "step": 613
+    },
+    {
+      "epoch": 1.220675944333996,
+      "grad_norm": 0.7486027346054723,
+      "learning_rate": 7.505440711645482e-05,
+      "loss": 0.5633,
+      "step": 614
+    },
+    {
+      "epoch": 1.2226640159045725,
+      "grad_norm": 0.6883033522680998,
+      "learning_rate": 7.502762711481464e-05,
+      "loss": 0.5653,
+      "step": 615
+    },
+    {
+      "epoch": 1.2246520874751492,
+      "grad_norm": 0.7389094070760283,
+      "learning_rate": 7.50007796073438e-05,
+      "loss": 0.5611,
+      "step": 616
+    },
+    {
+      "epoch": 1.2266401590457257,
+      "grad_norm": 0.8318535060569524,
+      "learning_rate": 7.497386464578329e-05,
+      "loss": 0.5619,
+      "step": 617
+    },
+    {
+      "epoch": 1.228628230616302,
+      "grad_norm": 3.438501053226825,
+      "learning_rate": 7.494688228200407e-05,
+      "loss": 0.5655,
+      "step": 618
+    },
+    {
+      "epoch": 1.2306163021868788,
+      "grad_norm": 1.3177086303181726,
+      "learning_rate": 7.4919832568007e-05,
+      "loss": 0.5725,
+      "step": 619
+    },
+    {
+      "epoch": 1.2326043737574552,
+      "grad_norm": 1.275711618972336,
+      "learning_rate": 7.489271555592274e-05,
+      "loss": 0.5619,
+      "step": 620
+    },
+    {
+      "epoch": 1.234592445328032,
+      "grad_norm": 0.816816158692742,
+      "learning_rate": 7.486553129801167e-05,
+      "loss": 0.5532,
+      "step": 621
+    },
+    {
+      "epoch": 1.2365805168986084,
+      "grad_norm": 1.0789526717800226,
+      "learning_rate": 7.483827984666372e-05,
+      "loss": 0.5636,
+      "step": 622
+    },
+    {
+      "epoch": 1.2385685884691848,
+      "grad_norm": 0.931570699652363,
+      "learning_rate": 7.481096125439836e-05,
+      "loss": 0.5586,
+      "step": 623
+    },
+    {
+      "epoch": 1.2405566600397615,
+      "grad_norm": 1.0353236647558606,
+      "learning_rate": 7.478357557386444e-05,
+      "loss": 0.5576,
+      "step": 624
+    },
+    {
+      "epoch": 1.242544731610338,
+      "grad_norm": 0.7121241024787056,
+      "learning_rate": 7.47561228578401e-05,
+      "loss": 0.5527,
+      "step": 625
+    },
+    {
+      "epoch": 1.2445328031809146,
+      "grad_norm": 0.9354960795229111,
+      "learning_rate": 7.472860315923269e-05,
+      "loss": 0.5563,
+      "step": 626
+    },
+    {
+      "epoch": 1.246520874751491,
+      "grad_norm": 0.6810399481449622,
+      "learning_rate": 7.47010165310786e-05,
+      "loss": 0.5484,
+      "step": 627
+    },
+    {
+      "epoch": 1.2485089463220675,
+      "grad_norm": 0.564339776191756,
+      "learning_rate": 7.467336302654327e-05,
+      "loss": 0.5464,
+      "step": 628
+    },
+    {
+      "epoch": 1.250497017892644,
+      "grad_norm": 0.6329153920308194,
+      "learning_rate": 7.4645642698921e-05,
+      "loss": 0.5541,
+      "step": 629
+    },
+    {
+      "epoch": 1.2524850894632207,
+      "grad_norm": 0.5058821553568681,
+      "learning_rate": 7.461785560163485e-05,
+      "loss": 0.546,
+      "step": 630
+    },
+    {
+      "epoch": 1.2544731610337971,
+      "grad_norm": 0.4873097467195202,
+      "learning_rate": 7.459000178823661e-05,
+      "loss": 0.5478,
+      "step": 631
+    },
+    {
+      "epoch": 1.2564612326043738,
+      "grad_norm": 0.36686120495912883,
+      "learning_rate": 7.456208131240658e-05,
+      "loss": 0.5501,
+      "step": 632
+    },
+    {
+      "epoch": 1.2584493041749503,
+      "grad_norm": 0.4802564106494995,
+      "learning_rate": 7.45340942279536e-05,
+      "loss": 0.5447,
+      "step": 633
+    },
+    {
+      "epoch": 1.2604373757455267,
+      "grad_norm": 0.5250721782990612,
+      "learning_rate": 7.450604058881481e-05,
+      "loss": 0.5447,
+      "step": 634
+    },
+    {
+      "epoch": 1.2624254473161034,
+      "grad_norm": 0.4944252793040313,
+      "learning_rate": 7.44779204490557e-05,
+      "loss": 0.5517,
+      "step": 635
+    },
+    {
+      "epoch": 1.2644135188866799,
+      "grad_norm": 0.719034766972675,
+      "learning_rate": 7.444973386286985e-05,
+      "loss": 0.5453,
+      "step": 636
+    },
+    {
+      "epoch": 1.2664015904572565,
+      "grad_norm": 0.41783092273177835,
+      "learning_rate": 7.442148088457894e-05,
+      "loss": 0.5445,
+      "step": 637
+    },
+    {
+      "epoch": 1.268389662027833,
+      "grad_norm": 0.553215774804116,
+      "learning_rate": 7.439316156863255e-05,
+      "loss": 0.5416,
+      "step": 638
+    },
+    {
+      "epoch": 1.2703777335984094,
+      "grad_norm": 0.5408797993525253,
+      "learning_rate": 7.436477596960817e-05,
+      "loss": 0.5497,
+      "step": 639
+    },
+    {
+      "epoch": 1.2723658051689861,
+      "grad_norm": 0.4548242204966512,
+      "learning_rate": 7.4336324142211e-05,
+      "loss": 0.5432,
+      "step": 640
+    },
+    {
+      "epoch": 1.2743538767395626,
+      "grad_norm": 0.39363033210078713,
+      "learning_rate": 7.430780614127387e-05,
+      "loss": 0.5427,
+      "step": 641
+    },
+    {
+      "epoch": 1.2763419483101393,
+      "grad_norm": 0.589540690704029,
+      "learning_rate": 7.427922202175715e-05,
+      "loss": 0.5427,
+      "step": 642
+    },
+    {
+      "epoch": 1.2783300198807157,
+      "grad_norm": 0.7312646026221609,
+      "learning_rate": 7.425057183874865e-05,
+      "loss": 0.5459,
+      "step": 643
+    },
+    {
+      "epoch": 1.2803180914512922,
+      "grad_norm": 0.8432430495180779,
+      "learning_rate": 7.422185564746347e-05,
+      "loss": 0.5516,
+      "step": 644
+    },
+    {
+      "epoch": 1.2823061630218688,
+      "grad_norm": 0.7907499027073033,
+      "learning_rate": 7.419307350324392e-05,
+      "loss": 0.5477,
+      "step": 645
+    },
+    {
+      "epoch": 1.2842942345924453,
+      "grad_norm": 0.7675777704728362,
+      "learning_rate": 7.416422546155947e-05,
+      "loss": 0.5477,
+      "step": 646
+    },
+    {
+      "epoch": 1.286282306163022,
+      "grad_norm": 0.7818952066954505,
+      "learning_rate": 7.41353115780065e-05,
+      "loss": 0.5445,
+      "step": 647
+    },
+    {
+      "epoch": 1.2882703777335984,
+      "grad_norm": 0.6321072667200449,
+      "learning_rate": 7.410633190830839e-05,
+      "loss": 0.5475,
+      "step": 648
+    },
+    {
+      "epoch": 1.2902584493041749,
+      "grad_norm": 0.5418717449116404,
+      "learning_rate": 7.407728650831519e-05,
+      "loss": 0.5381,
+      "step": 649
+    },
+    {
+      "epoch": 1.2922465208747516,
+      "grad_norm": 0.47733226736370155,
+      "learning_rate": 7.404817543400372e-05,
+      "loss": 0.5358,
+      "step": 650
+    },
+    {
+      "epoch": 1.294234592445328,
+      "grad_norm": 0.4087774811685782,
+      "learning_rate": 7.401899874147733e-05,
+      "loss": 0.5423,
+      "step": 651
+    },
+    {
+      "epoch": 1.2962226640159047,
+      "grad_norm": 0.34492073630409487,
+      "learning_rate": 7.39897564869658e-05,
+      "loss": 0.5377,
+      "step": 652
+    },
+    {
+      "epoch": 1.2982107355864811,
+      "grad_norm": 0.41364092660903046,
+      "learning_rate": 7.396044872682533e-05,
+      "loss": 0.5409,
+      "step": 653
+    },
+    {
+      "epoch": 1.3001988071570576,
+      "grad_norm": 0.49300751266085974,
+      "learning_rate": 7.393107551753829e-05,
+      "loss": 0.5418,
+      "step": 654
+    },
+    {
+      "epoch": 1.302186878727634,
+      "grad_norm": 0.5115889514144143,
+      "learning_rate": 7.390163691571325e-05,
+      "loss": 0.541,
+      "step": 655
+    },
+    {
+      "epoch": 1.3041749502982107,
+      "grad_norm": 0.4520722977521369,
+      "learning_rate": 7.387213297808477e-05,
+      "loss": 0.5364,
+      "step": 656
+    },
+    {
+      "epoch": 1.3061630218687874,
+      "grad_norm": 0.45730722103763394,
+      "learning_rate": 7.384256376151333e-05,
+      "loss": 0.5433,
+      "step": 657
+    },
+    {
+      "epoch": 1.3081510934393639,
+      "grad_norm": 0.35991381976270714,
+      "learning_rate": 7.381292932298524e-05,
+      "loss": 0.5393,
+      "step": 658
+    },
+    {
+      "epoch": 1.3101391650099403,
+      "grad_norm": 0.3715007427958041,
+      "learning_rate": 7.378322971961245e-05,
+      "loss": 0.5386,
+      "step": 659
+    },
+    {
+      "epoch": 1.3121272365805168,
+      "grad_norm": 0.5035889463103559,
+      "learning_rate": 7.375346500863256e-05,
+      "loss": 0.5355,
+      "step": 660
+    },
+    {
+      "epoch": 1.3141153081510935,
+      "grad_norm": 0.6845352231571457,
+      "learning_rate": 7.372363524740862e-05,
+      "loss": 0.5434,
+      "step": 661
+    },
+    {
+      "epoch": 1.31610337972167,
+      "grad_norm": 0.7310717499538159,
+      "learning_rate": 7.369374049342904e-05,
+      "loss": 0.531,
+      "step": 662
+    },
+    {
+      "epoch": 1.3180914512922466,
+      "grad_norm": 0.5741347090908184,
+      "learning_rate": 7.366378080430748e-05,
+      "loss": 0.5424,
+      "step": 663
+    },
+    {
+      "epoch": 1.320079522862823,
+      "grad_norm": 0.4353235586521284,
+      "learning_rate": 7.36337562377828e-05,
+      "loss": 0.5359,
+      "step": 664
+    },
+    {
+      "epoch": 1.3220675944333995,
+      "grad_norm": 0.33576960536141043,
+      "learning_rate": 7.360366685171878e-05,
+      "loss": 0.535,
+      "step": 665
+    },
+    {
+      "epoch": 1.3240556660039762,
+      "grad_norm": 0.28581021576787774,
+      "learning_rate": 7.357351270410424e-05,
+      "loss": 0.5337,
+      "step": 666
+    },
+    {
+      "epoch": 1.3260437375745526,
+      "grad_norm": 0.34667350546819015,
+      "learning_rate": 7.354329385305274e-05,
+      "loss": 0.5344,
+      "step": 667
+    },
+    {
+      "epoch": 1.3280318091451293,
+      "grad_norm": 0.38791896920198515,
+      "learning_rate": 7.351301035680256e-05,
+      "loss": 0.5336,
+      "step": 668
+    },
+    {
+      "epoch": 1.3300198807157058,
+      "grad_norm": 0.3445729798343968,
+      "learning_rate": 7.348266227371657e-05,
+      "loss": 0.5304,
+      "step": 669
+    },
+    {
+      "epoch": 1.3320079522862822,
+      "grad_norm": 0.36399698860400687,
+      "learning_rate": 7.34522496622821e-05,
+      "loss": 0.5377,
+      "step": 670
+    },
+    {
+      "epoch": 1.333996023856859,
+      "grad_norm": 0.4179520009930122,
+      "learning_rate": 7.342177258111085e-05,
+      "loss": 0.5345,
+      "step": 671
+    },
+    {
+      "epoch": 1.3359840954274353,
+      "grad_norm": 0.5056757926940009,
+      "learning_rate": 7.339123108893873e-05,
+      "loss": 0.5315,
+      "step": 672
+    },
+    {
+      "epoch": 1.337972166998012,
+      "grad_norm": 0.6568326228633579,
+      "learning_rate": 7.336062524462586e-05,
+      "loss": 0.5336,
+      "step": 673
+    },
+    {
+      "epoch": 1.3399602385685885,
+      "grad_norm": 0.8051581651906589,
+      "learning_rate": 7.332995510715633e-05,
+      "loss": 0.5321,
+      "step": 674
+    },
+    {
+      "epoch": 1.341948310139165,
+      "grad_norm": 1.0061718045293344,
+      "learning_rate": 7.329922073563812e-05,
+      "loss": 0.5389,
+      "step": 675
+    },
+    {
+      "epoch": 1.3439363817097416,
+      "grad_norm": 1.224198955636859,
+      "learning_rate": 7.326842218930303e-05,
+      "loss": 0.5323,
+      "step": 676
+    },
+    {
+      "epoch": 1.345924453280318,
+      "grad_norm": 0.6806294285463886,
+      "learning_rate": 7.323755952750654e-05,
+      "loss": 0.5358,
+      "step": 677
+    },
+    {
+      "epoch": 1.3479125248508947,
+      "grad_norm": 0.5216238484730512,
+      "learning_rate": 7.32066328097277e-05,
+      "loss": 0.535,
+      "step": 678
+    },
+    {
+      "epoch": 1.3499005964214712,
+      "grad_norm": 0.6531151290656164,
+      "learning_rate": 7.317564209556902e-05,
+      "loss": 0.5379,
+      "step": 679
+    },
+    {
+      "epoch": 1.3518886679920477,
+      "grad_norm": 0.9240874377962419,
+      "learning_rate": 7.314458744475627e-05,
+      "loss": 0.5361,
+      "step": 680
+    },
+    {
+      "epoch": 1.3538767395626243,
+      "grad_norm": 1.0912021119009678,
+      "learning_rate": 7.311346891713852e-05,
+      "loss": 0.5324,
+      "step": 681
+    },
+    {
+      "epoch": 1.3558648111332008,
+      "grad_norm": 0.7376919034069003,
+      "learning_rate": 7.308228657268793e-05,
+      "loss": 0.5342,
+      "step": 682
+    },
+    {
+      "epoch": 1.3578528827037775,
+      "grad_norm": 0.5350891144993892,
+      "learning_rate": 7.305104047149961e-05,
+      "loss": 0.5268,
+      "step": 683
+    },
+    {
+      "epoch": 1.359840954274354,
+      "grad_norm": 0.5603454960089224,
+      "learning_rate": 7.30197306737916e-05,
+      "loss": 0.5347,
+      "step": 684
+    },
+    {
+      "epoch": 1.3618290258449304,
+      "grad_norm": 0.6375682349348152,
+      "learning_rate": 7.298835723990465e-05,
+      "loss": 0.5256,
+      "step": 685
+    },
+    {
+      "epoch": 1.3638170974155068,
+      "grad_norm": 0.5979905914047293,
+      "learning_rate": 7.295692023030217e-05,
+      "loss": 0.5351,
+      "step": 686
+    },
+    {
+      "epoch": 1.3658051689860835,
+      "grad_norm": 0.6240089079981558,
+      "learning_rate": 7.29254197055701e-05,
+      "loss": 0.5297,
+      "step": 687
+    },
+    {
+      "epoch": 1.36779324055666,
+      "grad_norm": 0.8050947571296337,
+      "learning_rate": 7.289385572641678e-05,
+      "loss": 0.5335,
+      "step": 688
+    },
+    {
+      "epoch": 1.3697813121272366,
+      "grad_norm": 0.8059108298679306,
+      "learning_rate": 7.286222835367283e-05,
+      "loss": 0.5354,
+      "step": 689
+    },
+    {
+      "epoch": 1.371769383697813,
+      "grad_norm": 0.6260685868632353,
+      "learning_rate": 7.283053764829106e-05,
+      "loss": 0.5382,
+      "step": 690
+    },
+    {
+      "epoch": 1.3737574552683895,
+      "grad_norm": 0.48277003276738417,
+      "learning_rate": 7.279878367134633e-05,
+      "loss": 0.5366,
+      "step": 691
+    },
+    {
+      "epoch": 1.3757455268389662,
+      "grad_norm": 0.6008163142885341,
+      "learning_rate": 7.276696648403542e-05,
+      "loss": 0.5331,
+      "step": 692
+    },
+    {
+      "epoch": 1.3777335984095427,
+      "grad_norm": 0.6809704933082544,
+      "learning_rate": 7.273508614767698e-05,
+      "loss": 0.5314,
+      "step": 693
+    },
+    {
+      "epoch": 1.3797216699801194,
+      "grad_norm": 0.5791983308971101,
+      "learning_rate": 7.270314272371129e-05,
+      "loss": 0.5384,
+      "step": 694
+    },
+    {
+      "epoch": 1.3817097415506958,
+      "grad_norm": 0.46173801711286394,
+      "learning_rate": 7.26711362737003e-05,
+      "loss": 0.5402,
+      "step": 695
+    },
+    {
+      "epoch": 1.3836978131212723,
+      "grad_norm": 0.3789583200923854,
+      "learning_rate": 7.263906685932734e-05,
+      "loss": 0.5308,
+      "step": 696
+    },
+    {
+      "epoch": 1.385685884691849,
+      "grad_norm": 0.35662981556630236,
+      "learning_rate": 7.260693454239712e-05,
+      "loss": 0.533,
+      "step": 697
+    },
+    {
+      "epoch": 1.3876739562624254,
+      "grad_norm": 0.4313635452570124,
+      "learning_rate": 7.257473938483561e-05,
+      "loss": 0.5361,
+      "step": 698
+    },
+    {
+      "epoch": 1.389662027833002,
+      "grad_norm": 0.42693701726569916,
+      "learning_rate": 7.254248144868985e-05,
+      "loss": 0.5269,
+      "step": 699
+    },
+    {
+      "epoch": 1.3916500994035785,
+      "grad_norm": 0.46448401272739637,
+      "learning_rate": 7.251016079612786e-05,
+      "loss": 0.5281,
+      "step": 700
+    },
+    {
+      "epoch": 1.393638170974155,
+      "grad_norm": 0.45165956462569934,
+      "learning_rate": 7.247777748943854e-05,
+      "loss": 0.5302,
+      "step": 701
+    },
+    {
+      "epoch": 1.3956262425447317,
+      "grad_norm": 0.5051001158050499,
+      "learning_rate": 7.244533159103157e-05,
+      "loss": 0.5319,
+      "step": 702
+    },
+    {
+      "epoch": 1.3976143141153081,
+      "grad_norm": 0.5235203957856401,
+      "learning_rate": 7.24128231634372e-05,
+      "loss": 0.5293,
+      "step": 703
+    },
+    {
+      "epoch": 1.3996023856858848,
+      "grad_norm": 0.4566107243597437,
+      "learning_rate": 7.238025226930624e-05,
+      "loss": 0.5358,
+      "step": 704
+    },
+    {
+      "epoch": 1.4015904572564613,
+      "grad_norm": 0.4625460874004843,
+      "learning_rate": 7.234761897140986e-05,
+      "loss": 0.5304,
+      "step": 705
+    },
+    {
+      "epoch": 1.4035785288270377,
+      "grad_norm": 0.41859360558238123,
+      "learning_rate": 7.231492333263949e-05,
+      "loss": 0.5307,
+      "step": 706
+    },
+    {
+      "epoch": 1.4055666003976144,
+      "grad_norm": 0.48973969900503483,
+      "learning_rate": 7.228216541600671e-05,
+      "loss": 0.5331,
+      "step": 707
+    },
+    {
+      "epoch": 1.4075546719681908,
+      "grad_norm": 0.7812920436986611,
+      "learning_rate": 7.224934528464314e-05,
+      "loss": 0.5372,
+      "step": 708
+    },
+    {
+      "epoch": 1.4095427435387675,
+      "grad_norm": 0.7104262912732389,
+      "learning_rate": 7.221646300180027e-05,
+      "loss": 0.5318,
+      "step": 709
+    },
+    {
+      "epoch": 1.411530815109344,
+      "grad_norm": 0.7596596767184741,
+      "learning_rate": 7.218351863084942e-05,
+      "loss": 0.5328,
+      "step": 710
+    },
+    {
+      "epoch": 1.4135188866799204,
+      "grad_norm": 0.7856583093617283,
+      "learning_rate": 7.21505122352815e-05,
+      "loss": 0.5361,
+      "step": 711
+    },
+    {
+      "epoch": 1.4155069582504969,
+      "grad_norm": 0.7573687326885941,
+      "learning_rate": 7.2117443878707e-05,
+      "loss": 0.5369,
+      "step": 712
+    },
+    {
+      "epoch": 1.4174950298210736,
+      "grad_norm": 0.7884477181941297,
+      "learning_rate": 7.20843136248558e-05,
+      "loss": 0.5377,
+      "step": 713
+    },
+    {
+      "epoch": 1.4194831013916502,
+      "grad_norm": 0.7951212883224608,
+      "learning_rate": 7.205112153757713e-05,
+      "loss": 0.5294,
+      "step": 714
+    },
+    {
+      "epoch": 1.4214711729622267,
+      "grad_norm": 0.7295618655254292,
+      "learning_rate": 7.201786768083929e-05,
+      "loss": 0.5387,
+      "step": 715
+    },
+    {
+      "epoch": 1.4234592445328031,
+      "grad_norm": 0.6204183333887281,
+      "learning_rate": 7.198455211872966e-05,
+      "loss": 0.5328,
+      "step": 716
+    },
+    {
+      "epoch": 1.4254473161033796,
+      "grad_norm": 0.5131018383315537,
+      "learning_rate": 7.19511749154546e-05,
+      "loss": 0.5324,
+      "step": 717
+    },
+    {
+      "epoch": 1.4274353876739563,
+      "grad_norm": 0.6107281423459198,
+      "learning_rate": 7.191773613533919e-05,
+      "loss": 0.5376,
+      "step": 718
+    },
+    {
+      "epoch": 1.4294234592445327,
+      "grad_norm": 0.9103425902361078,
+      "learning_rate": 7.188423584282719e-05,
+      "loss": 0.5445,
+      "step": 719
+    },
+    {
+      "epoch": 1.4314115308151094,
+      "grad_norm": 1.145266589470004,
+      "learning_rate": 7.185067410248096e-05,
+      "loss": 0.5295,
+      "step": 720
+    },
+    {
+      "epoch": 1.4333996023856859,
+      "grad_norm": 0.6522122472416725,
+      "learning_rate": 7.181705097898122e-05,
+      "loss": 0.5314,
+      "step": 721
+    },
+    {
+      "epoch": 1.4353876739562623,
+      "grad_norm": 0.4808333705940563,
+      "learning_rate": 7.178336653712704e-05,
+      "loss": 0.5283,
+      "step": 722
+    },
+    {
+      "epoch": 1.437375745526839,
+      "grad_norm": 1.4207277257992248,
+      "learning_rate": 7.174962084183563e-05,
+      "loss": 0.5458,
+      "step": 723
+    },
+    {
+      "epoch": 1.4393638170974155,
+      "grad_norm": 0.5015133463112434,
+      "learning_rate": 7.171581395814225e-05,
+      "loss": 0.5323,
+      "step": 724
+    },
+    {
+      "epoch": 1.4413518886679921,
+      "grad_norm": 0.7617435821622607,
+      "learning_rate": 7.168194595120011e-05,
+      "loss": 0.5367,
+      "step": 725
+    },
+    {
+      "epoch": 1.4433399602385686,
+      "grad_norm": 0.9226878333434035,
+      "learning_rate": 7.16480168862802e-05,
+      "loss": 0.5336,
+      "step": 726
+    },
+    {
+      "epoch": 1.445328031809145,
+      "grad_norm": 1.0981884524716032,
+      "learning_rate": 7.161402682877117e-05,
+      "loss": 0.5382,
+      "step": 727
+    },
+    {
+      "epoch": 1.4473161033797217,
+      "grad_norm": 0.7264677197882256,
+      "learning_rate": 7.157997584417927e-05,
+      "loss": 0.5349,
+      "step": 728
+    },
+    {
+      "epoch": 1.4493041749502982,
+      "grad_norm": 0.7884303052990086,
+      "learning_rate": 7.154586399812808e-05,
+      "loss": 0.5363,
+      "step": 729
+    },
+    {
+      "epoch": 1.4512922465208749,
+      "grad_norm": 0.7470918521247032,
+      "learning_rate": 7.151169135635853e-05,
+      "loss": 0.5374,
+      "step": 730
+    },
+    {
+      "epoch": 1.4532803180914513,
+      "grad_norm": 0.6053092802449975,
+      "learning_rate": 7.147745798472871e-05,
+      "loss": 0.5401,
+      "step": 731
+    },
+    {
+      "epoch": 1.4552683896620278,
+      "grad_norm": 0.7160809940534572,
+      "learning_rate": 7.144316394921378e-05,
+      "loss": 0.5428,
+      "step": 732
+    },
+    {
+      "epoch": 1.4572564612326044,
+      "grad_norm": 0.46588637989423676,
+      "learning_rate": 7.140880931590575e-05,
+      "loss": 0.5409,
+      "step": 733
+    },
+    {
+      "epoch": 1.459244532803181,
+      "grad_norm": 0.42391000958772374,
+      "learning_rate": 7.137439415101345e-05,
+      "loss": 0.535,
+      "step": 734
+    },
+    {
+      "epoch": 1.4612326043737576,
+      "grad_norm": 0.4462128319583528,
+      "learning_rate": 7.133991852086235e-05,
+      "loss": 0.539,
+      "step": 735
+    },
+    {
+      "epoch": 1.463220675944334,
+      "grad_norm": 0.3737682098835748,
+      "learning_rate": 7.13053824918945e-05,
+      "loss": 0.5374,
+      "step": 736
+    },
+    {
+      "epoch": 1.4652087475149105,
+      "grad_norm": 0.4379515087175158,
+      "learning_rate": 7.12707861306683e-05,
+      "loss": 0.5395,
+      "step": 737
+    },
+    {
+      "epoch": 1.4671968190854872,
+      "grad_norm": 0.49246889187865955,
+      "learning_rate": 7.12361295038584e-05,
+      "loss": 0.5338,
+      "step": 738
+    },
+    {
+      "epoch": 1.4691848906560636,
+      "grad_norm": 1.0681705732859723,
+      "learning_rate": 7.120141267825567e-05,
+      "loss": 0.5564,
+      "step": 739
+    },
+    {
+      "epoch": 1.4711729622266403,
+      "grad_norm": 0.6301239024315092,
+      "learning_rate": 7.116663572076697e-05,
+      "loss": 0.5423,
+      "step": 740
+    },
+    {
+      "epoch": 1.4731610337972167,
+      "grad_norm": 0.5869716649862752,
+      "learning_rate": 7.113179869841502e-05,
+      "loss": 0.5423,
+      "step": 741
+    },
+    {
+      "epoch": 1.4751491053677932,
+      "grad_norm": 0.6262438855252204,
+      "learning_rate": 7.109690167833829e-05,
+      "loss": 0.5517,
+      "step": 742
+    },
+    {
+      "epoch": 1.4771371769383697,
+      "grad_norm": 0.5891280338185073,
+      "learning_rate": 7.106194472779093e-05,
+      "loss": 0.5433,
+      "step": 743
+    },
+    {
+      "epoch": 1.4791252485089463,
+      "grad_norm": 0.6468565422433893,
+      "learning_rate": 7.102692791414257e-05,
+      "loss": 0.549,
+      "step": 744
+    },
+    {
+      "epoch": 1.4811133200795228,
+      "grad_norm": 0.5481646916057995,
+      "learning_rate": 7.099185130487818e-05,
+      "loss": 0.5348,
+      "step": 745
+    },
+    {
+      "epoch": 1.4831013916500995,
+      "grad_norm": 0.4199924161739955,
+      "learning_rate": 7.0956714967598e-05,
+      "loss": 0.54,
+      "step": 746
+    },
+    {
+      "epoch": 1.485089463220676,
+      "grad_norm": 0.48686262677138054,
+      "learning_rate": 7.092151897001734e-05,
+      "loss": 0.5411,
+      "step": 747
+    },
+    {
+      "epoch": 1.4870775347912524,
+      "grad_norm": 0.4745863525083435,
+      "learning_rate": 7.088626337996657e-05,
+      "loss": 0.5342,
+      "step": 748
+    },
+    {
+      "epoch": 1.489065606361829,
+      "grad_norm": 0.42679968740429375,
+      "learning_rate": 7.085094826539083e-05,
+      "loss": 0.5346,
+      "step": 749
+    },
+    {
+      "epoch": 1.4910536779324055,
+      "grad_norm": 0.4972102291339057,
+      "learning_rate": 7.081557369434999e-05,
+      "loss": 0.5368,
+      "step": 750
+    },
+    {
+      "epoch": 1.4930417495029822,
+      "grad_norm": 0.5632447683848434,
+      "learning_rate": 7.078013973501853e-05,
+      "loss": 0.5338,
+      "step": 751
+    },
+    {
+      "epoch": 1.4950298210735586,
+      "grad_norm": 0.6050601239636209,
+      "learning_rate": 7.074464645568535e-05,
+      "loss": 0.5313,
+      "step": 752
+    },
+    {
+      "epoch": 1.497017892644135,
+      "grad_norm": 0.7162448792755521,
+      "learning_rate": 7.070909392475372e-05,
+      "loss": 0.531,
+      "step": 753
+    },
+    {
+      "epoch": 1.4990059642147118,
+      "grad_norm": 0.6731491376009496,
+      "learning_rate": 7.067348221074106e-05,
+      "loss": 0.5261,
+      "step": 754
+    },
+    {
+      "epoch": 1.5009940357852882,
+      "grad_norm": 0.8022863170941446,
+      "learning_rate": 7.063781138227888e-05,
+      "loss": 0.5447,
+      "step": 755
+    },
+    {
+      "epoch": 1.502982107355865,
+      "grad_norm": 0.5914280370903948,
+      "learning_rate": 7.060208150811257e-05,
+      "loss": 0.5379,
+      "step": 756
+    },
+    {
+      "epoch": 1.5049701789264414,
+      "grad_norm": 0.41190704446363857,
+      "learning_rate": 7.056629265710137e-05,
+      "loss": 0.5396,
+      "step": 757
+    },
+    {
+      "epoch": 1.5069582504970178,
+      "grad_norm": 0.401420962458254,
+      "learning_rate": 7.053044489821812e-05,
+      "loss": 0.5419,
+      "step": 758
+    },
+    {
+      "epoch": 1.5089463220675943,
+      "grad_norm": 0.5144759429269341,
+      "learning_rate": 7.049453830054927e-05,
+      "loss": 0.5388,
+      "step": 759
+    },
+    {
+      "epoch": 1.510934393638171,
+      "grad_norm": 0.5723949652263456,
+      "learning_rate": 7.045857293329457e-05,
+      "loss": 0.542,
+      "step": 760
+    },
+    {
+      "epoch": 1.5129224652087476,
+      "grad_norm": 1.1438855059147868,
+      "learning_rate": 7.042254886576712e-05,
+      "loss": 0.5351,
+      "step": 761
+    },
+    {
+      "epoch": 1.514910536779324,
+      "grad_norm": 0.4874743210763008,
+      "learning_rate": 7.038646616739309e-05,
+      "loss": 0.5408,
+      "step": 762
+    },
+    {
+      "epoch": 1.5168986083499005,
+      "grad_norm": 0.4142150568688663,
+      "learning_rate": 7.035032490771165e-05,
+      "loss": 0.5379,
+      "step": 763
+    },
+    {
+      "epoch": 1.518886679920477,
+      "grad_norm": 0.36121114882100824,
+      "learning_rate": 7.031412515637485e-05,
+      "loss": 0.5419,
+      "step": 764
+    },
+    {
+      "epoch": 1.5208747514910537,
+      "grad_norm": 0.35876241898828937,
+      "learning_rate": 7.027786698314746e-05,
+      "loss": 0.5457,
+      "step": 765
+    },
+    {
+      "epoch": 1.5228628230616303,
+      "grad_norm": 0.3641469911154147,
+      "learning_rate": 7.024155045790684e-05,
+      "loss": 0.5404,
+      "step": 766
+    },
+    {
+      "epoch": 1.5248508946322068,
+      "grad_norm": 0.33445543544945516,
+      "learning_rate": 7.020517565064282e-05,
+      "loss": 0.5333,
+      "step": 767
+    },
+    {
+      "epoch": 1.5268389662027833,
+      "grad_norm": 0.34325800594144645,
+      "learning_rate": 7.016874263145752e-05,
+      "loss": 0.5341,
+      "step": 768
+    },
+    {
+      "epoch": 1.5288270377733597,
+      "grad_norm": 0.4686449206678608,
+      "learning_rate": 7.013225147056526e-05,
+      "loss": 0.5351,
+      "step": 769
+    },
+    {
+      "epoch": 1.5308151093439364,
+      "grad_norm": 0.41485737972871756,
+      "learning_rate": 7.009570223829244e-05,
+      "loss": 0.5318,
+      "step": 770
+    },
+    {
+      "epoch": 1.532803180914513,
+      "grad_norm": 0.5248903855009908,
+      "learning_rate": 7.005909500507735e-05,
+      "loss": 0.5368,
+      "step": 771
+    },
+    {
+      "epoch": 1.5347912524850895,
+      "grad_norm": 0.47323182987265955,
+      "learning_rate": 7.002242984147005e-05,
+      "loss": 0.537,
+      "step": 772
+    },
+    {
+      "epoch": 1.536779324055666,
+      "grad_norm": 0.541703280414164,
+      "learning_rate": 6.998570681813226e-05,
+      "loss": 0.536,
+      "step": 773
+    },
+    {
+      "epoch": 1.5387673956262424,
+      "grad_norm": 0.5737224485492203,
+      "learning_rate": 6.994892600583722e-05,
+      "loss": 0.5372,
+      "step": 774
+    },
+    {
+      "epoch": 1.540755467196819,
+      "grad_norm": 0.4745838189750734,
+      "learning_rate": 6.991208747546951e-05,
+      "loss": 0.5403,
+      "step": 775
+    },
+    {
+      "epoch": 1.5427435387673958,
+      "grad_norm": 0.4354551274230896,
+      "learning_rate": 6.987519129802499e-05,
+      "loss": 0.5395,
+      "step": 776
+    },
+    {
+      "epoch": 1.5447316103379722,
+      "grad_norm": 0.4475851156104487,
+      "learning_rate": 6.983823754461056e-05,
+      "loss": 0.5293,
+      "step": 777
+    },
+    {
+      "epoch": 1.5467196819085487,
+      "grad_norm": 0.35033211689343957,
+      "learning_rate": 6.980122628644414e-05,
+      "loss": 0.5367,
+      "step": 778
+    },
+    {
+      "epoch": 1.5487077534791251,
+      "grad_norm": 0.32233645639354963,
+      "learning_rate": 6.976415759485443e-05,
+      "loss": 0.5327,
+      "step": 779
+    },
+    {
+      "epoch": 1.5506958250497018,
+      "grad_norm": 0.29258727348920655,
+      "learning_rate": 6.972703154128084e-05,
+      "loss": 0.529,
+      "step": 780
+    },
+    {
+      "epoch": 1.5526838966202783,
+      "grad_norm": 0.2799251045164188,
+      "learning_rate": 6.968984819727334e-05,
+      "loss": 0.5307,
+      "step": 781
+    },
+    {
+      "epoch": 1.554671968190855,
+      "grad_norm": 0.25309699205970554,
+      "learning_rate": 6.965260763449227e-05,
+      "loss": 0.5282,
+      "step": 782
+    },
+    {
+      "epoch": 1.5566600397614314,
+      "grad_norm": 0.32992663486582197,
+      "learning_rate": 6.961530992470828e-05,
+      "loss": 0.5383,
+      "step": 783
+    },
+    {
+      "epoch": 1.5586481113320079,
+      "grad_norm": 0.41878571772654466,
+      "learning_rate": 6.957795513980214e-05,
+      "loss": 0.533,
+      "step": 784
+    },
+    {
+      "epoch": 1.5606361829025845,
+      "grad_norm": 0.42375355349911037,
+      "learning_rate": 6.954054335176461e-05,
+      "loss": 0.5292,
+      "step": 785
+    },
+    {
+      "epoch": 1.562624254473161,
+      "grad_norm": 0.37729083805071806,
+      "learning_rate": 6.950307463269633e-05,
+      "loss": 0.5337,
+      "step": 786
+    },
+    {
+      "epoch": 1.5646123260437377,
+      "grad_norm": 0.5172787811154693,
+      "learning_rate": 6.946554905480764e-05,
+      "loss": 0.5311,
+      "step": 787
+    },
+    {
+      "epoch": 1.5666003976143141,
+      "grad_norm": 0.6430291497025111,
+      "learning_rate": 6.942796669041845e-05,
+      "loss": 0.5266,
+      "step": 788
+    },
+    {
+      "epoch": 1.5685884691848906,
+      "grad_norm": 0.7341682662010934,
+      "learning_rate": 6.939032761195814e-05,
+      "loss": 0.5319,
+      "step": 789
+    },
+    {
+      "epoch": 1.570576540755467,
+      "grad_norm": 0.7733283890973897,
+      "learning_rate": 6.935263189196536e-05,
+      "loss": 0.525,
+      "step": 790
+    },
+    {
+      "epoch": 1.5725646123260437,
+      "grad_norm": 0.7479417608111679,
+      "learning_rate": 6.931487960308795e-05,
+      "loss": 0.5311,
+      "step": 791
+    },
+    {
+      "epoch": 1.5745526838966204,
+      "grad_norm": 0.6757799604868806,
+      "learning_rate": 6.927707081808274e-05,
+      "loss": 0.5302,
+      "step": 792
+    },
+    {
+      "epoch": 1.5765407554671969,
+      "grad_norm": 0.7227696163173898,
+      "learning_rate": 6.923920560981547e-05,
+      "loss": 0.5289,
+      "step": 793
+    },
+    {
+      "epoch": 1.5785288270377733,
+      "grad_norm": 0.824107106148003,
+      "learning_rate": 6.920128405126057e-05,
+      "loss": 0.5256,
+      "step": 794
+    },
+    {
+      "epoch": 1.5805168986083498,
+      "grad_norm": 0.8628110973646821,
+      "learning_rate": 6.916330621550116e-05,
+      "loss": 0.5295,
+      "step": 795
+    },
+    {
+      "epoch": 1.5825049701789264,
+      "grad_norm": 0.7654075564561619,
+      "learning_rate": 6.912527217572872e-05,
+      "loss": 0.5295,
+      "step": 796
+    },
+    {
+      "epoch": 1.5844930417495031,
+      "grad_norm": 0.6046199689394902,
+      "learning_rate": 6.90871820052431e-05,
+      "loss": 0.5244,
+      "step": 797
+    },
+    {
+      "epoch": 1.5864811133200796,
+      "grad_norm": 0.4598376110935637,
+      "learning_rate": 6.90490357774523e-05,
+      "loss": 0.5242,
+      "step": 798
+    },
+    {
+      "epoch": 1.588469184890656,
+      "grad_norm": 0.383063601508151,
+      "learning_rate": 6.90108335658724e-05,
+      "loss": 0.528,
+      "step": 799
+    },
+    {
+      "epoch": 1.5904572564612325,
+      "grad_norm": 0.3312610747130047,
+      "learning_rate": 6.897257544412733e-05,
+      "loss": 0.5233,
+      "step": 800
+    },
+    {
+      "epoch": 1.5924453280318092,
+      "grad_norm": 0.37975991863998054,
+      "learning_rate": 6.893426148594878e-05,
+      "loss": 0.5258,
+      "step": 801
+    },
+    {
+      "epoch": 1.5944333996023858,
+      "grad_norm": 0.518783862803526,
+      "learning_rate": 6.889589176517605e-05,
+      "loss": 0.5279,
+      "step": 802
+    },
+    {
+      "epoch": 1.5964214711729623,
+      "grad_norm": 0.620415152981916,
+      "learning_rate": 6.885746635575594e-05,
+      "loss": 0.5277,
+      "step": 803
+    },
+    {
+      "epoch": 1.5984095427435387,
+      "grad_norm": 0.5372244798306283,
+      "learning_rate": 6.88189853317425e-05,
+      "loss": 0.528,
+      "step": 804
+    },
+    {
+      "epoch": 1.6003976143141152,
+      "grad_norm": 0.35429448129413066,
+      "learning_rate": 6.878044876729703e-05,
+      "loss": 0.5311,
+      "step": 805
+    },
+    {
+      "epoch": 1.6023856858846919,
+      "grad_norm": 0.471670336402686,
+      "learning_rate": 6.874185673668783e-05,
+      "loss": 0.5295,
+      "step": 806
+    },
+    {
+      "epoch": 1.6043737574552683,
+      "grad_norm": 0.29450520073608794,
+      "learning_rate": 6.870320931429013e-05,
+      "loss": 0.5227,
+      "step": 807
+    },
+    {
+      "epoch": 1.606361829025845,
+      "grad_norm": 0.3606020368465704,
+      "learning_rate": 6.866450657458587e-05,
+      "loss": 0.5269,
+      "step": 808
+    },
+    {
+      "epoch": 1.6083499005964215,
+      "grad_norm": 0.4286070793201051,
+      "learning_rate": 6.862574859216363e-05,
+      "loss": 0.5255,
+      "step": 809
+    },
+    {
+      "epoch": 1.610337972166998,
+      "grad_norm": 0.42927764814865915,
+      "learning_rate": 6.858693544171845e-05,
+      "loss": 0.5279,
+      "step": 810
+    },
+    {
+      "epoch": 1.6123260437375746,
+      "grad_norm": 0.4410104674784239,
+      "learning_rate": 6.854806719805166e-05,
+      "loss": 0.5265,
+      "step": 811
+    },
+    {
+      "epoch": 1.614314115308151,
+      "grad_norm": 0.531026735276354,
+      "learning_rate": 6.850914393607082e-05,
+      "loss": 0.5309,
+      "step": 812
+    },
+    {
+      "epoch": 1.6163021868787277,
+      "grad_norm": 0.6716814454381097,
+      "learning_rate": 6.847016573078947e-05,
+      "loss": 0.5366,
+      "step": 813
+    },
+    {
+      "epoch": 1.6182902584493042,
+      "grad_norm": 0.7934744929903685,
+      "learning_rate": 6.843113265732708e-05,
+      "loss": 0.5192,
+      "step": 814
+    },
+    {
+      "epoch": 1.6202783300198806,
+      "grad_norm": 0.9038605001312126,
+      "learning_rate": 6.839204479090883e-05,
+      "loss": 0.5283,
+      "step": 815
+    },
+    {
+      "epoch": 1.622266401590457,
+      "grad_norm": 0.9290349986214314,
+      "learning_rate": 6.835290220686553e-05,
+      "loss": 0.529,
+      "step": 816
+    },
+    {
+      "epoch": 1.6242544731610338,
+      "grad_norm": 0.835195207409939,
+      "learning_rate": 6.831370498063341e-05,
+      "loss": 0.5316,
+      "step": 817
+    },
+    {
+      "epoch": 1.6262425447316105,
+      "grad_norm": 0.6921457969116479,
+      "learning_rate": 6.827445318775401e-05,
+      "loss": 0.5268,
+      "step": 818
+    },
+    {
+      "epoch": 1.628230616302187,
+      "grad_norm": 0.4237508185833249,
+      "learning_rate": 6.823514690387407e-05,
+      "loss": 0.5195,
+      "step": 819
+    },
+    {
+      "epoch": 1.6302186878727634,
+      "grad_norm": 0.3942746067868676,
+      "learning_rate": 6.81957862047453e-05,
+      "loss": 0.5276,
+      "step": 820
+    },
+    {
+      "epoch": 1.6322067594433398,
+      "grad_norm": 0.5755721778151188,
+      "learning_rate": 6.815637116622432e-05,
+      "loss": 0.5344,
+      "step": 821
+    },
+    {
+      "epoch": 1.6341948310139165,
+      "grad_norm": 0.6478367776398117,
+      "learning_rate": 6.811690186427244e-05,
+      "loss": 0.5237,
+      "step": 822
+    },
+    {
+      "epoch": 1.6361829025844932,
+      "grad_norm": 0.5020337919490125,
+      "learning_rate": 6.807737837495556e-05,
+      "loss": 0.5285,
+      "step": 823
+    },
+    {
+      "epoch": 1.6381709741550696,
+      "grad_norm": 0.36435989435714544,
+      "learning_rate": 6.803780077444401e-05,
+      "loss": 0.5259,
+      "step": 824
+    },
+    {
+      "epoch": 1.640159045725646,
+      "grad_norm": 0.3882654380737712,
+      "learning_rate": 6.799816913901242e-05,
+      "loss": 0.5271,
+      "step": 825
+    },
+    {
+      "epoch": 1.6421471172962225,
+      "grad_norm": 0.49404535476174083,
+      "learning_rate": 6.795848354503951e-05,
+      "loss": 0.5247,
+      "step": 826
+    },
+    {
+      "epoch": 1.6441351888667992,
+      "grad_norm": 0.5112014431648076,
+      "learning_rate": 6.791874406900806e-05,
+      "loss": 0.5265,
+      "step": 827
+    },
+    {
+      "epoch": 1.646123260437376,
+      "grad_norm": 0.45808472140668066,
+      "learning_rate": 6.787895078750465e-05,
+      "loss": 0.525,
+      "step": 828
+    },
+    {
+      "epoch": 1.6481113320079523,
+      "grad_norm": 1.2914831242671405,
+      "learning_rate": 6.783910377721954e-05,
+      "loss": 0.52,
+      "step": 829
+    },
+    {
+      "epoch": 1.6500994035785288,
+      "grad_norm": 0.3384619854495341,
+      "learning_rate": 6.779920311494657e-05,
+      "loss": 0.5272,
+      "step": 830
+    },
+    {
+      "epoch": 1.6520874751491053,
+      "grad_norm": 0.5083556366105404,
+      "learning_rate": 6.775924887758296e-05,
+      "loss": 0.5195,
+      "step": 831
+    },
+    {
+      "epoch": 1.654075546719682,
+      "grad_norm": 0.6386003447389207,
+      "learning_rate": 6.771924114212917e-05,
+      "loss": 0.5243,
+      "step": 832
+    },
+    {
+      "epoch": 1.6560636182902586,
+      "grad_norm": 0.5423661412327074,
+      "learning_rate": 6.767917998568882e-05,
+      "loss": 0.5214,
+      "step": 833
+    },
+    {
+      "epoch": 1.658051689860835,
+      "grad_norm": 0.46973576508674253,
+      "learning_rate": 6.763906548546841e-05,
+      "loss": 0.53,
+      "step": 834
+    },
+    {
+      "epoch": 1.6600397614314115,
+      "grad_norm": 0.40006367573081403,
+      "learning_rate": 6.75988977187773e-05,
+      "loss": 0.5302,
+      "step": 835
+    },
+    {
+      "epoch": 1.662027833001988,
+      "grad_norm": 0.4694950009603063,
+      "learning_rate": 6.755867676302747e-05,
+      "loss": 0.5212,
+      "step": 836
+    },
+    {
+      "epoch": 1.6640159045725647,
+      "grad_norm": 0.5151875858962307,
+      "learning_rate": 6.751840269573342e-05,
+      "loss": 0.5286,
+      "step": 837
+    },
+    {
+      "epoch": 1.666003976143141,
+      "grad_norm": 1.2198920217976148,
+      "learning_rate": 6.747807559451201e-05,
+      "loss": 0.5271,
+      "step": 838
+    },
+    {
+      "epoch": 1.6679920477137178,
+      "grad_norm": 0.3313618013972345,
+      "learning_rate": 6.74376955370823e-05,
+      "loss": 0.5336,
+      "step": 839
+    },
+    {
+      "epoch": 1.6699801192842942,
+      "grad_norm": 0.4987390488162863,
+      "learning_rate": 6.739726260126545e-05,
+      "loss": 0.5317,
+      "step": 840
+    },
+    {
+      "epoch": 1.6719681908548707,
+      "grad_norm": 0.7074939333172882,
+      "learning_rate": 6.735677686498443e-05,
+      "loss": 0.537,
+      "step": 841
+    },
+    {
+      "epoch": 1.6739562624254472,
+      "grad_norm": 0.6747029464356521,
+      "learning_rate": 6.731623840626408e-05,
+      "loss": 0.5372,
+      "step": 842
+    },
+    {
+      "epoch": 1.6759443339960238,
+      "grad_norm": 0.6466193766227213,
+      "learning_rate": 6.727564730323077e-05,
+      "loss": 0.5296,
+      "step": 843
+    },
+    {
+      "epoch": 1.6779324055666005,
+      "grad_norm": 0.5584777552614117,
+      "learning_rate": 6.723500363411238e-05,
+      "loss": 0.5252,
+      "step": 844
+    },
+    {
+      "epoch": 1.679920477137177,
+      "grad_norm": 0.453649217028817,
+      "learning_rate": 6.719430747723805e-05,
+      "loss": 0.5288,
+      "step": 845
+    },
+    {
+      "epoch": 1.6819085487077534,
+      "grad_norm": 0.48188541292199083,
+      "learning_rate": 6.715355891103811e-05,
+      "loss": 0.5264,
+      "step": 846
+    },
+    {
+      "epoch": 1.6838966202783299,
+      "grad_norm": 0.4254994544193942,
+      "learning_rate": 6.711275801404387e-05,
+      "loss": 0.529,
+      "step": 847
+    },
+    {
+      "epoch": 1.6858846918489065,
+      "grad_norm": 0.380870644142891,
+      "learning_rate": 6.707190486488751e-05,
+      "loss": 0.5276,
+      "step": 848
+    },
+    {
+      "epoch": 1.6878727634194832,
+      "grad_norm": 0.4177074544613175,
+      "learning_rate": 6.70309995423019e-05,
+      "loss": 0.5318,
+      "step": 849
+    },
+    {
+      "epoch": 1.6898608349900597,
+      "grad_norm": 0.395911119103237,
+      "learning_rate": 6.699004212512046e-05,
+      "loss": 0.526,
+      "step": 850
+    },
+    {
+      "epoch": 1.6918489065606361,
+      "grad_norm": 0.45543127339921896,
+      "learning_rate": 6.694903269227703e-05,
+      "loss": 0.5248,
+      "step": 851
+    },
+    {
+      "epoch": 1.6938369781312126,
+      "grad_norm": 0.44150579657087613,
+      "learning_rate": 6.690797132280565e-05,
+      "loss": 0.5243,
+      "step": 852
+    },
+    {
+      "epoch": 1.6958250497017893,
+      "grad_norm": 0.3842977414436349,
+      "learning_rate": 6.686685809584051e-05,
+      "loss": 0.5172,
+      "step": 853
+    },
+    {
+      "epoch": 1.697813121272366,
+      "grad_norm": 0.3267241356619446,
+      "learning_rate": 6.682569309061567e-05,
+      "loss": 0.5359,
+      "step": 854
+    },
+    {
+      "epoch": 1.6998011928429424,
+      "grad_norm": 0.31023857384364334,
+      "learning_rate": 6.678447638646507e-05,
+      "loss": 0.5203,
+      "step": 855
+    },
+    {
+      "epoch": 1.7017892644135189,
+      "grad_norm": 0.4582609159970721,
+      "learning_rate": 6.67432080628222e-05,
+      "loss": 0.5246,
+      "step": 856
+    },
+    {
+      "epoch": 1.7037773359840953,
+      "grad_norm": 0.4371367746577152,
+      "learning_rate": 6.670188819922006e-05,
+      "loss": 0.5256,
+      "step": 857
+    },
+    {
+      "epoch": 1.705765407554672,
+      "grad_norm": 0.3601588642012915,
+      "learning_rate": 6.666051687529102e-05,
+      "loss": 0.527,
+      "step": 858
+    },
+    {
+      "epoch": 1.7077534791252487,
+      "grad_norm": 0.3598932666256967,
+      "learning_rate": 6.661909417076656e-05,
+      "loss": 0.522,
+      "step": 859
+    },
+    {
+      "epoch": 1.7097415506958251,
+      "grad_norm": 0.3814306793726626,
+      "learning_rate": 6.65776201654772e-05,
+      "loss": 0.5263,
+      "step": 860
+    },
+    {
+      "epoch": 1.7117296222664016,
+      "grad_norm": 0.5155709892788785,
+      "learning_rate": 6.653609493935239e-05,
+      "loss": 0.5277,
+      "step": 861
+    },
+    {
+      "epoch": 1.713717693836978,
+      "grad_norm": 0.592297522606716,
+      "learning_rate": 6.649451857242018e-05,
+      "loss": 0.5302,
+      "step": 862
+    },
+    {
+      "epoch": 1.7157057654075547,
+      "grad_norm": 0.62317820887808,
+      "learning_rate": 6.64528911448073e-05,
+      "loss": 0.5314,
+      "step": 863
+    },
+    {
+      "epoch": 1.7176938369781312,
+      "grad_norm": 0.5293921255569864,
+      "learning_rate": 6.641121273673877e-05,
+      "loss": 0.5341,
+      "step": 864
+    },
+    {
+      "epoch": 1.7196819085487078,
+      "grad_norm": 0.3877701813428404,
+      "learning_rate": 6.636948342853797e-05,
+      "loss": 0.5298,
+      "step": 865
+    },
+    {
+      "epoch": 1.7216699801192843,
+      "grad_norm": 0.3207068664780065,
+      "learning_rate": 6.632770330062628e-05,
+      "loss": 0.5185,
+      "step": 866
+    },
+    {
+      "epoch": 1.7236580516898607,
+      "grad_norm": 0.3000147535618842,
+      "learning_rate": 6.62858724335231e-05,
+      "loss": 0.5279,
+      "step": 867
+    },
+    {
+      "epoch": 1.7256461232604374,
+      "grad_norm": 0.27110479588599784,
+      "learning_rate": 6.624399090784555e-05,
+      "loss": 0.5166,
+      "step": 868
+    },
+    {
+      "epoch": 1.7276341948310139,
+      "grad_norm": 0.2534706020052216,
+      "learning_rate": 6.62020588043084e-05,
+      "loss": 0.5299,
+      "step": 869
+    },
+    {
+      "epoch": 1.7296222664015906,
+      "grad_norm": 0.2651273278605272,
+      "learning_rate": 6.616007620372396e-05,
+      "loss": 0.5213,
+      "step": 870
+    },
+    {
+      "epoch": 1.731610337972167,
+      "grad_norm": 0.3008823267991322,
+      "learning_rate": 6.611804318700174e-05,
+      "loss": 0.5231,
+      "step": 871
+    },
+    {
+      "epoch": 1.7335984095427435,
+      "grad_norm": 0.36886074204926794,
+      "learning_rate": 6.60759598351485e-05,
+      "loss": 0.5184,
+      "step": 872
+    },
+    {
+      "epoch": 1.73558648111332,
+      "grad_norm": 0.4053040242003246,
+      "learning_rate": 6.603382622926801e-05,
+      "loss": 0.5272,
+      "step": 873
+    },
+    {
+      "epoch": 1.7375745526838966,
+      "grad_norm": 0.35686447014595274,
+      "learning_rate": 6.599164245056081e-05,
+      "loss": 0.5087,
+      "step": 874
+    },
+    {
+      "epoch": 1.7395626242544733,
+      "grad_norm": 0.3429941917807466,
+      "learning_rate": 6.594940858032422e-05,
+      "loss": 0.5195,
+      "step": 875
+    },
+    {
+      "epoch": 1.7415506958250497,
+      "grad_norm": 0.34576654170390286,
+      "learning_rate": 6.590712469995204e-05,
+      "loss": 0.5175,
+      "step": 876
+    },
+    {
+      "epoch": 1.7435387673956262,
+      "grad_norm": 0.3310782562648362,
+      "learning_rate": 6.586479089093451e-05,
+      "loss": 0.5133,
+      "step": 877
+    },
+    {
+      "epoch": 1.7455268389662026,
+      "grad_norm": 0.4318684757964282,
+      "learning_rate": 6.582240723485801e-05,
+      "loss": 0.5184,
+      "step": 878
+    },
+    {
+      "epoch": 1.7475149105367793,
+      "grad_norm": 0.6198244221196105,
+      "learning_rate": 6.577997381340509e-05,
+      "loss": 0.5278,
+      "step": 879
+    },
+    {
+      "epoch": 1.749502982107356,
+      "grad_norm": 0.707683440817649,
+      "learning_rate": 6.573749070835409e-05,
+      "loss": 0.5237,
+      "step": 880
+    },
+    {
+      "epoch": 1.7514910536779325,
+      "grad_norm": 0.676898274391904,
+      "learning_rate": 6.56949580015792e-05,
+      "loss": 0.5203,
+      "step": 881
+    },
+    {
+      "epoch": 1.753479125248509,
+      "grad_norm": 0.6183840795701814,
+      "learning_rate": 6.565237577505016e-05,
+      "loss": 0.5108,
+      "step": 882
+    },
+    {
+      "epoch": 1.7554671968190854,
+      "grad_norm": 0.5793914378061121,
+      "learning_rate": 6.560974411083214e-05,
+      "loss": 0.5205,
+      "step": 883
+    },
+    {
+      "epoch": 1.757455268389662,
+      "grad_norm": 0.5772648421166539,
+      "learning_rate": 6.556706309108562e-05,
+      "loss": 0.5206,
+      "step": 884
+    },
+    {
+      "epoch": 1.7594433399602387,
+      "grad_norm": 0.6223903009611668,
+      "learning_rate": 6.552433279806617e-05,
+      "loss": 0.5275,
+      "step": 885
+    },
+    {
+      "epoch": 1.7614314115308152,
+      "grad_norm": 0.694303234381892,
+      "learning_rate": 6.548155331412431e-05,
+      "loss": 0.5181,
+      "step": 886
+    },
+    {
+      "epoch": 1.7634194831013916,
+      "grad_norm": 0.7367498764846339,
+      "learning_rate": 6.54387247217054e-05,
+      "loss": 0.5212,
+      "step": 887
+    },
+    {
+      "epoch": 1.765407554671968,
+      "grad_norm": 0.6573127727786128,
+      "learning_rate": 6.539584710334941e-05,
+      "loss": 0.5165,
+      "step": 888
+    },
+    {
+      "epoch": 1.7673956262425448,
+      "grad_norm": 0.5423711722715193,
+      "learning_rate": 6.535292054169083e-05,
+      "loss": 0.5255,
+      "step": 889
+    },
+    {
+      "epoch": 1.7693836978131214,
+      "grad_norm": 0.37296323316331076,
+      "learning_rate": 6.530994511945841e-05,
+      "loss": 0.5175,
+      "step": 890
+    },
+    {
+      "epoch": 1.771371769383698,
+      "grad_norm": 0.390630809690755,
+      "learning_rate": 6.526692091947513e-05,
+      "loss": 0.5231,
+      "step": 891
+    },
+    {
+      "epoch": 1.7733598409542743,
+      "grad_norm": 0.628946177002013,
+      "learning_rate": 6.522384802465792e-05,
+      "loss": 0.516,
+      "step": 892
+    },
+    {
+      "epoch": 1.7753479125248508,
+      "grad_norm": 0.7131786851847514,
+      "learning_rate": 6.518072651801763e-05,
+      "loss": 0.5139,
+      "step": 893
+    },
+    {
+      "epoch": 1.7773359840954275,
+      "grad_norm": 0.677048790263275,
+      "learning_rate": 6.513755648265874e-05,
+      "loss": 0.5085,
+      "step": 894
+    },
+    {
+      "epoch": 1.779324055666004,
+      "grad_norm": 0.5545892812569629,
+      "learning_rate": 6.509433800177925e-05,
+      "loss": 0.5157,
+      "step": 895
+    },
+    {
+      "epoch": 1.7813121272365806,
+      "grad_norm": 0.4221871273105676,
+      "learning_rate": 6.505107115867053e-05,
+      "loss": 0.5129,
+      "step": 896
+    },
+    {
+      "epoch": 1.783300198807157,
+      "grad_norm": 0.33895277193033474,
+      "learning_rate": 6.500775603671721e-05,
+      "loss": 0.5269,
+      "step": 897
+    },
+    {
+      "epoch": 1.7852882703777335,
+      "grad_norm": 0.4106684204987088,
+      "learning_rate": 6.496439271939689e-05,
+      "loss": 0.5251,
+      "step": 898
+    },
+    {
+      "epoch": 1.78727634194831,
+      "grad_norm": 0.4813385967804267,
+      "learning_rate": 6.492098129028008e-05,
+      "loss": 0.5208,
+      "step": 899
+    },
+    {
+      "epoch": 1.7892644135188867,
+      "grad_norm": 0.5095331170047483,
+      "learning_rate": 6.487752183303002e-05,
+      "loss": 0.5175,
+      "step": 900
+    },
+    {
+      "epoch": 1.7912524850894633,
+      "grad_norm": 0.5785645590013878,
+      "learning_rate": 6.483401443140251e-05,
+      "loss": 0.5182,
+      "step": 901
+    },
+    {
+      "epoch": 1.7932405566600398,
+      "grad_norm": 0.547793961103489,
+      "learning_rate": 6.479045916924574e-05,
+      "loss": 0.5165,
+      "step": 902
+    },
+    {
+      "epoch": 1.7952286282306162,
+      "grad_norm": 0.41600517135228604,
+      "learning_rate": 6.474685613050014e-05,
+      "loss": 0.5049,
+      "step": 903
+    },
+    {
+      "epoch": 1.7972166998011927,
+      "grad_norm": 0.3299862272982457,
+      "learning_rate": 6.470320539919823e-05,
+      "loss": 0.5191,
+      "step": 904
+    },
+    {
+      "epoch": 1.7992047713717694,
+      "grad_norm": 0.32073772048850396,
+      "learning_rate": 6.465950705946441e-05,
+      "loss": 0.5117,
+      "step": 905
+    },
+    {
+      "epoch": 1.801192842942346,
+      "grad_norm": 0.35624437991584945,
+      "learning_rate": 6.461576119551488e-05,
+      "loss": 0.5152,
+      "step": 906
+    },
+    {
+      "epoch": 1.8031809145129225,
+      "grad_norm": 0.3359272702550711,
+      "learning_rate": 6.457196789165737e-05,
+      "loss": 0.5208,
+      "step": 907
+    },
+    {
+      "epoch": 1.805168986083499,
+      "grad_norm": 0.3001896290540342,
+      "learning_rate": 6.452812723229111e-05,
+      "loss": 0.5155,
+      "step": 908
+    },
+    {
+      "epoch": 1.8071570576540754,
+      "grad_norm": 0.43244608682808966,
+      "learning_rate": 6.448423930190653e-05,
+      "loss": 0.5168,
+      "step": 909
+    },
+    {
+      "epoch": 1.809145129224652,
+      "grad_norm": 0.33914062705350506,
+      "learning_rate": 6.444030418508517e-05,
+      "loss": 0.5215,
+      "step": 910
+    },
+    {
+      "epoch": 1.8111332007952288,
+      "grad_norm": 0.3793585723969184,
+      "learning_rate": 6.439632196649955e-05,
+      "loss": 0.5235,
+      "step": 911
+    },
+    {
+      "epoch": 1.8131212723658052,
+      "grad_norm": 0.36209092348601185,
+      "learning_rate": 6.435229273091295e-05,
+      "loss": 0.5067,
+      "step": 912
+    },
+    {
+      "epoch": 1.8151093439363817,
+      "grad_norm": 0.39315793371620283,
+      "learning_rate": 6.430821656317921e-05,
+      "loss": 0.5202,
+      "step": 913
+    },
+    {
+      "epoch": 1.8170974155069581,
+      "grad_norm": 0.3870294576875787,
+      "learning_rate": 6.42640935482427e-05,
+      "loss": 0.525,
+      "step": 914
+    },
+    {
+      "epoch": 1.8190854870775348,
+      "grad_norm": 0.3740502813906515,
+      "learning_rate": 6.421992377113801e-05,
+      "loss": 0.5189,
+      "step": 915
+    },
+    {
+      "epoch": 1.8210735586481115,
+      "grad_norm": 0.36305841904786135,
+      "learning_rate": 6.417570731698989e-05,
+      "loss": 0.5191,
+      "step": 916
+    },
+    {
+      "epoch": 1.823061630218688,
+      "grad_norm": 0.37807892969813695,
+      "learning_rate": 6.413144427101303e-05,
+      "loss": 0.5144,
+      "step": 917
+    },
+    {
+      "epoch": 1.8250497017892644,
+      "grad_norm": 0.3458228510072188,
+      "learning_rate": 6.40871347185119e-05,
+      "loss": 0.5189,
+      "step": 918
+    },
+    {
+      "epoch": 1.8270377733598409,
+      "grad_norm": 0.38758188828148943,
+      "learning_rate": 6.404277874488063e-05,
+      "loss": 0.5108,
+      "step": 919
+    },
+    {
+      "epoch": 1.8290258449304175,
+      "grad_norm": 0.41572979496537815,
+      "learning_rate": 6.399837643560279e-05,
+      "loss": 0.5136,
+      "step": 920
+    },
+    {
+      "epoch": 1.831013916500994,
+      "grad_norm": 0.42898037039016224,
+      "learning_rate": 6.395392787625126e-05,
+      "loss": 0.5235,
+      "step": 921
+    },
+    {
+      "epoch": 1.8330019880715707,
+      "grad_norm": 0.4978473830264871,
+      "learning_rate": 6.390943315248804e-05,
+      "loss": 0.5175,
+      "step": 922
+    },
+    {
+      "epoch": 1.8349900596421471,
+      "grad_norm": 0.6629837956376023,
+      "learning_rate": 6.386489235006412e-05,
+      "loss": 0.5178,
+      "step": 923
+    },
+    {
+      "epoch": 1.8369781312127236,
+      "grad_norm": 0.5000042265219343,
+      "learning_rate": 6.382030555481928e-05,
+      "loss": 0.5158,
+      "step": 924
+    },
+    {
+      "epoch": 1.8389662027833003,
+      "grad_norm": 0.4279576070326289,
+      "learning_rate": 6.377567285268192e-05,
+      "loss": 0.5223,
+      "step": 925
+    },
+    {
+      "epoch": 1.8409542743538767,
+      "grad_norm": 0.3459736856516694,
+      "learning_rate": 6.373099432966895e-05,
+      "loss": 0.5189,
+      "step": 926
+    },
+    {
+      "epoch": 1.8429423459244534,
+      "grad_norm": 0.2961388042649445,
+      "learning_rate": 6.368627007188556e-05,
+      "loss": 0.5159,
+      "step": 927
+    },
+    {
+      "epoch": 1.8449304174950298,
+      "grad_norm": 1.158821404766412,
+      "learning_rate": 6.364150016552511e-05,
+      "loss": 0.5118,
+      "step": 928
+    },
+    {
+      "epoch": 1.8469184890656063,
+      "grad_norm": 0.3313243590186058,
+      "learning_rate": 6.359668469686891e-05,
+      "loss": 0.5133,
+      "step": 929
+    },
+    {
+      "epoch": 1.8489065606361827,
+      "grad_norm": 0.3442419248332627,
+      "learning_rate": 6.355182375228606e-05,
+      "loss": 0.5188,
+      "step": 930
+    },
+    {
+      "epoch": 1.8508946322067594,
+      "grad_norm": 0.46405524867421283,
+      "learning_rate": 6.350691741823334e-05,
+      "loss": 0.5234,
+      "step": 931
+    },
+    {
+      "epoch": 1.852882703777336,
+      "grad_norm": 0.5552544859564328,
+      "learning_rate": 6.346196578125499e-05,
+      "loss": 0.5283,
+      "step": 932
+    },
+    {
+      "epoch": 1.8548707753479126,
+      "grad_norm": 0.6149770086477317,
+      "learning_rate": 6.341696892798256e-05,
+      "loss": 0.5213,
+      "step": 933
+    },
+    {
+      "epoch": 1.856858846918489,
+      "grad_norm": 0.6066033014098224,
+      "learning_rate": 6.337192694513474e-05,
+      "loss": 0.5258,
+      "step": 934
+    },
+    {
+      "epoch": 1.8588469184890655,
+      "grad_norm": 0.5534949185853664,
+      "learning_rate": 6.33268399195172e-05,
+      "loss": 0.5217,
+      "step": 935
+    },
+    {
+      "epoch": 1.8608349900596421,
+      "grad_norm": 0.46419762093019756,
+      "learning_rate": 6.328170793802239e-05,
+      "loss": 0.5184,
+      "step": 936
+    },
+    {
+      "epoch": 1.8628230616302188,
+      "grad_norm": 0.40115737747047064,
+      "learning_rate": 6.323653108762946e-05,
+      "loss": 0.5136,
+      "step": 937
+    },
+    {
+      "epoch": 1.8648111332007953,
+      "grad_norm": 0.4968544590324144,
+      "learning_rate": 6.319130945540397e-05,
+      "loss": 0.5173,
+      "step": 938
+    },
+    {
+      "epoch": 1.8667992047713717,
+      "grad_norm": 0.8922179521301115,
+      "learning_rate": 6.31460431284978e-05,
+      "loss": 0.5387,
+      "step": 939
+    },
+    {
+      "epoch": 1.8687872763419482,
+      "grad_norm": 0.8274425993048475,
+      "learning_rate": 6.310073219414897e-05,
+      "loss": 0.5171,
+      "step": 940
+    },
+    {
+      "epoch": 1.8707753479125249,
+      "grad_norm": 0.9470368749962408,
+      "learning_rate": 6.305537673968152e-05,
+      "loss": 0.5233,
+      "step": 941
+    },
+    {
+      "epoch": 1.8727634194831015,
+      "grad_norm": 0.9116078804524538,
+      "learning_rate": 6.300997685250518e-05,
+      "loss": 0.519,
+      "step": 942
+    },
+    {
+      "epoch": 1.874751491053678,
+      "grad_norm": 0.7644404414742432,
+      "learning_rate": 6.296453262011543e-05,
+      "loss": 0.522,
+      "step": 943
+    },
+    {
+      "epoch": 1.8767395626242545,
+      "grad_norm": 0.6794360427532787,
+      "learning_rate": 6.291904413009313e-05,
+      "loss": 0.5068,
+      "step": 944
+    },
+    {
+      "epoch": 1.878727634194831,
+      "grad_norm": 6.258038066728078,
+      "learning_rate": 6.287351147010445e-05,
+      "loss": 0.5197,
+      "step": 945
+    },
+    {
+      "epoch": 1.8807157057654076,
+      "grad_norm": 1.0714907278173418,
+      "learning_rate": 6.282793472790073e-05,
+      "loss": 0.5129,
+      "step": 946
+    },
+    {
+      "epoch": 1.882703777335984,
+      "grad_norm": 1.6860408777693041,
+      "learning_rate": 6.27823139913182e-05,
+      "loss": 0.5158,
+      "step": 947
+    },
+    {
+      "epoch": 1.8846918489065607,
+      "grad_norm": 0.9157025608605575,
+      "learning_rate": 6.273664934827792e-05,
+      "loss": 0.5254,
+      "step": 948
+    },
+    {
+      "epoch": 1.8866799204771372,
+      "grad_norm": 2.3654517437481464,
+      "learning_rate": 6.269094088678555e-05,
+      "loss": 0.5227,
+      "step": 949
+    },
+    {
+      "epoch": 1.8886679920477136,
+      "grad_norm": 1.941802659426266,
+      "learning_rate": 6.264518869493123e-05,
+      "loss": 0.5288,
+      "step": 950
+    },
+    {
+      "epoch": 1.8906560636182903,
+      "grad_norm": 1.3723093581829342,
+      "learning_rate": 6.259939286088932e-05,
+      "loss": 0.5272,
+      "step": 951
+    },
+    {
+      "epoch": 1.8926441351888668,
+      "grad_norm": 1.1729092180787042,
+      "learning_rate": 6.255355347291833e-05,
+      "loss": 0.5269,
+      "step": 952
+    },
+    {
+      "epoch": 1.8946322067594434,
+      "grad_norm": 1.5119822561319634,
+      "learning_rate": 6.25076706193607e-05,
+      "loss": 0.5199,
+      "step": 953
+    },
+    {
+      "epoch": 1.89662027833002,
+      "grad_norm": 1.0227380539137576,
+      "learning_rate": 6.246174438864264e-05,
+      "loss": 0.5208,
+      "step": 954
+    },
+    {
+      "epoch": 1.8986083499005963,
+      "grad_norm": 1.6110735368657032,
+      "learning_rate": 6.241577486927396e-05,
+      "loss": 0.5267,
+      "step": 955
+    },
+    {
+      "epoch": 1.9005964214711728,
+      "grad_norm": 1.3696020584560245,
+      "learning_rate": 6.236976214984787e-05,
+      "loss": 0.5307,
+      "step": 956
+    },
+    {
+      "epoch": 1.9025844930417495,
+      "grad_norm": 1.1164032790705993,
+      "learning_rate": 6.232370631904085e-05,
+      "loss": 0.5235,
+      "step": 957
+    },
+    {
+      "epoch": 1.9045725646123262,
+      "grad_norm": 0.9403848803470333,
+      "learning_rate": 6.227760746561248e-05,
+      "loss": 0.5126,
+      "step": 958
+    },
+    {
+      "epoch": 1.9065606361829026,
+      "grad_norm": 1.1097967355382794,
+      "learning_rate": 6.223146567840526e-05,
+      "loss": 0.5213,
+      "step": 959
+    },
+    {
+      "epoch": 1.908548707753479,
+      "grad_norm": 0.7036299666285428,
+      "learning_rate": 6.218528104634439e-05,
+      "loss": 0.5165,
+      "step": 960
+    },
+    {
+      "epoch": 1.9105367793240555,
+      "grad_norm": 1.1611817296430542,
+      "learning_rate": 6.213905365843767e-05,
+      "loss": 0.5172,
+      "step": 961
+    },
+    {
+      "epoch": 1.9125248508946322,
+      "grad_norm": 0.6447298644520358,
+      "learning_rate": 6.209278360377532e-05,
+      "loss": 0.5239,
+      "step": 962
+    },
+    {
+      "epoch": 1.9145129224652089,
+      "grad_norm": 0.961035900760647,
+      "learning_rate": 6.204647097152976e-05,
+      "loss": 0.5158,
+      "step": 963
+    },
+    {
+      "epoch": 1.9165009940357853,
+      "grad_norm": 0.5611928909928571,
+      "learning_rate": 6.200011585095546e-05,
+      "loss": 0.5207,
+      "step": 964
+    },
+    {
+      "epoch": 1.9184890656063618,
+      "grad_norm": 0.8349582131183718,
+      "learning_rate": 6.195371833138878e-05,
+      "loss": 0.517,
+      "step": 965
+    },
+    {
+      "epoch": 1.9204771371769382,
+      "grad_norm": 0.4906951073133178,
+      "learning_rate": 6.190727850224781e-05,
+      "loss": 0.5205,
+      "step": 966
+    },
+    {
+      "epoch": 1.922465208747515,
+      "grad_norm": 0.5781023232362354,
+      "learning_rate": 6.186079645303217e-05,
+      "loss": 0.5109,
+      "step": 967
+    },
+    {
+      "epoch": 1.9244532803180916,
+      "grad_norm": 0.4523458594191586,
+      "learning_rate": 6.181427227332284e-05,
+      "loss": 0.5163,
+      "step": 968
+    },
+    {
+      "epoch": 1.926441351888668,
+      "grad_norm": 0.5895393535936876,
+      "learning_rate": 6.1767706052782e-05,
+      "loss": 0.5101,
+      "step": 969
+    },
+    {
+      "epoch": 1.9284294234592445,
+      "grad_norm": 0.4494991711450939,
+      "learning_rate": 6.172109788115286e-05,
+      "loss": 0.5134,
+      "step": 970
+    },
+    {
+      "epoch": 1.930417495029821,
+      "grad_norm": 0.49354196995819166,
+      "learning_rate": 6.167444784825945e-05,
+      "loss": 0.5209,
+      "step": 971
+    },
+    {
+      "epoch": 1.9324055666003976,
+      "grad_norm": 0.734680878386189,
+      "learning_rate": 6.162775604400653e-05,
+      "loss": 0.538,
+      "step": 972
+    },
+    {
+      "epoch": 1.9343936381709743,
+      "grad_norm": 0.4126375739890944,
+      "learning_rate": 6.158102255837928e-05,
+      "loss": 0.5089,
+      "step": 973
+    },
+    {
+      "epoch": 1.9363817097415508,
+      "grad_norm": 2.5240012713966222,
+      "learning_rate": 6.153424748144331e-05,
+      "loss": 0.5484,
+      "step": 974
+    },
+    {
+      "epoch": 1.9383697813121272,
+      "grad_norm": 0.6639225231429577,
+      "learning_rate": 6.148743090334429e-05,
+      "loss": 0.5243,
+      "step": 975
+    },
+    {
+      "epoch": 1.9403578528827037,
+      "grad_norm": 0.7918093768800656,
+      "learning_rate": 6.144057291430794e-05,
+      "loss": 0.5137,
+      "step": 976
+    },
+    {
+      "epoch": 1.9423459244532804,
+      "grad_norm": 1.4149190907951867,
+      "learning_rate": 6.139367360463976e-05,
+      "loss": 0.5224,
+      "step": 977
+    },
+    {
+      "epoch": 1.9443339960238568,
+      "grad_norm": 0.46582696350343256,
+      "learning_rate": 6.134673306472487e-05,
+      "loss": 0.5182,
+      "step": 978
+    },
+    {
+      "epoch": 1.9463220675944335,
+      "grad_norm": 0.9635660986087046,
+      "learning_rate": 6.129975138502788e-05,
+      "loss": 0.528,
+      "step": 979
+    },
+    {
+      "epoch": 1.94831013916501,
+      "grad_norm": 1.1033730040373586,
+      "learning_rate": 6.125272865609267e-05,
+      "loss": 0.5189,
+      "step": 980
+    },
+    {
+      "epoch": 1.9502982107355864,
+      "grad_norm": 0.6710161371168709,
+      "learning_rate": 6.120566496854225e-05,
+      "loss": 0.5233,
+      "step": 981
+    },
+    {
+      "epoch": 1.9522862823061629,
+      "grad_norm": 0.5342304493013306,
+      "learning_rate": 6.11585604130785e-05,
+      "loss": 0.5188,
+      "step": 982
+    },
+    {
+      "epoch": 1.9542743538767395,
+      "grad_norm": 0.7351872609836071,
+      "learning_rate": 6.111141508048217e-05,
+      "loss": 0.5083,
+      "step": 983
+    },
+    {
+      "epoch": 1.9562624254473162,
+      "grad_norm": 0.6600541489771485,
+      "learning_rate": 6.106422906161248e-05,
+      "loss": 0.5234,
+      "step": 984
+    },
+    {
+      "epoch": 1.9582504970178927,
+      "grad_norm": 0.5571169506680288,
+      "learning_rate": 6.101700244740715e-05,
+      "loss": 0.5258,
+      "step": 985
+    },
+    {
+      "epoch": 1.9602385685884691,
+      "grad_norm": 0.7165724959209299,
+      "learning_rate": 6.0969735328882086e-05,
+      "loss": 0.5238,
+      "step": 986
+    },
+    {
+      "epoch": 1.9622266401590456,
+      "grad_norm": 0.643330366441739,
+      "learning_rate": 6.092242779713131e-05,
+      "loss": 0.5156,
+      "step": 987
+    },
+    {
+      "epoch": 1.9642147117296223,
+      "grad_norm": 0.476323219464556,
+      "learning_rate": 6.087507994332663e-05,
+      "loss": 0.5165,
+      "step": 988
+    },
+    {
+      "epoch": 1.966202783300199,
+      "grad_norm": 0.41439350527321883,
+      "learning_rate": 6.0827691858717674e-05,
+      "loss": 0.521,
+      "step": 989
+    },
+    {
+      "epoch": 1.9681908548707754,
+      "grad_norm": 0.5424398593597871,
+      "learning_rate": 6.078026363463153e-05,
+      "loss": 0.5212,
+      "step": 990
+    },
+    {
+      "epoch": 1.9701789264413518,
+      "grad_norm": 0.4603082917256751,
+      "learning_rate": 6.073279536247268e-05,
+      "loss": 0.5138,
+      "step": 991
+    },
+    {
+      "epoch": 1.9721669980119283,
+      "grad_norm": 0.5001963424623456,
+      "learning_rate": 6.068528713372276e-05,
+      "loss": 0.5247,
+      "step": 992
+    },
+    {
+      "epoch": 1.974155069582505,
+      "grad_norm": 0.4730380781447341,
+      "learning_rate": 6.063773903994043e-05,
+      "loss": 0.5141,
+      "step": 993
+    },
+    {
+      "epoch": 1.9761431411530817,
+      "grad_norm": 0.43352243194755535,
+      "learning_rate": 6.0590151172761186e-05,
+      "loss": 0.5205,
+      "step": 994
+    },
+    {
+      "epoch": 1.978131212723658,
+      "grad_norm": 0.48285340824373285,
+      "learning_rate": 6.054252362389715e-05,
+      "loss": 0.5118,
+      "step": 995
+    },
+    {
+      "epoch": 1.9801192842942346,
+      "grad_norm": 0.40877908110477573,
+      "learning_rate": 6.049485648513696e-05,
+      "loss": 0.5225,
+      "step": 996
+    },
+    {
+      "epoch": 1.982107355864811,
+      "grad_norm": 0.3780132592901253,
+      "learning_rate": 6.0447149848345476e-05,
+      "loss": 0.5206,
+      "step": 997
+    },
+    {
+      "epoch": 1.9840954274353877,
+      "grad_norm": 0.35453842643953837,
+      "learning_rate": 6.0399403805463767e-05,
+      "loss": 0.5126,
+      "step": 998
+    },
+    {
+      "epoch": 1.9860834990059644,
+      "grad_norm": 0.3032849729752592,
+      "learning_rate": 6.03516184485088e-05,
+      "loss": 0.5185,
+      "step": 999
+    },
+    {
+      "epoch": 1.9880715705765408,
+      "grad_norm": 0.49672445891973527,
+      "learning_rate": 6.030379386957331e-05,
+      "loss": 0.5257,
+      "step": 1000
+    },
+    {
+      "epoch": 1.9900596421471173,
+      "grad_norm": 0.34605524527043097,
+      "learning_rate": 6.025593016082564e-05,
+      "loss": 0.5149,
+      "step": 1001
+    },
+    {
+      "epoch": 1.9920477137176937,
+      "grad_norm": 0.34364685471639445,
+      "learning_rate": 6.020802741450951e-05,
+      "loss": 0.5118,
+      "step": 1002
+    },
+    {
+      "epoch": 1.9940357852882704,
+      "grad_norm": 0.3179870833998065,
+      "learning_rate": 6.016008572294392e-05,
+      "loss": 0.515,
+      "step": 1003
+    },
+    {
+      "epoch": 1.9960238568588469,
+      "grad_norm": 0.3567363057139461,
+      "learning_rate": 6.01121051785229e-05,
+      "loss": 0.5226,
+      "step": 1004
+    },
+    {
+      "epoch": 1.9980119284294235,
+      "grad_norm": 0.29232456854575584,
+      "learning_rate": 6.006408587371535e-05,
+      "loss": 0.5145,
+      "step": 1005
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.3614805776515675,
+      "learning_rate": 6.001602790106488e-05,
+      "loss": 0.5081,
+      "step": 1006
+    },
+    {
+      "epoch": 2.0019880715705765,
+      "grad_norm": 0.4265486606705852,
+      "learning_rate": 5.996793135318962e-05,
+      "loss": 0.4912,
+      "step": 1007
+    },
+    {
+      "epoch": 2.003976143141153,
+      "grad_norm": 0.41944427653109145,
+      "learning_rate": 5.991979632278206e-05,
+      "loss": 0.4852,
+      "step": 1008
+    },
+    {
+      "epoch": 2.00596421471173,
+      "grad_norm": 0.46970990607031543,
+      "learning_rate": 5.987162290260883e-05,
+      "loss": 0.4798,
+      "step": 1009
+    },
+    {
+      "epoch": 2.0079522862823063,
+      "grad_norm": 0.3494875624015793,
+      "learning_rate": 5.9823411185510546e-05,
+      "loss": 0.48,
+      "step": 1010
+    },
+    {
+      "epoch": 2.0099403578528827,
+      "grad_norm": 0.33653022283326256,
+      "learning_rate": 5.9775161264401636e-05,
+      "loss": 0.4841,
+      "step": 1011
+    },
+    {
+      "epoch": 2.011928429423459,
+      "grad_norm": 0.3886220441073982,
+      "learning_rate": 5.972687323227016e-05,
+      "loss": 0.4829,
+      "step": 1012
+    },
+    {
+      "epoch": 2.0139165009940356,
+      "grad_norm": 0.35168024551500693,
+      "learning_rate": 5.967854718217764e-05,
+      "loss": 0.4802,
+      "step": 1013
+    },
+    {
+      "epoch": 2.0159045725646125,
+      "grad_norm": 0.29225835757850427,
+      "learning_rate": 5.963018320725884e-05,
+      "loss": 0.4809,
+      "step": 1014
+    },
+    {
+      "epoch": 2.017892644135189,
+      "grad_norm": 0.2910719808738326,
+      "learning_rate": 5.9581781400721605e-05,
+      "loss": 0.4722,
+      "step": 1015
+    },
+    {
+      "epoch": 2.0198807157057654,
+      "grad_norm": 0.2961140641840679,
+      "learning_rate": 5.9533341855846745e-05,
+      "loss": 0.4749,
+      "step": 1016
+    },
+    {
+      "epoch": 2.021868787276342,
+      "grad_norm": 0.2297321304146483,
+      "learning_rate": 5.948486466598773e-05,
+      "loss": 0.479,
+      "step": 1017
+    },
+    {
+      "epoch": 2.0238568588469183,
+      "grad_norm": 0.276146779314003,
+      "learning_rate": 5.943634992457065e-05,
+      "loss": 0.4816,
+      "step": 1018
+    },
+    {
+      "epoch": 2.0258449304174952,
+      "grad_norm": 0.3894209253927516,
+      "learning_rate": 5.938779772509388e-05,
+      "loss": 0.4845,
+      "step": 1019
+    },
+    {
+      "epoch": 2.0278330019880717,
+      "grad_norm": 0.3399631549933991,
+      "learning_rate": 5.933920816112807e-05,
+      "loss": 0.4718,
+      "step": 1020
+    },
+    {
+      "epoch": 2.029821073558648,
+      "grad_norm": 0.34376727093560305,
+      "learning_rate": 5.929058132631584e-05,
+      "loss": 0.4817,
+      "step": 1021
+    },
+    {
+      "epoch": 2.0318091451292246,
+      "grad_norm": 0.39737386982055917,
+      "learning_rate": 5.924191731437162e-05,
+      "loss": 0.4827,
+      "step": 1022
+    },
+    {
+      "epoch": 2.033797216699801,
+      "grad_norm": 0.2551201885904131,
+      "learning_rate": 5.9193216219081546e-05,
+      "loss": 0.4813,
+      "step": 1023
+    },
+    {
+      "epoch": 2.0357852882703775,
+      "grad_norm": 0.25758071434624624,
+      "learning_rate": 5.914447813430315e-05,
+      "loss": 0.475,
+      "step": 1024
+    },
+    {
+      "epoch": 2.0377733598409544,
+      "grad_norm": 0.2655696299091699,
+      "learning_rate": 5.909570315396529e-05,
+      "loss": 0.4821,
+      "step": 1025
+    },
+    {
+      "epoch": 2.039761431411531,
+      "grad_norm": 0.3183828091106384,
+      "learning_rate": 5.904689137206796e-05,
+      "loss": 0.4801,
+      "step": 1026
+    },
+    {
+      "epoch": 2.0417495029821073,
+      "grad_norm": 0.3018621992324004,
+      "learning_rate": 5.8998042882682024e-05,
+      "loss": 0.4754,
+      "step": 1027
+    },
+    {
+      "epoch": 2.043737574552684,
+      "grad_norm": 0.23706986324039556,
+      "learning_rate": 5.8949157779949094e-05,
+      "loss": 0.4699,
+      "step": 1028
+    },
+    {
+      "epoch": 2.0457256461232602,
+      "grad_norm": 0.21273568862734663,
+      "learning_rate": 5.8900236158081373e-05,
+      "loss": 0.476,
+      "step": 1029
+    },
+    {
+      "epoch": 2.047713717693837,
+      "grad_norm": 0.31270371437555916,
+      "learning_rate": 5.8851278111361444e-05,
+      "loss": 0.4799,
+      "step": 1030
+    },
+    {
+      "epoch": 2.0497017892644136,
+      "grad_norm": 0.39052899581994754,
+      "learning_rate": 5.8802283734142074e-05,
+      "loss": 0.4818,
+      "step": 1031
+    },
+    {
+      "epoch": 2.05168986083499,
+      "grad_norm": 0.3025793126930875,
+      "learning_rate": 5.875325312084603e-05,
+      "loss": 0.467,
+      "step": 1032
+    },
+    {
+      "epoch": 2.0536779324055665,
+      "grad_norm": 0.22993823523500936,
+      "learning_rate": 5.870418636596594e-05,
+      "loss": 0.4769,
+      "step": 1033
+    },
+    {
+      "epoch": 2.055666003976143,
+      "grad_norm": 0.3656561691652938,
+      "learning_rate": 5.865508356406408e-05,
+      "loss": 0.4777,
+      "step": 1034
+    },
+    {
+      "epoch": 2.05765407554672,
+      "grad_norm": 0.4134849810413476,
+      "learning_rate": 5.86059448097722e-05,
+      "loss": 0.4775,
+      "step": 1035
+    },
+    {
+      "epoch": 2.0596421471172963,
+      "grad_norm": 0.29706263006774,
+      "learning_rate": 5.8556770197791334e-05,
+      "loss": 0.4759,
+      "step": 1036
+    },
+    {
+      "epoch": 2.0616302186878728,
+      "grad_norm": 0.4075218161502847,
+      "learning_rate": 5.850755982289159e-05,
+      "loss": 0.4794,
+      "step": 1037
+    },
+    {
+      "epoch": 2.0636182902584492,
+      "grad_norm": 0.33462269323112476,
+      "learning_rate": 5.845831377991206e-05,
+      "loss": 0.4812,
+      "step": 1038
+    },
+    {
+      "epoch": 2.0656063618290257,
+      "grad_norm": 0.4806307657616057,
+      "learning_rate": 5.8409032163760536e-05,
+      "loss": 0.4781,
+      "step": 1039
+    },
+    {
+      "epoch": 2.0675944333996026,
+      "grad_norm": 0.5577446854340833,
+      "learning_rate": 5.835971506941338e-05,
+      "loss": 0.4889,
+      "step": 1040
+    },
+    {
+      "epoch": 2.069582504970179,
+      "grad_norm": 0.5725622742873799,
+      "learning_rate": 5.83103625919153e-05,
+      "loss": 0.4829,
+      "step": 1041
+    },
+    {
+      "epoch": 2.0715705765407555,
+      "grad_norm": 0.45898725687359726,
+      "learning_rate": 5.826097482637924e-05,
+      "loss": 0.4723,
+      "step": 1042
+    },
+    {
+      "epoch": 2.073558648111332,
+      "grad_norm": 0.2939925335093428,
+      "learning_rate": 5.8211551867986126e-05,
+      "loss": 0.4766,
+      "step": 1043
+    },
+    {
+      "epoch": 2.0755467196819084,
+      "grad_norm": 0.3276561950082936,
+      "learning_rate": 5.8162093811984714e-05,
+      "loss": 0.481,
+      "step": 1044
+    },
+    {
+      "epoch": 2.0775347912524853,
+      "grad_norm": 0.36792672163421464,
+      "learning_rate": 5.8112600753691404e-05,
+      "loss": 0.4844,
+      "step": 1045
+    },
+    {
+      "epoch": 2.0795228628230618,
+      "grad_norm": 0.41267081081703366,
+      "learning_rate": 5.806307278849003e-05,
+      "loss": 0.4824,
+      "step": 1046
+    },
+    {
+      "epoch": 2.081510934393638,
+      "grad_norm": 0.28869818706409717,
+      "learning_rate": 5.8013510011831726e-05,
+      "loss": 0.4808,
+      "step": 1047
+    },
+    {
+      "epoch": 2.0834990059642147,
+      "grad_norm": 0.22647781008653403,
+      "learning_rate": 5.7963912519234716e-05,
+      "loss": 0.4671,
+      "step": 1048
+    },
+    {
+      "epoch": 2.085487077534791,
+      "grad_norm": 0.2643026256357433,
+      "learning_rate": 5.791428040628413e-05,
+      "loss": 0.4763,
+      "step": 1049
+    },
+    {
+      "epoch": 2.0874751491053676,
+      "grad_norm": 0.3184898788545838,
+      "learning_rate": 5.7864613768631764e-05,
+      "loss": 0.4704,
+      "step": 1050
+    },
+    {
+      "epoch": 2.0894632206759445,
+      "grad_norm": 0.3180159256253107,
+      "learning_rate": 5.781491270199604e-05,
+      "loss": 0.4766,
+      "step": 1051
+    },
+    {
+      "epoch": 2.091451292246521,
+      "grad_norm": 0.2508605320554848,
+      "learning_rate": 5.776517730216166e-05,
+      "loss": 0.4752,
+      "step": 1052
+    },
+    {
+      "epoch": 2.0934393638170974,
+      "grad_norm": 0.23712239688416356,
+      "learning_rate": 5.7715407664979545e-05,
+      "loss": 0.4774,
+      "step": 1053
+    },
+    {
+      "epoch": 2.095427435387674,
+      "grad_norm": 0.2527292992526699,
+      "learning_rate": 5.7665603886366556e-05,
+      "loss": 0.4758,
+      "step": 1054
+    },
+    {
+      "epoch": 2.0974155069582503,
+      "grad_norm": 0.27237694330213974,
+      "learning_rate": 5.761576606230538e-05,
+      "loss": 0.474,
+      "step": 1055
+    },
+    {
+      "epoch": 2.099403578528827,
+      "grad_norm": 0.3069196097337193,
+      "learning_rate": 5.75658942888443e-05,
+      "loss": 0.4788,
+      "step": 1056
+    },
+    {
+      "epoch": 2.1013916500994037,
+      "grad_norm": 0.24640187280522557,
+      "learning_rate": 5.751598866209704e-05,
+      "loss": 0.4766,
+      "step": 1057
+    },
+    {
+      "epoch": 2.10337972166998,
+      "grad_norm": 0.3103053687041832,
+      "learning_rate": 5.746604927824257e-05,
+      "loss": 0.4706,
+      "step": 1058
+    },
+    {
+      "epoch": 2.1053677932405566,
+      "grad_norm": 0.40603031427539615,
+      "learning_rate": 5.741607623352488e-05,
+      "loss": 0.4736,
+      "step": 1059
+    },
+    {
+      "epoch": 2.107355864811133,
+      "grad_norm": 0.4925667009481897,
+      "learning_rate": 5.736606962425288e-05,
+      "loss": 0.4812,
+      "step": 1060
+    },
+    {
+      "epoch": 2.10934393638171,
+      "grad_norm": 0.5061455538281211,
+      "learning_rate": 5.731602954680016e-05,
+      "loss": 0.4767,
+      "step": 1061
+    },
+    {
+      "epoch": 2.1113320079522864,
+      "grad_norm": 0.3629127454241806,
+      "learning_rate": 5.726595609760477e-05,
+      "loss": 0.4811,
+      "step": 1062
+    },
+    {
+      "epoch": 2.113320079522863,
+      "grad_norm": 0.26290247683922574,
+      "learning_rate": 5.721584937316912e-05,
+      "loss": 0.4827,
+      "step": 1063
+    },
+    {
+      "epoch": 2.1153081510934393,
+      "grad_norm": 0.2265400929052351,
+      "learning_rate": 5.716570947005972e-05,
+      "loss": 0.4734,
+      "step": 1064
+    },
+    {
+      "epoch": 2.1172962226640157,
+      "grad_norm": 0.28429428323409667,
+      "learning_rate": 5.7115536484907034e-05,
+      "loss": 0.4671,
+      "step": 1065
+    },
+    {
+      "epoch": 2.1192842942345926,
+      "grad_norm": 0.3535690679025093,
+      "learning_rate": 5.706533051440529e-05,
+      "loss": 0.4877,
+      "step": 1066
+    },
+    {
+      "epoch": 2.121272365805169,
+      "grad_norm": 0.4987602517860749,
+      "learning_rate": 5.701509165531226e-05,
+      "loss": 0.4828,
+      "step": 1067
+    },
+    {
+      "epoch": 2.1232604373757455,
+      "grad_norm": 0.44471007932552276,
+      "learning_rate": 5.696482000444913e-05,
+      "loss": 0.479,
+      "step": 1068
+    },
+    {
+      "epoch": 2.125248508946322,
+      "grad_norm": 0.7333534347504769,
+      "learning_rate": 5.691451565870026e-05,
+      "loss": 0.4833,
+      "step": 1069
+    },
+    {
+      "epoch": 2.1272365805168985,
+      "grad_norm": 0.41727447435676135,
+      "learning_rate": 5.686417871501304e-05,
+      "loss": 0.4772,
+      "step": 1070
+    },
+    {
+      "epoch": 2.1292246520874754,
+      "grad_norm": 0.4401221787273442,
+      "learning_rate": 5.681380927039767e-05,
+      "loss": 0.4746,
+      "step": 1071
+    },
+    {
+      "epoch": 2.131212723658052,
+      "grad_norm": 0.5208507936208158,
+      "learning_rate": 5.676340742192697e-05,
+      "loss": 0.4704,
+      "step": 1072
+    },
+    {
+      "epoch": 2.1332007952286283,
+      "grad_norm": 0.4273310201857551,
+      "learning_rate": 5.671297326673625e-05,
+      "loss": 0.4735,
+      "step": 1073
+    },
+    {
+      "epoch": 2.1351888667992047,
+      "grad_norm": 0.32801220821324406,
+      "learning_rate": 5.666250690202304e-05,
+      "loss": 0.4749,
+      "step": 1074
+    },
+    {
+      "epoch": 2.137176938369781,
+      "grad_norm": 0.3278924275873546,
+      "learning_rate": 5.661200842504698e-05,
+      "loss": 0.4756,
+      "step": 1075
+    },
+    {
+      "epoch": 2.139165009940358,
+      "grad_norm": 0.34059455870639094,
+      "learning_rate": 5.6561477933129605e-05,
+      "loss": 0.484,
+      "step": 1076
+    },
+    {
+      "epoch": 2.1411530815109345,
+      "grad_norm": 0.3902473929759777,
+      "learning_rate": 5.651091552365407e-05,
+      "loss": 0.4747,
+      "step": 1077
+    },
+    {
+      "epoch": 2.143141153081511,
+      "grad_norm": 0.45074247141393353,
+      "learning_rate": 5.646032129406514e-05,
+      "loss": 0.4738,
+      "step": 1078
+    },
+    {
+      "epoch": 2.1451292246520874,
+      "grad_norm": 0.5573194267206498,
+      "learning_rate": 5.640969534186886e-05,
+      "loss": 0.4848,
+      "step": 1079
+    },
+    {
+      "epoch": 2.147117296222664,
+      "grad_norm": 0.5620699864501703,
+      "learning_rate": 5.635903776463242e-05,
+      "loss": 0.4831,
+      "step": 1080
+    },
+    {
+      "epoch": 2.1491053677932404,
+      "grad_norm": 0.3560378056402117,
+      "learning_rate": 5.630834865998391e-05,
+      "loss": 0.4806,
+      "step": 1081
+    },
+    {
+      "epoch": 2.1510934393638173,
+      "grad_norm": 0.259702279452807,
+      "learning_rate": 5.625762812561226e-05,
+      "loss": 0.473,
+      "step": 1082
+    },
+    {
+      "epoch": 2.1530815109343937,
+      "grad_norm": 0.37619445513774913,
+      "learning_rate": 5.620687625926693e-05,
+      "loss": 0.4815,
+      "step": 1083
+    },
+    {
+      "epoch": 2.15506958250497,
+      "grad_norm": 0.46465125114676165,
+      "learning_rate": 5.6156093158757776e-05,
+      "loss": 0.4686,
+      "step": 1084
+    },
+    {
+      "epoch": 2.1570576540755466,
+      "grad_norm": 0.49932017268218537,
+      "learning_rate": 5.6105278921954834e-05,
+      "loss": 0.4756,
+      "step": 1085
+    },
+    {
+      "epoch": 2.159045725646123,
+      "grad_norm": 0.4587350037975631,
+      "learning_rate": 5.605443364678813e-05,
+      "loss": 0.4805,
+      "step": 1086
+    },
+    {
+      "epoch": 2.1610337972167,
+      "grad_norm": 0.2846640559400187,
+      "learning_rate": 5.600355743124756e-05,
+      "loss": 0.4666,
+      "step": 1087
+    },
+    {
+      "epoch": 2.1630218687872764,
+      "grad_norm": 0.2747078862140328,
+      "learning_rate": 5.595265037338262e-05,
+      "loss": 0.4741,
+      "step": 1088
+    },
+    {
+      "epoch": 2.165009940357853,
+      "grad_norm": 0.34597139882807243,
+      "learning_rate": 5.590171257130225e-05,
+      "loss": 0.4731,
+      "step": 1089
+    },
+    {
+      "epoch": 2.1669980119284293,
+      "grad_norm": 0.32077150910072605,
+      "learning_rate": 5.58507441231746e-05,
+      "loss": 0.4782,
+      "step": 1090
+    },
+    {
+      "epoch": 2.168986083499006,
+      "grad_norm": 0.325683453143863,
+      "learning_rate": 5.579974512722695e-05,
+      "loss": 0.4778,
+      "step": 1091
+    },
+    {
+      "epoch": 2.1709741550695827,
+      "grad_norm": 0.23366416594281797,
+      "learning_rate": 5.574871568174542e-05,
+      "loss": 0.4801,
+      "step": 1092
+    },
+    {
+      "epoch": 2.172962226640159,
+      "grad_norm": 0.2615690824569999,
+      "learning_rate": 5.569765588507481e-05,
+      "loss": 0.478,
+      "step": 1093
+    },
+    {
+      "epoch": 2.1749502982107356,
+      "grad_norm": 0.24438298458929783,
+      "learning_rate": 5.5646565835618414e-05,
+      "loss": 0.4732,
+      "step": 1094
+    },
+    {
+      "epoch": 2.176938369781312,
+      "grad_norm": 0.25924952804351686,
+      "learning_rate": 5.5595445631837826e-05,
+      "loss": 0.485,
+      "step": 1095
+    },
+    {
+      "epoch": 2.1789264413518885,
+      "grad_norm": 0.26261769213553204,
+      "learning_rate": 5.554429537225277e-05,
+      "loss": 0.4675,
+      "step": 1096
+    },
+    {
+      "epoch": 2.1809145129224654,
+      "grad_norm": 0.25706757072349795,
+      "learning_rate": 5.549311515544088e-05,
+      "loss": 0.4755,
+      "step": 1097
+    },
+    {
+      "epoch": 2.182902584493042,
+      "grad_norm": 0.263310508037158,
+      "learning_rate": 5.544190508003755e-05,
+      "loss": 0.479,
+      "step": 1098
+    },
+    {
+      "epoch": 2.1848906560636183,
+      "grad_norm": 0.282958409875485,
+      "learning_rate": 5.5390665244735644e-05,
+      "loss": 0.477,
+      "step": 1099
+    },
+    {
+      "epoch": 2.1868787276341948,
+      "grad_norm": 0.32353132229505105,
+      "learning_rate": 5.533939574828548e-05,
+      "loss": 0.4725,
+      "step": 1100
+    },
+    {
+      "epoch": 2.1888667992047712,
+      "grad_norm": 0.27942916331621664,
+      "learning_rate": 5.528809668949448e-05,
+      "loss": 0.4801,
+      "step": 1101
+    },
+    {
+      "epoch": 2.1908548707753477,
+      "grad_norm": 0.26257454184150586,
+      "learning_rate": 5.523676816722705e-05,
+      "loss": 0.4749,
+      "step": 1102
+    },
+    {
+      "epoch": 2.1928429423459246,
+      "grad_norm": 0.3092973711478264,
+      "learning_rate": 5.518541028040437e-05,
+      "loss": 0.4705,
+      "step": 1103
+    },
+    {
+      "epoch": 2.194831013916501,
+      "grad_norm": 0.289875960480621,
+      "learning_rate": 5.5134023128004236e-05,
+      "loss": 0.4698,
+      "step": 1104
+    },
+    {
+      "epoch": 2.1968190854870775,
+      "grad_norm": 0.3272973896667466,
+      "learning_rate": 5.508260680906081e-05,
+      "loss": 0.476,
+      "step": 1105
+    },
+    {
+      "epoch": 2.198807157057654,
+      "grad_norm": 0.3043857410112243,
+      "learning_rate": 5.50311614226645e-05,
+      "loss": 0.4781,
+      "step": 1106
+    },
+    {
+      "epoch": 2.2007952286282304,
+      "grad_norm": 0.25728137972199694,
+      "learning_rate": 5.497968706796173e-05,
+      "loss": 0.4788,
+      "step": 1107
+    },
+    {
+      "epoch": 2.2027833001988073,
+      "grad_norm": 0.2496025926087033,
+      "learning_rate": 5.49281838441547e-05,
+      "loss": 0.4689,
+      "step": 1108
+    },
+    {
+      "epoch": 2.2047713717693838,
+      "grad_norm": 0.2552013584123278,
+      "learning_rate": 5.4876651850501325e-05,
+      "loss": 0.4721,
+      "step": 1109
+    },
+    {
+      "epoch": 2.20675944333996,
+      "grad_norm": 0.2508265413893804,
+      "learning_rate": 5.482509118631491e-05,
+      "loss": 0.47,
+      "step": 1110
+    },
+    {
+      "epoch": 2.2087475149105367,
+      "grad_norm": 0.31405764647416207,
+      "learning_rate": 5.4773501950964055e-05,
+      "loss": 0.4795,
+      "step": 1111
+    },
+    {
+      "epoch": 2.210735586481113,
+      "grad_norm": 0.28939539182682655,
+      "learning_rate": 5.472188424387236e-05,
+      "loss": 0.4703,
+      "step": 1112
+    },
+    {
+      "epoch": 2.21272365805169,
+      "grad_norm": 0.21249749698185788,
+      "learning_rate": 5.467023816451836e-05,
+      "loss": 0.4804,
+      "step": 1113
+    },
+    {
+      "epoch": 2.2147117296222665,
+      "grad_norm": 0.3614011385798107,
+      "learning_rate": 5.461856381243524e-05,
+      "loss": 0.4747,
+      "step": 1114
+    },
+    {
+      "epoch": 2.216699801192843,
+      "grad_norm": 0.21423784260449702,
+      "learning_rate": 5.456686128721069e-05,
+      "loss": 0.4725,
+      "step": 1115
+    },
+    {
+      "epoch": 2.2186878727634194,
+      "grad_norm": 0.3235629875050971,
+      "learning_rate": 5.451513068848668e-05,
+      "loss": 0.4772,
+      "step": 1116
+    },
+    {
+      "epoch": 2.220675944333996,
+      "grad_norm": 0.25439106188409116,
+      "learning_rate": 5.4463372115959266e-05,
+      "loss": 0.4726,
+      "step": 1117
+    },
+    {
+      "epoch": 2.2226640159045727,
+      "grad_norm": 0.2590887452156136,
+      "learning_rate": 5.4411585669378464e-05,
+      "loss": 0.4732,
+      "step": 1118
+    },
+    {
+      "epoch": 2.224652087475149,
+      "grad_norm": 0.26919093398382715,
+      "learning_rate": 5.4359771448547966e-05,
+      "loss": 0.4728,
+      "step": 1119
+    },
+    {
+      "epoch": 2.2266401590457257,
+      "grad_norm": 0.36365985427405484,
+      "learning_rate": 5.430792955332503e-05,
+      "loss": 0.4732,
+      "step": 1120
+    },
+    {
+      "epoch": 2.228628230616302,
+      "grad_norm": 0.3915941810323402,
+      "learning_rate": 5.425606008362019e-05,
+      "loss": 0.4781,
+      "step": 1121
+    },
+    {
+      "epoch": 2.2306163021868786,
+      "grad_norm": 0.2610839854362163,
+      "learning_rate": 5.420416313939719e-05,
+      "loss": 0.4831,
+      "step": 1122
+    },
+    {
+      "epoch": 2.2326043737574555,
+      "grad_norm": 0.4388412247539103,
+      "learning_rate": 5.415223882067269e-05,
+      "loss": 0.4973,
+      "step": 1123
+    },
+    {
+      "epoch": 2.234592445328032,
+      "grad_norm": 0.2698215651291218,
+      "learning_rate": 5.410028722751609e-05,
+      "loss": 0.476,
+      "step": 1124
+    },
+    {
+      "epoch": 2.2365805168986084,
+      "grad_norm": 0.26389275172674886,
+      "learning_rate": 5.404830846004938e-05,
+      "loss": 0.4703,
+      "step": 1125
+    },
+    {
+      "epoch": 2.238568588469185,
+      "grad_norm": 0.3120918809111035,
+      "learning_rate": 5.3996302618446906e-05,
+      "loss": 0.4769,
+      "step": 1126
+    },
+    {
+      "epoch": 2.2405566600397613,
+      "grad_norm": 0.32108060374276975,
+      "learning_rate": 5.394426980293522e-05,
+      "loss": 0.4669,
+      "step": 1127
+    },
+    {
+      "epoch": 2.242544731610338,
+      "grad_norm": 0.36707667301171554,
+      "learning_rate": 5.389221011379281e-05,
+      "loss": 0.4715,
+      "step": 1128
+    },
+    {
+      "epoch": 2.2445328031809146,
+      "grad_norm": 0.3816486880160509,
+      "learning_rate": 5.384012365135002e-05,
+      "loss": 0.4767,
+      "step": 1129
+    },
+    {
+      "epoch": 2.246520874751491,
+      "grad_norm": 0.46296188269694993,
+      "learning_rate": 5.3788010515988695e-05,
+      "loss": 0.4702,
+      "step": 1130
+    },
+    {
+      "epoch": 2.2485089463220675,
+      "grad_norm": 0.4896953557986629,
+      "learning_rate": 5.37358708081422e-05,
+      "loss": 0.4685,
+      "step": 1131
+    },
+    {
+      "epoch": 2.250497017892644,
+      "grad_norm": 0.43869657749170266,
+      "learning_rate": 5.368370462829503e-05,
+      "loss": 0.465,
+      "step": 1132
+    },
+    {
+      "epoch": 2.252485089463221,
+      "grad_norm": 0.5371500014095143,
+      "learning_rate": 5.363151207698273e-05,
+      "loss": 0.4706,
+      "step": 1133
+    },
+    {
+      "epoch": 2.2544731610337974,
+      "grad_norm": 0.5863174393137796,
+      "learning_rate": 5.357929325479164e-05,
+      "loss": 0.4746,
+      "step": 1134
+    },
+    {
+      "epoch": 2.256461232604374,
+      "grad_norm": 0.39887424277932715,
+      "learning_rate": 5.3527048262358774e-05,
+      "loss": 0.4712,
+      "step": 1135
+    },
+    {
+      "epoch": 2.2584493041749503,
+      "grad_norm": 0.30753503519660036,
+      "learning_rate": 5.347477720037156e-05,
+      "loss": 0.4745,
+      "step": 1136
+    },
+    {
+      "epoch": 2.2604373757455267,
+      "grad_norm": 0.45617523478874633,
+      "learning_rate": 5.342248016956765e-05,
+      "loss": 0.4696,
+      "step": 1137
+    },
+    {
+      "epoch": 2.2624254473161036,
+      "grad_norm": 0.5535883206642056,
+      "learning_rate": 5.3370157270734764e-05,
+      "loss": 0.4705,
+      "step": 1138
+    },
+    {
+      "epoch": 2.26441351888668,
+      "grad_norm": 0.4652282323599418,
+      "learning_rate": 5.331780860471046e-05,
+      "loss": 0.4743,
+      "step": 1139
+    },
+    {
+      "epoch": 2.2664015904572565,
+      "grad_norm": 0.36379549274018436,
+      "learning_rate": 5.326543427238199e-05,
+      "loss": 0.4726,
+      "step": 1140
+    },
+    {
+      "epoch": 2.268389662027833,
+      "grad_norm": 0.3171847241802109,
+      "learning_rate": 5.321303437468603e-05,
+      "loss": 0.4655,
+      "step": 1141
+    },
+    {
+      "epoch": 2.2703777335984094,
+      "grad_norm": 0.2881646521576697,
+      "learning_rate": 5.3160609012608545e-05,
+      "loss": 0.4633,
+      "step": 1142
+    },
+    {
+      "epoch": 2.272365805168986,
+      "grad_norm": 0.29641268848872576,
+      "learning_rate": 5.310815828718457e-05,
+      "loss": 0.4708,
+      "step": 1143
+    },
+    {
+      "epoch": 2.274353876739563,
+      "grad_norm": 0.40275523394654794,
+      "learning_rate": 5.305568229949802e-05,
+      "loss": 0.4673,
+      "step": 1144
+    },
+    {
+      "epoch": 2.2763419483101393,
+      "grad_norm": 0.35883228801414907,
+      "learning_rate": 5.3003181150681506e-05,
+      "loss": 0.4712,
+      "step": 1145
+    },
+    {
+      "epoch": 2.2783300198807157,
+      "grad_norm": 0.2687510267313907,
+      "learning_rate": 5.295065494191612e-05,
+      "loss": 0.468,
+      "step": 1146
+    },
+    {
+      "epoch": 2.280318091451292,
+      "grad_norm": 0.2892946413120926,
+      "learning_rate": 5.289810377443127e-05,
+      "loss": 0.473,
+      "step": 1147
+    },
+    {
+      "epoch": 2.2823061630218686,
+      "grad_norm": 0.2943293689076449,
+      "learning_rate": 5.284552774950441e-05,
+      "loss": 0.4711,
+      "step": 1148
+    },
+    {
+      "epoch": 2.2842942345924455,
+      "grad_norm": 0.2434140998108816,
+      "learning_rate": 5.279292696846096e-05,
+      "loss": 0.4687,
+      "step": 1149
+    },
+    {
+      "epoch": 2.286282306163022,
+      "grad_norm": 0.22090774927042578,
+      "learning_rate": 5.274030153267403e-05,
+      "loss": 0.4756,
+      "step": 1150
+    },
+    {
+      "epoch": 2.2882703777335984,
+      "grad_norm": 0.2475754664083401,
+      "learning_rate": 5.268765154356425e-05,
+      "loss": 0.4748,
+      "step": 1151
+    },
+    {
+      "epoch": 2.290258449304175,
+      "grad_norm": 0.24994849706871775,
+      "learning_rate": 5.263497710259953e-05,
+      "loss": 0.4737,
+      "step": 1152
+    },
+    {
+      "epoch": 2.2922465208747513,
+      "grad_norm": 0.2215857652666882,
+      "learning_rate": 5.2582278311294975e-05,
+      "loss": 0.4639,
+      "step": 1153
+    },
+    {
+      "epoch": 2.294234592445328,
+      "grad_norm": 0.2170348912725335,
+      "learning_rate": 5.252955527121255e-05,
+      "loss": 0.4733,
+      "step": 1154
+    },
+    {
+      "epoch": 2.2962226640159047,
+      "grad_norm": 0.21834748510730956,
+      "learning_rate": 5.2476808083961e-05,
+      "loss": 0.4635,
+      "step": 1155
+    },
+    {
+      "epoch": 2.298210735586481,
+      "grad_norm": 0.2351001402329863,
+      "learning_rate": 5.242403685119559e-05,
+      "loss": 0.4729,
+      "step": 1156
+    },
+    {
+      "epoch": 2.3001988071570576,
+      "grad_norm": 0.5861647894402029,
+      "learning_rate": 5.2371241674617924e-05,
+      "loss": 0.4885,
+      "step": 1157
+    },
+    {
+      "epoch": 2.302186878727634,
+      "grad_norm": 0.26021493801320544,
+      "learning_rate": 5.231842265597573e-05,
+      "loss": 0.4698,
+      "step": 1158
+    },
+    {
+      "epoch": 2.3041749502982105,
+      "grad_norm": 0.33674055818667714,
+      "learning_rate": 5.2265579897062724e-05,
+      "loss": 0.4695,
+      "step": 1159
+    },
+    {
+      "epoch": 2.3061630218687874,
+      "grad_norm": 0.30732795040623506,
+      "learning_rate": 5.2212713499718374e-05,
+      "loss": 0.4682,
+      "step": 1160
+    },
+    {
+      "epoch": 2.308151093439364,
+      "grad_norm": 0.2351863418623793,
+      "learning_rate": 5.2159823565827656e-05,
+      "loss": 0.4692,
+      "step": 1161
+    },
+    {
+      "epoch": 2.3101391650099403,
+      "grad_norm": 0.24745733032782763,
+      "learning_rate": 5.210691019732096e-05,
+      "loss": 0.4676,
+      "step": 1162
+    },
+    {
+      "epoch": 2.3121272365805168,
+      "grad_norm": 0.21317022863905552,
+      "learning_rate": 5.2053973496173814e-05,
+      "loss": 0.4708,
+      "step": 1163
+    },
+    {
+      "epoch": 2.3141153081510932,
+      "grad_norm": 0.27443341498457285,
+      "learning_rate": 5.2001013564406704e-05,
+      "loss": 0.4704,
+      "step": 1164
+    },
+    {
+      "epoch": 2.31610337972167,
+      "grad_norm": 0.2621084832032726,
+      "learning_rate": 5.194803050408491e-05,
+      "loss": 0.4673,
+      "step": 1165
+    },
+    {
+      "epoch": 2.3180914512922466,
+      "grad_norm": 0.23873064722197226,
+      "learning_rate": 5.1895024417318275e-05,
+      "loss": 0.4702,
+      "step": 1166
+    },
+    {
+      "epoch": 2.320079522862823,
+      "grad_norm": 0.2860158902919176,
+      "learning_rate": 5.184199540626102e-05,
+      "loss": 0.4741,
+      "step": 1167
+    },
+    {
+      "epoch": 2.3220675944333995,
+      "grad_norm": 0.3120161852959375,
+      "learning_rate": 5.178894357311155e-05,
+      "loss": 0.4651,
+      "step": 1168
+    },
+    {
+      "epoch": 2.324055666003976,
+      "grad_norm": 0.2686323464939365,
+      "learning_rate": 5.173586902011224e-05,
+      "loss": 0.4694,
+      "step": 1169
+    },
+    {
+      "epoch": 2.326043737574553,
+      "grad_norm": 0.25928705925335244,
+      "learning_rate": 5.168277184954924e-05,
+      "loss": 0.4615,
+      "step": 1170
+    },
+    {
+      "epoch": 2.3280318091451293,
+      "grad_norm": 0.2401126856878729,
+      "learning_rate": 5.1629652163752326e-05,
+      "loss": 0.4675,
+      "step": 1171
+    },
+    {
+      "epoch": 2.3300198807157058,
+      "grad_norm": 0.23622257410914968,
+      "learning_rate": 5.157651006509467e-05,
+      "loss": 0.4738,
+      "step": 1172
+    },
+    {
+      "epoch": 2.332007952286282,
+      "grad_norm": 0.2398042204246071,
+      "learning_rate": 5.152334565599257e-05,
+      "loss": 0.4685,
+      "step": 1173
+    },
+    {
+      "epoch": 2.3339960238568587,
+      "grad_norm": 0.3099714198955679,
+      "learning_rate": 5.147015903890541e-05,
+      "loss": 0.4734,
+      "step": 1174
+    },
+    {
+      "epoch": 2.3359840954274356,
+      "grad_norm": 0.2410518103522495,
+      "learning_rate": 5.1416950316335305e-05,
+      "loss": 0.4692,
+      "step": 1175
+    },
+    {
+      "epoch": 2.337972166998012,
+      "grad_norm": 0.22715511588645854,
+      "learning_rate": 5.1363719590827005e-05,
+      "loss": 0.4723,
+      "step": 1176
+    },
+    {
+      "epoch": 2.3399602385685885,
+      "grad_norm": 0.23992255270613838,
+      "learning_rate": 5.131046696496766e-05,
+      "loss": 0.4682,
+      "step": 1177
+    },
+    {
+      "epoch": 2.341948310139165,
+      "grad_norm": 0.2148629396505402,
+      "learning_rate": 5.1257192541386617e-05,
+      "loss": 0.4715,
+      "step": 1178
+    },
+    {
+      "epoch": 2.3439363817097414,
+      "grad_norm": 0.21680845387800704,
+      "learning_rate": 5.120389642275525e-05,
+      "loss": 0.4668,
+      "step": 1179
+    },
+    {
+      "epoch": 2.3459244532803183,
+      "grad_norm": 0.24226307278763232,
+      "learning_rate": 5.115057871178672e-05,
+      "loss": 0.4618,
+      "step": 1180
+    },
+    {
+      "epoch": 2.3479125248508947,
+      "grad_norm": 0.2861167812722164,
+      "learning_rate": 5.1097239511235825e-05,
+      "loss": 0.4723,
+      "step": 1181
+    },
+    {
+      "epoch": 2.349900596421471,
+      "grad_norm": 0.27609528840749975,
+      "learning_rate": 5.104387892389877e-05,
+      "loss": 0.467,
+      "step": 1182
+    },
+    {
+      "epoch": 2.3518886679920477,
+      "grad_norm": 0.38245826721422277,
+      "learning_rate": 5.099049705261294e-05,
+      "loss": 0.4755,
+      "step": 1183
+    },
+    {
+      "epoch": 2.353876739562624,
+      "grad_norm": 0.29912531232454764,
+      "learning_rate": 5.093709400025682e-05,
+      "loss": 0.4721,
+      "step": 1184
+    },
+    {
+      "epoch": 2.355864811133201,
+      "grad_norm": 0.2502048153204002,
+      "learning_rate": 5.088366986974965e-05,
+      "loss": 0.4667,
+      "step": 1185
+    },
+    {
+      "epoch": 2.3578528827037775,
+      "grad_norm": 0.2203348693194436,
+      "learning_rate": 5.0830224764051304e-05,
+      "loss": 0.4663,
+      "step": 1186
+    },
+    {
+      "epoch": 2.359840954274354,
+      "grad_norm": 0.3196618924057666,
+      "learning_rate": 5.0776758786162087e-05,
+      "loss": 0.4702,
+      "step": 1187
+    },
+    {
+      "epoch": 2.3618290258449304,
+      "grad_norm": 0.2756931543786987,
+      "learning_rate": 5.072327203912254e-05,
+      "loss": 0.4605,
+      "step": 1188
+    },
+    {
+      "epoch": 2.363817097415507,
+      "grad_norm": 0.20491551253798782,
+      "learning_rate": 5.066976462601321e-05,
+      "loss": 0.4665,
+      "step": 1189
+    },
+    {
+      "epoch": 2.3658051689860837,
+      "grad_norm": 0.2784356458213145,
+      "learning_rate": 5.0616236649954486e-05,
+      "loss": 0.4707,
+      "step": 1190
+    },
+    {
+      "epoch": 2.36779324055666,
+      "grad_norm": 0.3183004798915327,
+      "learning_rate": 5.056268821410638e-05,
+      "loss": 0.4727,
+      "step": 1191
+    },
+    {
+      "epoch": 2.3697813121272366,
+      "grad_norm": 0.42961542435749583,
+      "learning_rate": 5.0509119421668315e-05,
+      "loss": 0.4751,
+      "step": 1192
+    },
+    {
+      "epoch": 2.371769383697813,
+      "grad_norm": 0.3334202159622228,
+      "learning_rate": 5.0455530375879e-05,
+      "loss": 0.4743,
+      "step": 1193
+    },
+    {
+      "epoch": 2.3737574552683895,
+      "grad_norm": 0.41604702703185387,
+      "learning_rate": 5.040192118001614e-05,
+      "loss": 0.4687,
+      "step": 1194
+    },
+    {
+      "epoch": 2.3757455268389664,
+      "grad_norm": 0.4822424655743456,
+      "learning_rate": 5.034829193739625e-05,
+      "loss": 0.4729,
+      "step": 1195
+    },
+    {
+      "epoch": 2.377733598409543,
+      "grad_norm": 0.3387106786694204,
+      "learning_rate": 5.029464275137453e-05,
+      "loss": 0.4665,
+      "step": 1196
+    },
+    {
+      "epoch": 2.3797216699801194,
+      "grad_norm": 0.3595173653277191,
+      "learning_rate": 5.0240973725344584e-05,
+      "loss": 0.4794,
+      "step": 1197
+    },
+    {
+      "epoch": 2.381709741550696,
+      "grad_norm": 0.46424663733541854,
+      "learning_rate": 5.0187284962738254e-05,
+      "loss": 0.4742,
+      "step": 1198
+    },
+    {
+      "epoch": 2.3836978131212723,
+      "grad_norm": 0.3198119518435221,
+      "learning_rate": 5.0133576567025436e-05,
+      "loss": 0.479,
+      "step": 1199
+    },
+    {
+      "epoch": 2.3856858846918487,
+      "grad_norm": 0.32264968680235223,
+      "learning_rate": 5.007984864171384e-05,
+      "loss": 0.4755,
+      "step": 1200
+    },
+    {
+      "epoch": 2.3876739562624256,
+      "grad_norm": 0.2680481041894317,
+      "learning_rate": 5.002610129034883e-05,
+      "loss": 0.4682,
+      "step": 1201
+    },
+    {
+      "epoch": 2.389662027833002,
+      "grad_norm": 0.24662752780204666,
+      "learning_rate": 4.99723346165132e-05,
+      "loss": 0.4733,
+      "step": 1202
+    },
+    {
+      "epoch": 2.3916500994035785,
+      "grad_norm": 0.2585957373284863,
+      "learning_rate": 4.991854872382699e-05,
+      "loss": 0.4779,
+      "step": 1203
+    },
+    {
+      "epoch": 2.393638170974155,
+      "grad_norm": 0.23387709556355635,
+      "learning_rate": 4.9864743715947275e-05,
+      "loss": 0.4764,
+      "step": 1204
+    },
+    {
+      "epoch": 2.3956262425447314,
+      "grad_norm": 0.2909207043279319,
+      "learning_rate": 4.9810919696567966e-05,
+      "loss": 0.4763,
+      "step": 1205
+    },
+    {
+      "epoch": 2.3976143141153083,
+      "grad_norm": 0.24287573819888061,
+      "learning_rate": 4.9757076769419616e-05,
+      "loss": 0.474,
+      "step": 1206
+    },
+    {
+      "epoch": 2.399602385685885,
+      "grad_norm": 0.26216025207203086,
+      "learning_rate": 4.970321503826922e-05,
+      "loss": 0.4804,
+      "step": 1207
+    },
+    {
+      "epoch": 2.4015904572564613,
+      "grad_norm": 0.2677040788546583,
+      "learning_rate": 4.964933460692001e-05,
+      "loss": 0.4783,
+      "step": 1208
+    },
+    {
+      "epoch": 2.4035785288270377,
+      "grad_norm": 0.2552777608049991,
+      "learning_rate": 4.959543557921126e-05,
+      "loss": 0.473,
+      "step": 1209
+    },
+    {
+      "epoch": 2.405566600397614,
+      "grad_norm": 0.22672195489552985,
+      "learning_rate": 4.9541518059018073e-05,
+      "loss": 0.4702,
+      "step": 1210
+    },
+    {
+      "epoch": 2.4075546719681906,
+      "grad_norm": 0.2515384275613973,
+      "learning_rate": 4.94875821502512e-05,
+      "loss": 0.466,
+      "step": 1211
+    },
+    {
+      "epoch": 2.4095427435387675,
+      "grad_norm": 0.24081500343214746,
+      "learning_rate": 4.9433627956856835e-05,
+      "loss": 0.4731,
+      "step": 1212
+    },
+    {
+      "epoch": 2.411530815109344,
+      "grad_norm": 0.23961419355896613,
+      "learning_rate": 4.9379655582816396e-05,
+      "loss": 0.4644,
+      "step": 1213
+    },
+    {
+      "epoch": 2.4135188866799204,
+      "grad_norm": 0.2378021223509304,
+      "learning_rate": 4.9325665132146354e-05,
+      "loss": 0.4659,
+      "step": 1214
+    },
+    {
+      "epoch": 2.415506958250497,
+      "grad_norm": 0.2803020254737336,
+      "learning_rate": 4.927165670889801e-05,
+      "loss": 0.4731,
+      "step": 1215
+    },
+    {
+      "epoch": 2.4174950298210733,
+      "grad_norm": 0.31478058063257236,
+      "learning_rate": 4.9217630417157306e-05,
+      "loss": 0.4691,
+      "step": 1216
+    },
+    {
+      "epoch": 2.4194831013916502,
+      "grad_norm": 0.27040398082481854,
+      "learning_rate": 4.9163586361044615e-05,
+      "loss": 0.4739,
+      "step": 1217
+    },
+    {
+      "epoch": 2.4214711729622267,
+      "grad_norm": 0.22978169619172967,
+      "learning_rate": 4.910952464471455e-05,
+      "loss": 0.4738,
+      "step": 1218
+    },
+    {
+      "epoch": 2.423459244532803,
+      "grad_norm": 0.3003478653396795,
+      "learning_rate": 4.905544537235575e-05,
+      "loss": 0.4768,
+      "step": 1219
+    },
+    {
+      "epoch": 2.4254473161033796,
+      "grad_norm": 0.3230073628448877,
+      "learning_rate": 4.9001348648190715e-05,
+      "loss": 0.4724,
+      "step": 1220
+    },
+    {
+      "epoch": 2.427435387673956,
+      "grad_norm": 0.38849544363848615,
+      "learning_rate": 4.894723457647554e-05,
+      "loss": 0.4813,
+      "step": 1221
+    },
+    {
+      "epoch": 2.429423459244533,
+      "grad_norm": 0.355162411969774,
+      "learning_rate": 4.889310326149979e-05,
+      "loss": 0.4696,
+      "step": 1222
+    },
+    {
+      "epoch": 2.4314115308151094,
+      "grad_norm": 0.3414594723704106,
+      "learning_rate": 4.883895480758621e-05,
+      "loss": 0.4706,
+      "step": 1223
+    },
+    {
+      "epoch": 2.433399602385686,
+      "grad_norm": 0.30815325615464595,
+      "learning_rate": 4.8784789319090644e-05,
+      "loss": 0.4667,
+      "step": 1224
+    },
+    {
+      "epoch": 2.4353876739562623,
+      "grad_norm": 0.341796797546769,
+      "learning_rate": 4.8730606900401715e-05,
+      "loss": 0.4756,
+      "step": 1225
+    },
+    {
+      "epoch": 2.4373757455268388,
+      "grad_norm": 0.47306285227016065,
+      "learning_rate": 4.8676407655940694e-05,
+      "loss": 0.4726,
+      "step": 1226
+    },
+    {
+      "epoch": 2.4393638170974157,
+      "grad_norm": 0.3497635037514457,
+      "learning_rate": 4.862219169016127e-05,
+      "loss": 0.4709,
+      "step": 1227
+    },
+    {
+      "epoch": 2.441351888667992,
+      "grad_norm": 0.4504529577189983,
+      "learning_rate": 4.856795910754937e-05,
+      "loss": 0.4654,
+      "step": 1228
+    },
+    {
+      "epoch": 2.4433399602385686,
+      "grad_norm": 0.34689558387001274,
+      "learning_rate": 4.851371001262292e-05,
+      "loss": 0.4717,
+      "step": 1229
+    },
+    {
+      "epoch": 2.445328031809145,
+      "grad_norm": 0.2872733269653433,
+      "learning_rate": 4.84594445099317e-05,
+      "loss": 0.4733,
+      "step": 1230
+    },
+    {
+      "epoch": 2.4473161033797215,
+      "grad_norm": 0.27243647138976984,
+      "learning_rate": 4.840516270405712e-05,
+      "loss": 0.4742,
+      "step": 1231
+    },
+    {
+      "epoch": 2.4493041749502984,
+      "grad_norm": 0.30305149495329026,
+      "learning_rate": 4.8350864699611924e-05,
+      "loss": 0.4626,
+      "step": 1232
+    },
+    {
+      "epoch": 2.451292246520875,
+      "grad_norm": 0.27978195741025513,
+      "learning_rate": 4.8296550601240186e-05,
+      "loss": 0.4753,
+      "step": 1233
+    },
+    {
+      "epoch": 2.4532803180914513,
+      "grad_norm": 0.29267098699687766,
+      "learning_rate": 4.8242220513616945e-05,
+      "loss": 0.471,
+      "step": 1234
+    },
+    {
+      "epoch": 2.4552683896620278,
+      "grad_norm": 0.23437662644156787,
+      "learning_rate": 4.8187874541448046e-05,
+      "loss": 0.4688,
+      "step": 1235
+    },
+    {
+      "epoch": 2.457256461232604,
+      "grad_norm": 0.23103521775278718,
+      "learning_rate": 4.8133512789469945e-05,
+      "loss": 0.4823,
+      "step": 1236
+    },
+    {
+      "epoch": 2.459244532803181,
+      "grad_norm": 0.20865348483141843,
+      "learning_rate": 4.807913536244955e-05,
+      "loss": 0.4743,
+      "step": 1237
+    },
+    {
+      "epoch": 2.4612326043737576,
+      "grad_norm": 0.3100863032015534,
+      "learning_rate": 4.8024742365183946e-05,
+      "loss": 0.4706,
+      "step": 1238
+    },
+    {
+      "epoch": 2.463220675944334,
+      "grad_norm": 0.27950183594554934,
+      "learning_rate": 4.797033390250021e-05,
+      "loss": 0.4634,
+      "step": 1239
+    },
+    {
+      "epoch": 2.4652087475149105,
+      "grad_norm": 0.25304117732962556,
+      "learning_rate": 4.791591007925526e-05,
+      "loss": 0.4638,
+      "step": 1240
+    },
+    {
+      "epoch": 2.467196819085487,
+      "grad_norm": 0.28945569502035434,
+      "learning_rate": 4.7861471000335604e-05,
+      "loss": 0.4744,
+      "step": 1241
+    },
+    {
+      "epoch": 2.469184890656064,
+      "grad_norm": 0.23301689545756774,
+      "learning_rate": 4.780701677065714e-05,
+      "loss": 0.4689,
+      "step": 1242
+    },
+    {
+      "epoch": 2.4711729622266403,
+      "grad_norm": 0.2884587227277944,
+      "learning_rate": 4.775254749516497e-05,
+      "loss": 0.4713,
+      "step": 1243
+    },
+    {
+      "epoch": 2.4731610337972167,
+      "grad_norm": 0.31338204568964245,
+      "learning_rate": 4.76980632788332e-05,
+      "loss": 0.4797,
+      "step": 1244
+    },
+    {
+      "epoch": 2.475149105367793,
+      "grad_norm": 0.29265797464452964,
+      "learning_rate": 4.764356422666473e-05,
+      "loss": 0.4726,
+      "step": 1245
+    },
+    {
+      "epoch": 2.4771371769383697,
+      "grad_norm": 0.23282823698357283,
+      "learning_rate": 4.758905044369105e-05,
+      "loss": 0.4671,
+      "step": 1246
+    },
+    {
+      "epoch": 2.4791252485089466,
+      "grad_norm": 0.2929501814614166,
+      "learning_rate": 4.753452203497202e-05,
+      "loss": 0.473,
+      "step": 1247
+    },
+    {
+      "epoch": 2.481113320079523,
+      "grad_norm": 0.297087580600837,
+      "learning_rate": 4.747997910559573e-05,
+      "loss": 0.4698,
+      "step": 1248
+    },
+    {
+      "epoch": 2.4831013916500995,
+      "grad_norm": 0.83213556842206,
+      "learning_rate": 4.7425421760678195e-05,
+      "loss": 0.4798,
+      "step": 1249
+    },
+    {
+      "epoch": 2.485089463220676,
+      "grad_norm": 0.4143963209466649,
+      "learning_rate": 4.737085010536328e-05,
+      "loss": 0.47,
+      "step": 1250
+    },
+    {
+      "epoch": 2.4870775347912524,
+      "grad_norm": 0.3331788869276754,
+      "learning_rate": 4.731626424482238e-05,
+      "loss": 0.473,
+      "step": 1251
+    },
+    {
+      "epoch": 2.4890656063618293,
+      "grad_norm": 0.29929653590699296,
+      "learning_rate": 4.726166428425427e-05,
+      "loss": 0.4771,
+      "step": 1252
+    },
+    {
+      "epoch": 2.4910536779324057,
+      "grad_norm": 0.34503964283270694,
+      "learning_rate": 4.720705032888492e-05,
+      "loss": 0.4672,
+      "step": 1253
+    },
+    {
+      "epoch": 2.493041749502982,
+      "grad_norm": 0.3732218590332561,
+      "learning_rate": 4.715242248396728e-05,
+      "loss": 0.471,
+      "step": 1254
+    },
+    {
+      "epoch": 2.4950298210735586,
+      "grad_norm": 0.3583811530564549,
+      "learning_rate": 4.709778085478101e-05,
+      "loss": 0.4761,
+      "step": 1255
+    },
+    {
+      "epoch": 2.497017892644135,
+      "grad_norm": 0.3230101452445875,
+      "learning_rate": 4.70431255466324e-05,
+      "loss": 0.4614,
+      "step": 1256
+    },
+    {
+      "epoch": 2.4990059642147116,
+      "grad_norm": 0.28794588330894255,
+      "learning_rate": 4.698845666485408e-05,
+      "loss": 0.4699,
+      "step": 1257
+    },
+    {
+      "epoch": 2.500994035785288,
+      "grad_norm": 0.3780763497100984,
+      "learning_rate": 4.693377431480482e-05,
+      "loss": 0.4694,
+      "step": 1258
+    },
+    {
+      "epoch": 2.502982107355865,
+      "grad_norm": 0.45136607030451903,
+      "learning_rate": 4.6879078601869374e-05,
+      "loss": 0.4747,
+      "step": 1259
+    },
+    {
+      "epoch": 2.5049701789264414,
+      "grad_norm": 0.24953986065483028,
+      "learning_rate": 4.682436963145823e-05,
+      "loss": 0.4685,
+      "step": 1260
+    },
+    {
+      "epoch": 2.506958250497018,
+      "grad_norm": 0.336007300317526,
+      "learning_rate": 4.6769647509007443e-05,
+      "loss": 0.4738,
+      "step": 1261
+    },
+    {
+      "epoch": 2.5089463220675943,
+      "grad_norm": 0.33628369876467346,
+      "learning_rate": 4.671491233997841e-05,
+      "loss": 0.468,
+      "step": 1262
+    },
+    {
+      "epoch": 2.5109343936381707,
+      "grad_norm": 0.2711421941817858,
+      "learning_rate": 4.666016422985764e-05,
+      "loss": 0.4701,
+      "step": 1263
+    },
+    {
+      "epoch": 2.5129224652087476,
+      "grad_norm": 0.34640170723170177,
+      "learning_rate": 4.660540328415663e-05,
+      "loss": 0.4595,
+      "step": 1264
+    },
+    {
+      "epoch": 2.514910536779324,
+      "grad_norm": 0.3054801483553342,
+      "learning_rate": 4.6550629608411596e-05,
+      "loss": 0.4558,
+      "step": 1265
+    },
+    {
+      "epoch": 2.5168986083499005,
+      "grad_norm": 0.2278535685525667,
+      "learning_rate": 4.6495843308183275e-05,
+      "loss": 0.4648,
+      "step": 1266
+    },
+    {
+      "epoch": 2.518886679920477,
+      "grad_norm": 0.2947939908046167,
+      "learning_rate": 4.644104448905675e-05,
+      "loss": 0.4698,
+      "step": 1267
+    },
+    {
+      "epoch": 2.5208747514910534,
+      "grad_norm": 0.43545344724955837,
+      "learning_rate": 4.638623325664123e-05,
+      "loss": 0.4786,
+      "step": 1268
+    },
+    {
+      "epoch": 2.5228628230616303,
+      "grad_norm": 0.45373043954413755,
+      "learning_rate": 4.633140971656983e-05,
+      "loss": 0.4713,
+      "step": 1269
+    },
+    {
+      "epoch": 2.524850894632207,
+      "grad_norm": 0.33146377812850386,
+      "learning_rate": 4.6276573974499394e-05,
+      "loss": 0.4684,
+      "step": 1270
+    },
+    {
+      "epoch": 2.5268389662027833,
+      "grad_norm": 0.2680817520803437,
+      "learning_rate": 4.622172613611029e-05,
+      "loss": 0.4691,
+      "step": 1271
+    },
+    {
+      "epoch": 2.5288270377733597,
+      "grad_norm": 0.29840968057455064,
+      "learning_rate": 4.6166866307106195e-05,
+      "loss": 0.4681,
+      "step": 1272
+    },
+    {
+      "epoch": 2.530815109343936,
+      "grad_norm": 0.2928890490184993,
+      "learning_rate": 4.611199459321389e-05,
+      "loss": 0.4788,
+      "step": 1273
+    },
+    {
+      "epoch": 2.532803180914513,
+      "grad_norm": 0.24379232071112766,
+      "learning_rate": 4.605711110018307e-05,
+      "loss": 0.4738,
+      "step": 1274
+    },
+    {
+      "epoch": 2.5347912524850895,
+      "grad_norm": 0.22108967349738823,
+      "learning_rate": 4.600221593378611e-05,
+      "loss": 0.4688,
+      "step": 1275
+    },
+    {
+      "epoch": 2.536779324055666,
+      "grad_norm": 0.26959273399137357,
+      "learning_rate": 4.594730919981791e-05,
+      "loss": 0.4573,
+      "step": 1276
+    },
+    {
+      "epoch": 2.5387673956262424,
+      "grad_norm": 0.2989159213403466,
+      "learning_rate": 4.589239100409563e-05,
+      "loss": 0.465,
+      "step": 1277
+    },
+    {
+      "epoch": 2.540755467196819,
+      "grad_norm": 0.21391977504533247,
+      "learning_rate": 4.583746145245854e-05,
+      "loss": 0.467,
+      "step": 1278
+    },
+    {
+      "epoch": 2.542743538767396,
+      "grad_norm": 0.29971289779725213,
+      "learning_rate": 4.578252065076782e-05,
+      "loss": 0.4614,
+      "step": 1279
+    },
+    {
+      "epoch": 2.5447316103379722,
+      "grad_norm": 0.36235889419318323,
+      "learning_rate": 4.5727568704906285e-05,
+      "loss": 0.4709,
+      "step": 1280
+    },
+    {
+      "epoch": 2.5467196819085487,
+      "grad_norm": 0.3184243494913832,
+      "learning_rate": 4.5672605720778235e-05,
+      "loss": 0.4702,
+      "step": 1281
+    },
+    {
+      "epoch": 2.548707753479125,
+      "grad_norm": 0.3367075112146258,
+      "learning_rate": 4.561763180430928e-05,
+      "loss": 0.4681,
+      "step": 1282
+    },
+    {
+      "epoch": 2.5506958250497016,
+      "grad_norm": 0.38423599258480656,
+      "learning_rate": 4.5562647061446065e-05,
+      "loss": 0.4658,
+      "step": 1283
+    },
+    {
+      "epoch": 2.5526838966202785,
+      "grad_norm": 0.30156605639178163,
+      "learning_rate": 4.5507651598156104e-05,
+      "loss": 0.4715,
+      "step": 1284
+    },
+    {
+      "epoch": 2.554671968190855,
+      "grad_norm": 0.2539054825100618,
+      "learning_rate": 4.545264552042758e-05,
+      "loss": 0.4718,
+      "step": 1285
+    },
+    {
+      "epoch": 2.5566600397614314,
+      "grad_norm": 0.2832379512707468,
+      "learning_rate": 4.539762893426913e-05,
+      "loss": 0.4705,
+      "step": 1286
+    },
+    {
+      "epoch": 2.558648111332008,
+      "grad_norm": 0.2729857124864947,
+      "learning_rate": 4.534260194570964e-05,
+      "loss": 0.4629,
+      "step": 1287
+    },
+    {
+      "epoch": 2.5606361829025843,
+      "grad_norm": 0.9159869395713054,
+      "learning_rate": 4.528756466079804e-05,
+      "loss": 0.4611,
+      "step": 1288
+    },
+    {
+      "epoch": 2.5626242544731612,
+      "grad_norm": 0.2947830909553326,
+      "learning_rate": 4.5232517185603124e-05,
+      "loss": 0.4625,
+      "step": 1289
+    },
+    {
+      "epoch": 2.5646123260437377,
+      "grad_norm": 0.30617876929071663,
+      "learning_rate": 4.5177459626213294e-05,
+      "loss": 0.4575,
+      "step": 1290
+    },
+    {
+      "epoch": 2.566600397614314,
+      "grad_norm": 0.3906173771205208,
+      "learning_rate": 4.51223920887364e-05,
+      "loss": 0.4697,
+      "step": 1291
+    },
+    {
+      "epoch": 2.5685884691848906,
+      "grad_norm": 0.5107505465341972,
+      "learning_rate": 4.506731467929955e-05,
+      "loss": 0.4843,
+      "step": 1292
+    },
+    {
+      "epoch": 2.570576540755467,
+      "grad_norm": 0.26067879748684747,
+      "learning_rate": 4.501222750404883e-05,
+      "loss": 0.4687,
+      "step": 1293
+    },
+    {
+      "epoch": 2.572564612326044,
+      "grad_norm": 0.27991711122868407,
+      "learning_rate": 4.495713066914917e-05,
+      "loss": 0.483,
+      "step": 1294
+    },
+    {
+      "epoch": 2.5745526838966204,
+      "grad_norm": 0.3234266564676175,
+      "learning_rate": 4.4902024280784114e-05,
+      "loss": 0.4634,
+      "step": 1295
+    },
+    {
+      "epoch": 2.576540755467197,
+      "grad_norm": 0.29382217694112184,
+      "learning_rate": 4.4846908445155625e-05,
+      "loss": 0.4622,
+      "step": 1296
+    },
+    {
+      "epoch": 2.5785288270377733,
+      "grad_norm": 0.4240599462547126,
+      "learning_rate": 4.4791783268483864e-05,
+      "loss": 0.4707,
+      "step": 1297
+    },
+    {
+      "epoch": 2.5805168986083498,
+      "grad_norm": 0.29831663140195225,
+      "learning_rate": 4.473664885700699e-05,
+      "loss": 0.473,
+      "step": 1298
+    },
+    {
+      "epoch": 2.5825049701789267,
+      "grad_norm": 0.2867259906466539,
+      "learning_rate": 4.468150531698098e-05,
+      "loss": 0.474,
+      "step": 1299
+    },
+    {
+      "epoch": 2.584493041749503,
+      "grad_norm": 0.2380064860884348,
+      "learning_rate": 4.462635275467937e-05,
+      "loss": 0.4718,
+      "step": 1300
+    },
+    {
+      "epoch": 2.5864811133200796,
+      "grad_norm": 0.2817938743929372,
+      "learning_rate": 4.4571191276393125e-05,
+      "loss": 0.4721,
+      "step": 1301
+    },
+    {
+      "epoch": 2.588469184890656,
+      "grad_norm": 0.2792752528880275,
+      "learning_rate": 4.4516020988430346e-05,
+      "loss": 0.4677,
+      "step": 1302
+    },
+    {
+      "epoch": 2.5904572564612325,
+      "grad_norm": 0.27683589472201287,
+      "learning_rate": 4.4460841997116166e-05,
+      "loss": 0.4701,
+      "step": 1303
+    },
+    {
+      "epoch": 2.5924453280318094,
+      "grad_norm": 0.31437221453408365,
+      "learning_rate": 4.440565440879244e-05,
+      "loss": 0.4684,
+      "step": 1304
+    },
+    {
+      "epoch": 2.594433399602386,
+      "grad_norm": 0.30714606345078876,
+      "learning_rate": 4.435045832981763e-05,
+      "loss": 0.4642,
+      "step": 1305
+    },
+    {
+      "epoch": 2.5964214711729623,
+      "grad_norm": 0.2974009434374631,
+      "learning_rate": 4.429525386656653e-05,
+      "loss": 0.4744,
+      "step": 1306
+    },
+    {
+      "epoch": 2.5984095427435387,
+      "grad_norm": 0.3586097986600065,
+      "learning_rate": 4.424004112543012e-05,
+      "loss": 0.468,
+      "step": 1307
+    },
+    {
+      "epoch": 2.600397614314115,
+      "grad_norm": 0.43266237162260085,
+      "learning_rate": 4.4184820212815306e-05,
+      "loss": 0.4655,
+      "step": 1308
+    },
+    {
+      "epoch": 2.602385685884692,
+      "grad_norm": 0.3732894233625949,
+      "learning_rate": 4.4129591235144776e-05,
+      "loss": 0.4682,
+      "step": 1309
+    },
+    {
+      "epoch": 2.604373757455268,
+      "grad_norm": 0.42145804477336907,
+      "learning_rate": 4.407435429885674e-05,
+      "loss": 0.4598,
+      "step": 1310
+    },
+    {
+      "epoch": 2.606361829025845,
+      "grad_norm": 0.3143273970433766,
+      "learning_rate": 4.401910951040473e-05,
+      "loss": 0.4707,
+      "step": 1311
+    },
+    {
+      "epoch": 2.6083499005964215,
+      "grad_norm": 0.2402074218552476,
+      "learning_rate": 4.396385697625745e-05,
+      "loss": 0.4649,
+      "step": 1312
+    },
+    {
+      "epoch": 2.610337972166998,
+      "grad_norm": 0.3650833856208346,
+      "learning_rate": 4.39085968028985e-05,
+      "loss": 0.4705,
+      "step": 1313
+    },
+    {
+      "epoch": 2.612326043737575,
+      "grad_norm": 0.3102695268556827,
+      "learning_rate": 4.385332909682623e-05,
+      "loss": 0.4581,
+      "step": 1314
+    },
+    {
+      "epoch": 2.614314115308151,
+      "grad_norm": 0.2274936765796469,
+      "learning_rate": 4.3798053964553464e-05,
+      "loss": 0.4707,
+      "step": 1315
+    },
+    {
+      "epoch": 2.6163021868787277,
+      "grad_norm": 0.21562809171671227,
+      "learning_rate": 4.374277151260737e-05,
+      "loss": 0.4702,
+      "step": 1316
+    },
+    {
+      "epoch": 2.618290258449304,
+      "grad_norm": 0.2554204822921317,
+      "learning_rate": 4.3687481847529233e-05,
+      "loss": 0.4614,
+      "step": 1317
+    },
+    {
+      "epoch": 2.6202783300198806,
+      "grad_norm": 0.29349056607983115,
+      "learning_rate": 4.3632185075874214e-05,
+      "loss": 0.4704,
+      "step": 1318
+    },
+    {
+      "epoch": 2.622266401590457,
+      "grad_norm": 0.20690342951276153,
+      "learning_rate": 4.357688130421117e-05,
+      "loss": 0.477,
+      "step": 1319
+    },
+    {
+      "epoch": 2.6242544731610336,
+      "grad_norm": 0.24912650600700706,
+      "learning_rate": 4.3521570639122474e-05,
+      "loss": 0.4676,
+      "step": 1320
+    },
+    {
+      "epoch": 2.6262425447316105,
+      "grad_norm": 0.31060871647011534,
+      "learning_rate": 4.346625318720376e-05,
+      "loss": 0.4731,
+      "step": 1321
+    },
+    {
+      "epoch": 2.628230616302187,
+      "grad_norm": 0.3175262763718928,
+      "learning_rate": 4.341092905506375e-05,
+      "loss": 0.4647,
+      "step": 1322
+    },
+    {
+      "epoch": 2.6302186878727634,
+      "grad_norm": 0.22988315191632902,
+      "learning_rate": 4.3355598349324034e-05,
+      "loss": 0.4652,
+      "step": 1323
+    },
+    {
+      "epoch": 2.63220675944334,
+      "grad_norm": 0.29964998822607253,
+      "learning_rate": 4.330026117661888e-05,
+      "loss": 0.4597,
+      "step": 1324
+    },
+    {
+      "epoch": 2.6341948310139163,
+      "grad_norm": 0.2862821590914649,
+      "learning_rate": 4.324491764359502e-05,
+      "loss": 0.4705,
+      "step": 1325
+    },
+    {
+      "epoch": 2.636182902584493,
+      "grad_norm": 0.22238282029132211,
+      "learning_rate": 4.318956785691144e-05,
+      "loss": 0.4743,
+      "step": 1326
+    },
+    {
+      "epoch": 2.6381709741550696,
+      "grad_norm": 0.20835203612719516,
+      "learning_rate": 4.313421192323917e-05,
+      "loss": 0.4674,
+      "step": 1327
+    },
+    {
+      "epoch": 2.640159045725646,
+      "grad_norm": 0.21296138418439783,
+      "learning_rate": 4.30788499492611e-05,
+      "loss": 0.4669,
+      "step": 1328
+    },
+    {
+      "epoch": 2.6421471172962225,
+      "grad_norm": 0.6063626470982033,
+      "learning_rate": 4.3023482041671754e-05,
+      "loss": 0.4646,
+      "step": 1329
+    },
+    {
+      "epoch": 2.644135188866799,
+      "grad_norm": 0.22456488467008662,
+      "learning_rate": 4.296810830717709e-05,
+      "loss": 0.4576,
+      "step": 1330
+    },
+    {
+      "epoch": 2.646123260437376,
+      "grad_norm": 0.21173368053034997,
+      "learning_rate": 4.29127288524943e-05,
+      "loss": 0.4627,
+      "step": 1331
+    },
+    {
+      "epoch": 2.6481113320079523,
+      "grad_norm": 0.20369594046938577,
+      "learning_rate": 4.28573437843516e-05,
+      "loss": 0.4623,
+      "step": 1332
+    },
+    {
+      "epoch": 2.650099403578529,
+      "grad_norm": 0.23192364825311867,
+      "learning_rate": 4.280195320948801e-05,
+      "loss": 0.4562,
+      "step": 1333
+    },
+    {
+      "epoch": 2.6520874751491053,
+      "grad_norm": 0.20702404270208882,
+      "learning_rate": 4.274655723465319e-05,
+      "loss": 0.4587,
+      "step": 1334
+    },
+    {
+      "epoch": 2.6540755467196817,
+      "grad_norm": 0.2043349186973809,
+      "learning_rate": 4.26911559666072e-05,
+      "loss": 0.4686,
+      "step": 1335
+    },
+    {
+      "epoch": 2.6560636182902586,
+      "grad_norm": 0.2140823713672654,
+      "learning_rate": 4.263574951212027e-05,
+      "loss": 0.467,
+      "step": 1336
+    },
+    {
+      "epoch": 2.658051689860835,
+      "grad_norm": 0.21150791516481285,
+      "learning_rate": 4.2580337977972664e-05,
+      "loss": 0.4664,
+      "step": 1337
+    },
+    {
+      "epoch": 2.6600397614314115,
+      "grad_norm": 0.22487112272684076,
+      "learning_rate": 4.2524921470954406e-05,
+      "loss": 0.4693,
+      "step": 1338
+    },
+    {
+      "epoch": 2.662027833001988,
+      "grad_norm": 0.21793132020525952,
+      "learning_rate": 4.2469500097865134e-05,
+      "loss": 0.4729,
+      "step": 1339
+    },
+    {
+      "epoch": 2.6640159045725644,
+      "grad_norm": 0.2133804189419038,
+      "learning_rate": 4.2414073965513846e-05,
+      "loss": 0.4667,
+      "step": 1340
+    },
+    {
+      "epoch": 2.6660039761431413,
+      "grad_norm": 0.18948465569873307,
+      "learning_rate": 4.235864318071871e-05,
+      "loss": 0.4583,
+      "step": 1341
+    },
+    {
+      "epoch": 2.667992047713718,
+      "grad_norm": 0.19523854934494658,
+      "learning_rate": 4.230320785030686e-05,
+      "loss": 0.4624,
+      "step": 1342
+    },
+    {
+      "epoch": 2.6699801192842942,
+      "grad_norm": 1.6838215088450827,
+      "learning_rate": 4.2247768081114204e-05,
+      "loss": 0.486,
+      "step": 1343
+    },
+    {
+      "epoch": 2.6719681908548707,
+      "grad_norm": 0.22540813024187262,
+      "learning_rate": 4.219232397998518e-05,
+      "loss": 0.4574,
+      "step": 1344
+    },
+    {
+      "epoch": 2.673956262425447,
+      "grad_norm": 0.28028646989057787,
+      "learning_rate": 4.2136875653772614e-05,
+      "loss": 0.4695,
+      "step": 1345
+    },
+    {
+      "epoch": 2.675944333996024,
+      "grad_norm": 0.23378583807339337,
+      "learning_rate": 4.2081423209337426e-05,
+      "loss": 0.4677,
+      "step": 1346
+    },
+    {
+      "epoch": 2.6779324055666005,
+      "grad_norm": 0.23288832160542997,
+      "learning_rate": 4.202596675354851e-05,
+      "loss": 0.4641,
+      "step": 1347
+    },
+    {
+      "epoch": 2.679920477137177,
+      "grad_norm": 0.34485144522745426,
+      "learning_rate": 4.1970506393282483e-05,
+      "loss": 0.465,
+      "step": 1348
+    },
+    {
+      "epoch": 2.6819085487077534,
+      "grad_norm": 0.22517376125305072,
+      "learning_rate": 4.191504223542348e-05,
+      "loss": 0.4698,
+      "step": 1349
+    },
+    {
+      "epoch": 2.68389662027833,
+      "grad_norm": 0.2601189519472589,
+      "learning_rate": 4.185957438686294e-05,
+      "loss": 0.4748,
+      "step": 1350
+    },
+    {
+      "epoch": 2.6858846918489068,
+      "grad_norm": 0.3474035598879338,
+      "learning_rate": 4.180410295449944e-05,
+      "loss": 0.4668,
+      "step": 1351
+    },
+    {
+      "epoch": 2.6878727634194832,
+      "grad_norm": 0.24685855580960245,
+      "learning_rate": 4.174862804523847e-05,
+      "loss": 0.4577,
+      "step": 1352
+    },
+    {
+      "epoch": 2.6898608349900597,
+      "grad_norm": 0.21318028985851886,
+      "learning_rate": 4.1693149765992185e-05,
+      "loss": 0.4677,
+      "step": 1353
+    },
+    {
+      "epoch": 2.691848906560636,
+      "grad_norm": 0.25285941693159386,
+      "learning_rate": 4.1637668223679265e-05,
+      "loss": 0.4649,
+      "step": 1354
+    },
+    {
+      "epoch": 2.6938369781312126,
+      "grad_norm": 0.25170908983832746,
+      "learning_rate": 4.158218352522468e-05,
+      "loss": 0.4576,
+      "step": 1355
+    },
+    {
+      "epoch": 2.6958250497017895,
+      "grad_norm": 0.24902625836971093,
+      "learning_rate": 4.152669577755945e-05,
+      "loss": 0.4694,
+      "step": 1356
+    },
+    {
+      "epoch": 2.697813121272366,
+      "grad_norm": 0.26522375080926597,
+      "learning_rate": 4.147120508762051e-05,
+      "loss": 0.4685,
+      "step": 1357
+    },
+    {
+      "epoch": 2.6998011928429424,
+      "grad_norm": 0.21543694379170508,
+      "learning_rate": 4.141571156235044e-05,
+      "loss": 0.4684,
+      "step": 1358
+    },
+    {
+      "epoch": 2.701789264413519,
+      "grad_norm": 0.19796151498045514,
+      "learning_rate": 4.1360215308697294e-05,
+      "loss": 0.4586,
+      "step": 1359
+    },
+    {
+      "epoch": 2.7037773359840953,
+      "grad_norm": 0.2360071821488426,
+      "learning_rate": 4.130471643361438e-05,
+      "loss": 0.4622,
+      "step": 1360
+    },
+    {
+      "epoch": 2.705765407554672,
+      "grad_norm": 0.2401591059548515,
+      "learning_rate": 4.124921504406008e-05,
+      "loss": 0.4697,
+      "step": 1361
+    },
+    {
+      "epoch": 2.7077534791252487,
+      "grad_norm": 0.19632371914009855,
+      "learning_rate": 4.119371124699757e-05,
+      "loss": 0.4637,
+      "step": 1362
+    },
+    {
+      "epoch": 2.709741550695825,
+      "grad_norm": 0.24309468602335702,
+      "learning_rate": 4.1138205149394723e-05,
+      "loss": 0.4702,
+      "step": 1363
+    },
+    {
+      "epoch": 2.7117296222664016,
+      "grad_norm": 0.22446170452361522,
+      "learning_rate": 4.1082696858223826e-05,
+      "loss": 0.4625,
+      "step": 1364
+    },
+    {
+      "epoch": 2.713717693836978,
+      "grad_norm": 0.2258465756554639,
+      "learning_rate": 4.1027186480461375e-05,
+      "loss": 0.4579,
+      "step": 1365
+    },
+    {
+      "epoch": 2.715705765407555,
+      "grad_norm": 0.2644422390314895,
+      "learning_rate": 4.09716741230879e-05,
+      "loss": 0.4709,
+      "step": 1366
+    },
+    {
+      "epoch": 2.717693836978131,
+      "grad_norm": 0.2103180179626641,
+      "learning_rate": 4.0916159893087756e-05,
+      "loss": 0.4731,
+      "step": 1367
+    },
+    {
+      "epoch": 2.719681908548708,
+      "grad_norm": 0.27361880683699913,
+      "learning_rate": 4.0860643897448894e-05,
+      "loss": 0.4678,
+      "step": 1368
+    },
+    {
+      "epoch": 2.7216699801192843,
+      "grad_norm": 0.28798144219878985,
+      "learning_rate": 4.080512624316267e-05,
+      "loss": 0.4726,
+      "step": 1369
+    },
+    {
+      "epoch": 2.7236580516898607,
+      "grad_norm": 0.2264593548600379,
+      "learning_rate": 4.074960703722365e-05,
+      "loss": 0.4662,
+      "step": 1370
+    },
+    {
+      "epoch": 2.7256461232604376,
+      "grad_norm": 0.2242660243057448,
+      "learning_rate": 4.069408638662937e-05,
+      "loss": 0.4644,
+      "step": 1371
+    },
+    {
+      "epoch": 2.7276341948310137,
+      "grad_norm": 0.2621409442094008,
+      "learning_rate": 4.0638564398380163e-05,
+      "loss": 0.4655,
+      "step": 1372
+    },
+    {
+      "epoch": 2.7296222664015906,
+      "grad_norm": 0.2595547616945105,
+      "learning_rate": 4.0583041179478936e-05,
+      "loss": 0.4638,
+      "step": 1373
+    },
+    {
+      "epoch": 2.731610337972167,
+      "grad_norm": 0.2274598504625181,
+      "learning_rate": 4.052751683693095e-05,
+      "loss": 0.4633,
+      "step": 1374
+    },
+    {
+      "epoch": 2.7335984095427435,
+      "grad_norm": 0.23999707275096588,
+      "learning_rate": 4.047199147774368e-05,
+      "loss": 0.4631,
+      "step": 1375
+    },
+    {
+      "epoch": 2.73558648111332,
+      "grad_norm": 0.22767172991322615,
+      "learning_rate": 4.041646520892651e-05,
+      "loss": 0.4665,
+      "step": 1376
+    },
+    {
+      "epoch": 2.7375745526838964,
+      "grad_norm": 0.27587767902251026,
+      "learning_rate": 4.0360938137490604e-05,
+      "loss": 0.473,
+      "step": 1377
+    },
+    {
+      "epoch": 2.7395626242544733,
+      "grad_norm": 0.21764221482701487,
+      "learning_rate": 4.030541037044866e-05,
+      "loss": 0.4566,
+      "step": 1378
+    },
+    {
+      "epoch": 2.7415506958250497,
+      "grad_norm": 0.2666834855042023,
+      "learning_rate": 4.0249882014814735e-05,
+      "loss": 0.4624,
+      "step": 1379
+    },
+    {
+      "epoch": 2.743538767395626,
+      "grad_norm": 0.2364553717028977,
+      "learning_rate": 4.019435317760399e-05,
+      "loss": 0.4616,
+      "step": 1380
+    },
+    {
+      "epoch": 2.7455268389662026,
+      "grad_norm": 0.19496138813490002,
+      "learning_rate": 4.013882396583255e-05,
+      "loss": 0.4587,
+      "step": 1381
+    },
+    {
+      "epoch": 2.747514910536779,
+      "grad_norm": 0.20073634106913804,
+      "learning_rate": 4.008329448651722e-05,
+      "loss": 0.46,
+      "step": 1382
+    },
+    {
+      "epoch": 2.749502982107356,
+      "grad_norm": 0.2095676619154563,
+      "learning_rate": 4.0027764846675375e-05,
+      "loss": 0.4615,
+      "step": 1383
+    },
+    {
+      "epoch": 2.7514910536779325,
+      "grad_norm": 0.20024726668039622,
+      "learning_rate": 3.997223515332464e-05,
+      "loss": 0.4627,
+      "step": 1384
+    },
+    {
+      "epoch": 2.753479125248509,
+      "grad_norm": 0.18961125621675926,
+      "learning_rate": 3.991670551348279e-05,
+      "loss": 0.4655,
+      "step": 1385
+    },
+    {
+      "epoch": 2.7554671968190854,
+      "grad_norm": 0.1941167507355979,
+      "learning_rate": 3.9861176034167464e-05,
+      "loss": 0.4662,
+      "step": 1386
+    },
+    {
+      "epoch": 2.757455268389662,
+      "grad_norm": 0.23045374898538842,
+      "learning_rate": 3.980564682239603e-05,
+      "loss": 0.4686,
+      "step": 1387
+    },
+    {
+      "epoch": 2.7594433399602387,
+      "grad_norm": 0.3605198895596153,
+      "learning_rate": 3.9750117985185265e-05,
+      "loss": 0.4621,
+      "step": 1388
+    },
+    {
+      "epoch": 2.761431411530815,
+      "grad_norm": 0.29058441579436517,
+      "learning_rate": 3.969458962955134e-05,
+      "loss": 0.4699,
+      "step": 1389
+    },
+    {
+      "epoch": 2.7634194831013916,
+      "grad_norm": 0.2299369126959643,
+      "learning_rate": 3.96390618625094e-05,
+      "loss": 0.4644,
+      "step": 1390
+    },
+    {
+      "epoch": 2.765407554671968,
+      "grad_norm": 0.19284704992279508,
+      "learning_rate": 3.9583534791073496e-05,
+      "loss": 0.474,
+      "step": 1391
+    },
+    {
+      "epoch": 2.7673956262425445,
+      "grad_norm": 0.2294396001810075,
+      "learning_rate": 3.952800852225633e-05,
+      "loss": 0.4692,
+      "step": 1392
+    },
+    {
+      "epoch": 2.7693836978131214,
+      "grad_norm": 0.21459748499096268,
+      "learning_rate": 3.947248316306906e-05,
+      "loss": 0.4549,
+      "step": 1393
+    },
+    {
+      "epoch": 2.771371769383698,
+      "grad_norm": 0.21514167913930246,
+      "learning_rate": 3.9416958820521084e-05,
+      "loss": 0.4676,
+      "step": 1394
+    },
+    {
+      "epoch": 2.7733598409542743,
+      "grad_norm": 0.21325784114659638,
+      "learning_rate": 3.936143560161985e-05,
+      "loss": 0.4642,
+      "step": 1395
+    },
+    {
+      "epoch": 2.775347912524851,
+      "grad_norm": 0.24281523036646144,
+      "learning_rate": 3.930591361337065e-05,
+      "loss": 0.4628,
+      "step": 1396
+    },
+    {
+      "epoch": 2.7773359840954273,
+      "grad_norm": 0.265175164446194,
+      "learning_rate": 3.925039296277635e-05,
+      "loss": 0.4579,
+      "step": 1397
+    },
+    {
+      "epoch": 2.779324055666004,
+      "grad_norm": 0.25122618281563747,
+      "learning_rate": 3.919487375683733e-05,
+      "loss": 0.4571,
+      "step": 1398
+    },
+    {
+      "epoch": 2.7813121272365806,
+      "grad_norm": 0.21129482504489955,
+      "learning_rate": 3.913935610255112e-05,
+      "loss": 0.4486,
+      "step": 1399
+    },
+    {
+      "epoch": 2.783300198807157,
+      "grad_norm": 0.3147948552145016,
+      "learning_rate": 3.908384010691226e-05,
+      "loss": 0.4579,
+      "step": 1400
+    },
+    {
+      "epoch": 2.7852882703777335,
+      "grad_norm": 0.2455473863371928,
+      "learning_rate": 3.902832587691211e-05,
+      "loss": 0.4512,
+      "step": 1401
+    },
+    {
+      "epoch": 2.78727634194831,
+      "grad_norm": 0.2735323746731553,
+      "learning_rate": 3.897281351953864e-05,
+      "loss": 0.4549,
+      "step": 1402
+    },
+    {
+      "epoch": 2.789264413518887,
+      "grad_norm": 0.28524076892759626,
+      "learning_rate": 3.891730314177619e-05,
+      "loss": 0.4635,
+      "step": 1403
+    },
+    {
+      "epoch": 2.7912524850894633,
+      "grad_norm": 0.2566773378554369,
+      "learning_rate": 3.886179485060528e-05,
+      "loss": 0.4602,
+      "step": 1404
+    },
+    {
+      "epoch": 2.79324055666004,
+      "grad_norm": 0.2357393748079123,
+      "learning_rate": 3.8806288753002446e-05,
+      "loss": 0.4738,
+      "step": 1405
+    },
+    {
+      "epoch": 2.7952286282306162,
+      "grad_norm": 0.19262960449941974,
+      "learning_rate": 3.875078495593993e-05,
+      "loss": 0.4552,
+      "step": 1406
+    },
+    {
+      "epoch": 2.7972166998011927,
+      "grad_norm": 0.2201663667450906,
+      "learning_rate": 3.869528356638562e-05,
+      "loss": 0.4611,
+      "step": 1407
+    },
+    {
+      "epoch": 2.7992047713717696,
+      "grad_norm": 0.27795764096966974,
+      "learning_rate": 3.863978469130272e-05,
+      "loss": 0.4572,
+      "step": 1408
+    },
+    {
+      "epoch": 2.801192842942346,
+      "grad_norm": 0.2335116245887707,
+      "learning_rate": 3.8584288437649574e-05,
+      "loss": 0.4694,
+      "step": 1409
+    },
+    {
+      "epoch": 2.8031809145129225,
+      "grad_norm": 0.21856549040110454,
+      "learning_rate": 3.8528794912379506e-05,
+      "loss": 0.4577,
+      "step": 1410
+    },
+    {
+      "epoch": 2.805168986083499,
+      "grad_norm": 0.25356968098822286,
+      "learning_rate": 3.8473304222440556e-05,
+      "loss": 0.4636,
+      "step": 1411
+    },
+    {
+      "epoch": 2.8071570576540754,
+      "grad_norm": 0.21605953966990488,
+      "learning_rate": 3.841781647477533e-05,
+      "loss": 0.4671,
+      "step": 1412
+    },
+    {
+      "epoch": 2.8091451292246523,
+      "grad_norm": 0.25322156738563867,
+      "learning_rate": 3.836233177632075e-05,
+      "loss": 0.4584,
+      "step": 1413
+    },
+    {
+      "epoch": 2.8111332007952288,
+      "grad_norm": 0.2229043276554981,
+      "learning_rate": 3.8306850234007835e-05,
+      "loss": 0.46,
+      "step": 1414
+    },
+    {
+      "epoch": 2.8131212723658052,
+      "grad_norm": 0.22038124684629573,
+      "learning_rate": 3.825137195476154e-05,
+      "loss": 0.4614,
+      "step": 1415
+    },
+    {
+      "epoch": 2.8151093439363817,
+      "grad_norm": 0.25286192519399997,
+      "learning_rate": 3.819589704550056e-05,
+      "loss": 0.4535,
+      "step": 1416
+    },
+    {
+      "epoch": 2.817097415506958,
+      "grad_norm": 0.25163637021592183,
+      "learning_rate": 3.814042561313708e-05,
+      "loss": 0.4622,
+      "step": 1417
+    },
+    {
+      "epoch": 2.819085487077535,
+      "grad_norm": 0.21455983513210608,
+      "learning_rate": 3.808495776457654e-05,
+      "loss": 0.456,
+      "step": 1418
+    },
+    {
+      "epoch": 2.8210735586481115,
+      "grad_norm": 0.2410203088285688,
+      "learning_rate": 3.802949360671753e-05,
+      "loss": 0.4593,
+      "step": 1419
+    },
+    {
+      "epoch": 2.823061630218688,
+      "grad_norm": 0.22666498127120266,
+      "learning_rate": 3.7974033246451496e-05,
+      "loss": 0.4674,
+      "step": 1420
+    },
+    {
+      "epoch": 2.8250497017892644,
+      "grad_norm": 0.2678467891624614,
+      "learning_rate": 3.791857679066259e-05,
+      "loss": 0.4648,
+      "step": 1421
+    },
+    {
+      "epoch": 2.827037773359841,
+      "grad_norm": 0.24714758794770142,
+      "learning_rate": 3.786312434622741e-05,
+      "loss": 0.4563,
+      "step": 1422
+    },
+    {
+      "epoch": 2.8290258449304178,
+      "grad_norm": 0.3159535557855933,
+      "learning_rate": 3.780767602001484e-05,
+      "loss": 0.4691,
+      "step": 1423
+    },
+    {
+      "epoch": 2.8310139165009938,
+      "grad_norm": 0.24286637586695814,
+      "learning_rate": 3.77522319188858e-05,
+      "loss": 0.4564,
+      "step": 1424
+    },
+    {
+      "epoch": 2.8330019880715707,
+      "grad_norm": 0.2686674277466147,
+      "learning_rate": 3.7696792149693144e-05,
+      "loss": 0.46,
+      "step": 1425
+    },
+    {
+      "epoch": 2.834990059642147,
+      "grad_norm": 0.2813965137999818,
+      "learning_rate": 3.764135681928131e-05,
+      "loss": 0.4676,
+      "step": 1426
+    },
+    {
+      "epoch": 2.8369781312127236,
+      "grad_norm": 0.31227772048480595,
+      "learning_rate": 3.758592603448617e-05,
+      "loss": 0.4614,
+      "step": 1427
+    },
+    {
+      "epoch": 2.8389662027833005,
+      "grad_norm": 0.27050892530225235,
+      "learning_rate": 3.753049990213487e-05,
+      "loss": 0.4595,
+      "step": 1428
+    },
+    {
+      "epoch": 2.8409542743538765,
+      "grad_norm": 0.3253972996223399,
+      "learning_rate": 3.74750785290456e-05,
+      "loss": 0.4541,
+      "step": 1429
+    },
+    {
+      "epoch": 2.8429423459244534,
+      "grad_norm": 0.8290885706508856,
+      "learning_rate": 3.7419662022027356e-05,
+      "loss": 0.4626,
+      "step": 1430
+    },
+    {
+      "epoch": 2.84493041749503,
+      "grad_norm": 0.3278453956679973,
+      "learning_rate": 3.736425048787976e-05,
+      "loss": 0.4595,
+      "step": 1431
+    },
+    {
+      "epoch": 2.8469184890656063,
+      "grad_norm": 0.3380893038877377,
+      "learning_rate": 3.730884403339282e-05,
+      "loss": 0.458,
+      "step": 1432
+    },
+    {
+      "epoch": 2.8489065606361827,
+      "grad_norm": 0.2208597079738313,
+      "learning_rate": 3.725344276534681e-05,
+      "loss": 0.4588,
+      "step": 1433
+    },
+    {
+      "epoch": 2.850894632206759,
+      "grad_norm": 0.2229228221920045,
+      "learning_rate": 3.7198046790512e-05,
+      "loss": 0.472,
+      "step": 1434
+    },
+    {
+      "epoch": 2.852882703777336,
+      "grad_norm": 0.34642900940428056,
+      "learning_rate": 3.714265621564842e-05,
+      "loss": 0.4594,
+      "step": 1435
+    },
+    {
+      "epoch": 2.8548707753479126,
+      "grad_norm": 0.26258153715517213,
+      "learning_rate": 3.708727114750572e-05,
+      "loss": 0.4698,
+      "step": 1436
+    },
+    {
+      "epoch": 2.856858846918489,
+      "grad_norm": 0.21133696660065732,
+      "learning_rate": 3.7031891692822924e-05,
+      "loss": 0.47,
+      "step": 1437
+    },
+    {
+      "epoch": 2.8588469184890655,
+      "grad_norm": 0.2805733965805607,
+      "learning_rate": 3.697651795832826e-05,
+      "loss": 0.4618,
+      "step": 1438
+    },
+    {
+      "epoch": 2.860834990059642,
+      "grad_norm": 0.28771730407413304,
+      "learning_rate": 3.692115005073891e-05,
+      "loss": 0.4622,
+      "step": 1439
+    },
+    {
+      "epoch": 2.862823061630219,
+      "grad_norm": 0.257745593969288,
+      "learning_rate": 3.686578807676085e-05,
+      "loss": 0.4601,
+      "step": 1440
+    },
+    {
+      "epoch": 2.8648111332007953,
+      "grad_norm": 0.29359923657472803,
+      "learning_rate": 3.681043214308856e-05,
+      "loss": 0.4672,
+      "step": 1441
+    },
+    {
+      "epoch": 2.8667992047713717,
+      "grad_norm": 0.193949825815246,
+      "learning_rate": 3.6755082356404976e-05,
+      "loss": 0.4547,
+      "step": 1442
+    },
+    {
+      "epoch": 2.868787276341948,
+      "grad_norm": 0.26037388451793225,
+      "learning_rate": 3.6699738823381125e-05,
+      "loss": 0.4653,
+      "step": 1443
+    },
+    {
+      "epoch": 2.8707753479125246,
+      "grad_norm": 0.337949377251768,
+      "learning_rate": 3.664440165067598e-05,
+      "loss": 0.4588,
+      "step": 1444
+    },
+    {
+      "epoch": 2.8727634194831015,
+      "grad_norm": 0.9950831510061162,
+      "learning_rate": 3.6589070944936264e-05,
+      "loss": 0.4635,
+      "step": 1445
+    },
+    {
+      "epoch": 2.874751491053678,
+      "grad_norm": 0.2688681333914108,
+      "learning_rate": 3.653374681279625e-05,
+      "loss": 0.4609,
+      "step": 1446
+    },
+    {
+      "epoch": 2.8767395626242545,
+      "grad_norm": 0.26931338222240986,
+      "learning_rate": 3.647842936087754e-05,
+      "loss": 0.4537,
+      "step": 1447
+    },
+    {
+      "epoch": 2.878727634194831,
+      "grad_norm": 0.2657472291582106,
+      "learning_rate": 3.6423118695788834e-05,
+      "loss": 0.4639,
+      "step": 1448
+    },
+    {
+      "epoch": 2.8807157057654074,
+      "grad_norm": 0.30395367100041154,
+      "learning_rate": 3.6367814924125806e-05,
+      "loss": 0.4579,
+      "step": 1449
+    },
+    {
+      "epoch": 2.8827037773359843,
+      "grad_norm": 0.26228486695021913,
+      "learning_rate": 3.631251815247077e-05,
+      "loss": 0.4622,
+      "step": 1450
+    },
+    {
+      "epoch": 2.8846918489065607,
+      "grad_norm": 0.25042552063623835,
+      "learning_rate": 3.625722848739263e-05,
+      "loss": 0.4578,
+      "step": 1451
+    },
+    {
+      "epoch": 2.886679920477137,
+      "grad_norm": 0.37353549732274993,
+      "learning_rate": 3.620194603544655e-05,
+      "loss": 0.4583,
+      "step": 1452
+    },
+    {
+      "epoch": 2.8886679920477136,
+      "grad_norm": 0.2702383408957765,
+      "learning_rate": 3.6146670903173785e-05,
+      "loss": 0.4605,
+      "step": 1453
+    },
+    {
+      "epoch": 2.89065606361829,
+      "grad_norm": 0.20949177632500207,
+      "learning_rate": 3.6091403197101504e-05,
+      "loss": 0.4594,
+      "step": 1454
+    },
+    {
+      "epoch": 2.892644135188867,
+      "grad_norm": 0.20249165415901857,
+      "learning_rate": 3.6036143023742553e-05,
+      "loss": 0.4586,
+      "step": 1455
+    },
+    {
+      "epoch": 2.8946322067594434,
+      "grad_norm": 0.2013107734031283,
+      "learning_rate": 3.5980890489595275e-05,
+      "loss": 0.4582,
+      "step": 1456
+    },
+    {
+      "epoch": 2.89662027833002,
+      "grad_norm": 0.17825803453477496,
+      "learning_rate": 3.592564570114327e-05,
+      "loss": 0.4667,
+      "step": 1457
+    },
+    {
+      "epoch": 2.8986083499005963,
+      "grad_norm": 0.2043117194799363,
+      "learning_rate": 3.587040876485524e-05,
+      "loss": 0.4523,
+      "step": 1458
+    },
+    {
+      "epoch": 2.900596421471173,
+      "grad_norm": 0.1889988045208227,
+      "learning_rate": 3.5815179787184694e-05,
+      "loss": 0.4471,
+      "step": 1459
+    },
+    {
+      "epoch": 2.9025844930417497,
+      "grad_norm": 0.17301877144085692,
+      "learning_rate": 3.575995887456989e-05,
+      "loss": 0.4616,
+      "step": 1460
+    },
+    {
+      "epoch": 2.904572564612326,
+      "grad_norm": 0.1858563184396588,
+      "learning_rate": 3.570474613343348e-05,
+      "loss": 0.4538,
+      "step": 1461
+    },
+    {
+      "epoch": 2.9065606361829026,
+      "grad_norm": 0.17475933284775572,
+      "learning_rate": 3.5649541670182385e-05,
+      "loss": 0.4644,
+      "step": 1462
+    },
+    {
+      "epoch": 2.908548707753479,
+      "grad_norm": 0.19020230936864846,
+      "learning_rate": 3.559434559120757e-05,
+      "loss": 0.4662,
+      "step": 1463
+    },
+    {
+      "epoch": 2.9105367793240555,
+      "grad_norm": 0.19660105128297872,
+      "learning_rate": 3.553915800288385e-05,
+      "loss": 0.4679,
+      "step": 1464
+    },
+    {
+      "epoch": 2.9125248508946324,
+      "grad_norm": 0.2046735778964775,
+      "learning_rate": 3.548397901156966e-05,
+      "loss": 0.4664,
+      "step": 1465
+    },
+    {
+      "epoch": 2.914512922465209,
+      "grad_norm": 0.21619552346200047,
+      "learning_rate": 3.542880872360689e-05,
+      "loss": 0.4566,
+      "step": 1466
+    },
+    {
+      "epoch": 2.9165009940357853,
+      "grad_norm": 0.19360108054912628,
+      "learning_rate": 3.537364724532065e-05,
+      "loss": 0.4655,
+      "step": 1467
+    },
+    {
+      "epoch": 2.918489065606362,
+      "grad_norm": 0.18241404196740055,
+      "learning_rate": 3.5318494683019025e-05,
+      "loss": 0.4662,
+      "step": 1468
+    },
+    {
+      "epoch": 2.9204771371769382,
+      "grad_norm": 0.21577796418109538,
+      "learning_rate": 3.5263351142993e-05,
+      "loss": 0.4599,
+      "step": 1469
+    },
+    {
+      "epoch": 2.922465208747515,
+      "grad_norm": 0.2087684977429425,
+      "learning_rate": 3.520821673151614e-05,
+      "loss": 0.465,
+      "step": 1470
+    },
+    {
+      "epoch": 2.9244532803180916,
+      "grad_norm": 0.6884222775945866,
+      "learning_rate": 3.515309155484438e-05,
+      "loss": 0.4607,
+      "step": 1471
+    },
+    {
+      "epoch": 2.926441351888668,
+      "grad_norm": 0.224933874693449,
+      "learning_rate": 3.509797571921589e-05,
+      "loss": 0.4556,
+      "step": 1472
+    },
+    {
+      "epoch": 2.9284294234592445,
+      "grad_norm": 0.23471362139521423,
+      "learning_rate": 3.5042869330850844e-05,
+      "loss": 0.4643,
+      "step": 1473
+    },
+    {
+      "epoch": 2.930417495029821,
+      "grad_norm": 0.233852872228427,
+      "learning_rate": 3.498777249595118e-05,
+      "loss": 0.452,
+      "step": 1474
+    },
+    {
+      "epoch": 2.932405566600398,
+      "grad_norm": 0.2210891714850947,
+      "learning_rate": 3.493268532070046e-05,
+      "loss": 0.4678,
+      "step": 1475
+    },
+    {
+      "epoch": 2.9343936381709743,
+      "grad_norm": 0.3535398087630285,
+      "learning_rate": 3.487760791126361e-05,
+      "loss": 0.469,
+      "step": 1476
+    },
+    {
+      "epoch": 2.9363817097415508,
+      "grad_norm": 0.25585773718055604,
+      "learning_rate": 3.482254037378671e-05,
+      "loss": 0.4532,
+      "step": 1477
+    },
+    {
+      "epoch": 2.9383697813121272,
+      "grad_norm": 0.23694109296102414,
+      "learning_rate": 3.476748281439688e-05,
+      "loss": 0.4564,
+      "step": 1478
+    },
+    {
+      "epoch": 2.9403578528827037,
+      "grad_norm": 0.19910372458518003,
+      "learning_rate": 3.471243533920197e-05,
+      "loss": 0.4621,
+      "step": 1479
+    },
+    {
+      "epoch": 2.9423459244532806,
+      "grad_norm": 0.2237982620944968,
+      "learning_rate": 3.465739805429038e-05,
+      "loss": 0.4668,
+      "step": 1480
+    },
+    {
+      "epoch": 2.9443339960238566,
+      "grad_norm": 0.21397024941277665,
+      "learning_rate": 3.460237106573089e-05,
+      "loss": 0.4585,
+      "step": 1481
+    },
+    {
+      "epoch": 2.9463220675944335,
+      "grad_norm": 0.21613520324167362,
+      "learning_rate": 3.454735447957244e-05,
+      "loss": 0.4643,
+      "step": 1482
+    },
+    {
+      "epoch": 2.94831013916501,
+      "grad_norm": 0.20222193441002365,
+      "learning_rate": 3.44923484018439e-05,
+      "loss": 0.4553,
+      "step": 1483
+    },
+    {
+      "epoch": 2.9502982107355864,
+      "grad_norm": 0.20180090085759034,
+      "learning_rate": 3.443735293855395e-05,
+      "loss": 0.4608,
+      "step": 1484
+    },
+    {
+      "epoch": 2.952286282306163,
+      "grad_norm": 0.1888317542842441,
+      "learning_rate": 3.438236819569074e-05,
+      "loss": 0.4578,
+      "step": 1485
+    },
+    {
+      "epoch": 2.9542743538767393,
+      "grad_norm": 0.1921413976782572,
+      "learning_rate": 3.432739427922176e-05,
+      "loss": 0.4525,
+      "step": 1486
+    },
+    {
+      "epoch": 2.956262425447316,
+      "grad_norm": 0.18167928678331619,
+      "learning_rate": 3.427243129509372e-05,
+      "loss": 0.4519,
+      "step": 1487
+    },
+    {
+      "epoch": 2.9582504970178927,
+      "grad_norm": 0.18458189230588334,
+      "learning_rate": 3.421747934923219e-05,
+      "loss": 0.4683,
+      "step": 1488
+    },
+    {
+      "epoch": 2.960238568588469,
+      "grad_norm": 0.2082868127458548,
+      "learning_rate": 3.4162538547541465e-05,
+      "loss": 0.4552,
+      "step": 1489
+    },
+    {
+      "epoch": 2.9622266401590456,
+      "grad_norm": 0.18584151995338427,
+      "learning_rate": 3.410760899590439e-05,
+      "loss": 0.4631,
+      "step": 1490
+    },
+    {
+      "epoch": 2.964214711729622,
+      "grad_norm": 0.2383682645766151,
+      "learning_rate": 3.4052690800182104e-05,
+      "loss": 0.4689,
+      "step": 1491
+    },
+    {
+      "epoch": 2.966202783300199,
+      "grad_norm": 0.2404793594300668,
+      "learning_rate": 3.39977840662139e-05,
+      "loss": 0.467,
+      "step": 1492
+    },
+    {
+      "epoch": 2.9681908548707754,
+      "grad_norm": 0.20090767861090764,
+      "learning_rate": 3.394288889981695e-05,
+      "loss": 0.4533,
+      "step": 1493
+    },
+    {
+      "epoch": 2.970178926441352,
+      "grad_norm": 0.22162509945461517,
+      "learning_rate": 3.38880054067861e-05,
+      "loss": 0.4591,
+      "step": 1494
+    },
+    {
+      "epoch": 2.9721669980119283,
+      "grad_norm": 0.18800071463657672,
+      "learning_rate": 3.38331336928938e-05,
+      "loss": 0.4561,
+      "step": 1495
+    },
+    {
+      "epoch": 2.9741550695825048,
+      "grad_norm": 0.206244471673574,
+      "learning_rate": 3.377827386388971e-05,
+      "loss": 0.4654,
+      "step": 1496
+    },
+    {
+      "epoch": 2.9761431411530817,
+      "grad_norm": 0.22880555976938174,
+      "learning_rate": 3.372342602550062e-05,
+      "loss": 0.4558,
+      "step": 1497
+    },
+    {
+      "epoch": 2.978131212723658,
+      "grad_norm": 0.17949305996147424,
+      "learning_rate": 3.3668590283430185e-05,
+      "loss": 0.4627,
+      "step": 1498
+    },
+    {
+      "epoch": 2.9801192842942346,
+      "grad_norm": 0.18517810607930787,
+      "learning_rate": 3.3613766743358785e-05,
+      "loss": 0.4597,
+      "step": 1499
+    },
+    {
+      "epoch": 2.982107355864811,
+      "grad_norm": 0.23291876613271462,
+      "learning_rate": 3.3558955510943256e-05,
+      "loss": 0.4669,
+      "step": 1500
+    },
+    {
+      "epoch": 2.9840954274353875,
+      "grad_norm": 0.1820730482432967,
+      "learning_rate": 3.350415669181673e-05,
+      "loss": 0.4555,
+      "step": 1501
+    },
+    {
+      "epoch": 2.9860834990059644,
+      "grad_norm": 0.19413753257771896,
+      "learning_rate": 3.3449370391588424e-05,
+      "loss": 0.4633,
+      "step": 1502
+    },
+    {
+      "epoch": 2.988071570576541,
+      "grad_norm": 0.19346750713312172,
+      "learning_rate": 3.339459671584337e-05,
+      "loss": 0.4555,
+      "step": 1503
+    },
+    {
+      "epoch": 2.9900596421471173,
+      "grad_norm": 0.18401415157508247,
+      "learning_rate": 3.3339835770142364e-05,
+      "loss": 0.4621,
+      "step": 1504
+    },
+    {
+      "epoch": 2.9920477137176937,
+      "grad_norm": 0.19007332079668393,
+      "learning_rate": 3.32850876600216e-05,
+      "loss": 0.4604,
+      "step": 1505
+    },
+    {
+      "epoch": 2.99403578528827,
+      "grad_norm": 0.20364951200142725,
+      "learning_rate": 3.323035249099256e-05,
+      "loss": 0.4643,
+      "step": 1506
+    },
+    {
+      "epoch": 2.996023856858847,
+      "grad_norm": 0.17759550097595844,
+      "learning_rate": 3.3175630368541776e-05,
+      "loss": 0.4589,
+      "step": 1507
+    },
+    {
+      "epoch": 2.9980119284294235,
+      "grad_norm": 0.19573811055059043,
+      "learning_rate": 3.312092139813064e-05,
+      "loss": 0.461,
+      "step": 1508
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.18051472969551505,
+      "learning_rate": 3.3066225685195194e-05,
+      "loss": 0.4438,
+      "step": 1509
+    },
+    {
+      "epoch": 3.0019880715705765,
+      "grad_norm": 0.21866713244429514,
+      "learning_rate": 3.301154333514594e-05,
+      "loss": 0.4275,
+      "step": 1510
+    },
+    {
+      "epoch": 3.003976143141153,
+      "grad_norm": 0.19247081598906532,
+      "learning_rate": 3.295687445336762e-05,
+      "loss": 0.4225,
+      "step": 1511
+    },
+    {
+      "epoch": 3.00596421471173,
+      "grad_norm": 0.2309539724817374,
+      "learning_rate": 3.290221914521899e-05,
+      "loss": 0.4219,
+      "step": 1512
+    },
+    {
+      "epoch": 3.0079522862823063,
+      "grad_norm": 0.21933537357774005,
+      "learning_rate": 3.284757751603273e-05,
+      "loss": 0.4293,
+      "step": 1513
+    },
+    {
+      "epoch": 3.0099403578528827,
+      "grad_norm": 0.23026407562644793,
+      "learning_rate": 3.279294967111508e-05,
+      "loss": 0.4192,
+      "step": 1514
+    },
+    {
+      "epoch": 3.011928429423459,
+      "grad_norm": 0.23983833817576242,
+      "learning_rate": 3.273833571574574e-05,
+      "loss": 0.4207,
+      "step": 1515
+    },
+    {
+      "epoch": 3.0139165009940356,
+      "grad_norm": 0.18930871849206557,
+      "learning_rate": 3.268373575517764e-05,
+      "loss": 0.4326,
+      "step": 1516
+    },
+    {
+      "epoch": 3.0159045725646125,
+      "grad_norm": 0.1911013501133307,
+      "learning_rate": 3.262914989463673e-05,
+      "loss": 0.429,
+      "step": 1517
+    },
+    {
+      "epoch": 3.017892644135189,
+      "grad_norm": 0.20511791924121267,
+      "learning_rate": 3.257457823932181e-05,
+      "loss": 0.426,
+      "step": 1518
+    },
+    {
+      "epoch": 3.0198807157057654,
+      "grad_norm": 0.20338555749017675,
+      "learning_rate": 3.252002089440429e-05,
+      "loss": 0.4288,
+      "step": 1519
+    },
+    {
+      "epoch": 3.021868787276342,
+      "grad_norm": 0.2266632220816163,
+      "learning_rate": 3.2465477965028e-05,
+      "loss": 0.4095,
+      "step": 1520
+    },
+    {
+      "epoch": 3.0238568588469183,
+      "grad_norm": 0.1895880120844912,
+      "learning_rate": 3.241094955630895e-05,
+      "loss": 0.4261,
+      "step": 1521
+    },
+    {
+      "epoch": 3.0258449304174952,
+      "grad_norm": 0.23483445478957796,
+      "learning_rate": 3.235643577333527e-05,
+      "loss": 0.4193,
+      "step": 1522
+    },
+    {
+      "epoch": 3.0278330019880717,
+      "grad_norm": 0.3732412917899254,
+      "learning_rate": 3.230193672116681e-05,
+      "loss": 0.4239,
+      "step": 1523
+    },
+    {
+      "epoch": 3.029821073558648,
+      "grad_norm": 0.21687047596976053,
+      "learning_rate": 3.224745250483504e-05,
+      "loss": 0.4278,
+      "step": 1524
+    },
+    {
+      "epoch": 3.0318091451292246,
+      "grad_norm": 0.2039436298095718,
+      "learning_rate": 3.219298322934287e-05,
+      "loss": 0.4184,
+      "step": 1525
+    },
+    {
+      "epoch": 3.033797216699801,
+      "grad_norm": 0.19594156067182716,
+      "learning_rate": 3.213852899966441e-05,
+      "loss": 0.4259,
+      "step": 1526
+    },
+    {
+      "epoch": 3.0357852882703775,
+      "grad_norm": 0.20179469518562662,
+      "learning_rate": 3.2084089920744746e-05,
+      "loss": 0.4309,
+      "step": 1527
+    },
+    {
+      "epoch": 3.0377733598409544,
+      "grad_norm": 0.19670377201579528,
+      "learning_rate": 3.2029666097499804e-05,
+      "loss": 0.416,
+      "step": 1528
+    },
+    {
+      "epoch": 3.039761431411531,
+      "grad_norm": 0.17957230536692131,
+      "learning_rate": 3.197525763481608e-05,
+      "loss": 0.422,
+      "step": 1529
+    },
+    {
+      "epoch": 3.0417495029821073,
+      "grad_norm": 0.15981443945964646,
+      "learning_rate": 3.1920864637550454e-05,
+      "loss": 0.4205,
+      "step": 1530
+    },
+    {
+      "epoch": 3.043737574552684,
+      "grad_norm": 0.20291061700552804,
+      "learning_rate": 3.186648721053005e-05,
+      "loss": 0.4269,
+      "step": 1531
+    },
+    {
+      "epoch": 3.0457256461232602,
+      "grad_norm": 0.17478282257132408,
+      "learning_rate": 3.1812125458551974e-05,
+      "loss": 0.4132,
+      "step": 1532
+    },
+    {
+      "epoch": 3.047713717693837,
+      "grad_norm": 0.16948145356552408,
+      "learning_rate": 3.175777948638307e-05,
+      "loss": 0.4289,
+      "step": 1533
+    },
+    {
+      "epoch": 3.0497017892644136,
+      "grad_norm": 0.16854257729121044,
+      "learning_rate": 3.170344939875982e-05,
+      "loss": 0.4282,
+      "step": 1534
+    },
+    {
+      "epoch": 3.05168986083499,
+      "grad_norm": 0.16732134458370454,
+      "learning_rate": 3.164913530038808e-05,
+      "loss": 0.4177,
+      "step": 1535
+    },
+    {
+      "epoch": 3.0536779324055665,
+      "grad_norm": 0.17964783746191185,
+      "learning_rate": 3.15948372959429e-05,
+      "loss": 0.4156,
+      "step": 1536
+    },
+    {
+      "epoch": 3.055666003976143,
+      "grad_norm": 0.17083246059369028,
+      "learning_rate": 3.1540555490068304e-05,
+      "loss": 0.4212,
+      "step": 1537
+    },
+    {
+      "epoch": 3.05765407554672,
+      "grad_norm": 0.17480502617064078,
+      "learning_rate": 3.14862899873771e-05,
+      "loss": 0.4201,
+      "step": 1538
+    },
+    {
+      "epoch": 3.0596421471172963,
+      "grad_norm": 0.15811181049785739,
+      "learning_rate": 3.143204089245063e-05,
+      "loss": 0.4236,
+      "step": 1539
+    },
+    {
+      "epoch": 3.0616302186878728,
+      "grad_norm": 0.20090880336661895,
+      "learning_rate": 3.137780830983873e-05,
+      "loss": 0.4164,
+      "step": 1540
+    },
+    {
+      "epoch": 3.0636182902584492,
+      "grad_norm": 0.2116311394564602,
+      "learning_rate": 3.132359234405931e-05,
+      "loss": 0.4285,
+      "step": 1541
+    },
+    {
+      "epoch": 3.0656063618290257,
+      "grad_norm": 0.17291665249112143,
+      "learning_rate": 3.12693930995983e-05,
+      "loss": 0.4227,
+      "step": 1542
+    },
+    {
+      "epoch": 3.0675944333996026,
+      "grad_norm": 0.20723641513115693,
+      "learning_rate": 3.121521068090937e-05,
+      "loss": 0.4265,
+      "step": 1543
+    },
+    {
+      "epoch": 3.069582504970179,
+      "grad_norm": 0.1839141741250624,
+      "learning_rate": 3.11610451924138e-05,
+      "loss": 0.4234,
+      "step": 1544
+    },
+    {
+      "epoch": 3.0715705765407555,
+      "grad_norm": 0.19373127551015817,
+      "learning_rate": 3.110689673850023e-05,
+      "loss": 0.4191,
+      "step": 1545
+    },
+    {
+      "epoch": 3.073558648111332,
+      "grad_norm": 0.1660825761301819,
+      "learning_rate": 3.105276542352447e-05,
+      "loss": 0.4267,
+      "step": 1546
+    },
+    {
+      "epoch": 3.0755467196819084,
+      "grad_norm": 0.1648308561329296,
+      "learning_rate": 3.0998651351809305e-05,
+      "loss": 0.4217,
+      "step": 1547
+    },
+    {
+      "epoch": 3.0775347912524853,
+      "grad_norm": 0.18042851709340152,
+      "learning_rate": 3.094455462764425e-05,
+      "loss": 0.4211,
+      "step": 1548
+    },
+    {
+      "epoch": 3.0795228628230618,
+      "grad_norm": 0.18388627104468944,
+      "learning_rate": 3.089047535528545e-05,
+      "loss": 0.4187,
+      "step": 1549
+    },
+    {
+      "epoch": 3.081510934393638,
+      "grad_norm": 0.18338332550558414,
+      "learning_rate": 3.08364136389554e-05,
+      "loss": 0.4222,
+      "step": 1550
+    },
+    {
+      "epoch": 3.0834990059642147,
+      "grad_norm": 0.5457393482733713,
+      "learning_rate": 3.07823695828427e-05,
+      "loss": 0.4392,
+      "step": 1551
+    },
+    {
+      "epoch": 3.085487077534791,
+      "grad_norm": 0.16946702499846994,
+      "learning_rate": 3.0728343291102e-05,
+      "loss": 0.425,
+      "step": 1552
+    },
+    {
+      "epoch": 3.0874751491053676,
+      "grad_norm": 0.21982162026792132,
+      "learning_rate": 3.067433486785366e-05,
+      "loss": 0.4252,
+      "step": 1553
+    },
+    {
+      "epoch": 3.0894632206759445,
+      "grad_norm": 0.20012161267571643,
+      "learning_rate": 3.062034441718362e-05,
+      "loss": 0.4271,
+      "step": 1554
+    },
+    {
+      "epoch": 3.091451292246521,
+      "grad_norm": 0.17533381870496842,
+      "learning_rate": 3.056637204314318e-05,
+      "loss": 0.4179,
+      "step": 1555
+    },
+    {
+      "epoch": 3.0934393638170974,
+      "grad_norm": 0.3163113330290082,
+      "learning_rate": 3.0512417849748805e-05,
+      "loss": 0.4221,
+      "step": 1556
+    },
+    {
+      "epoch": 3.095427435387674,
+      "grad_norm": 0.1649768820807243,
+      "learning_rate": 3.0458481940981933e-05,
+      "loss": 0.432,
+      "step": 1557
+    },
+    {
+      "epoch": 3.0974155069582503,
+      "grad_norm": 0.15470699826266648,
+      "learning_rate": 3.0404564420788745e-05,
+      "loss": 0.4308,
+      "step": 1558
+    },
+    {
+      "epoch": 3.099403578528827,
+      "grad_norm": 0.1702220476604976,
+      "learning_rate": 3.035066539308e-05,
+      "loss": 0.427,
+      "step": 1559
+    },
+    {
+      "epoch": 3.1013916500994037,
+      "grad_norm": 0.18688827814809164,
+      "learning_rate": 3.029678496173079e-05,
+      "loss": 0.4289,
+      "step": 1560
+    },
+    {
+      "epoch": 3.10337972166998,
+      "grad_norm": 0.14448275902724284,
+      "learning_rate": 3.024292323058039e-05,
+      "loss": 0.4204,
+      "step": 1561
+    },
+    {
+      "epoch": 3.1053677932405566,
+      "grad_norm": 0.1825200775041438,
+      "learning_rate": 3.0189080303432044e-05,
+      "loss": 0.4264,
+      "step": 1562
+    },
+    {
+      "epoch": 3.107355864811133,
+      "grad_norm": 0.15196658182073003,
+      "learning_rate": 3.013525628405274e-05,
+      "loss": 0.4248,
+      "step": 1563
+    },
+    {
+      "epoch": 3.10934393638171,
+      "grad_norm": 0.23580394328013937,
+      "learning_rate": 3.008145127617302e-05,
+      "loss": 0.4274,
+      "step": 1564
+    },
+    {
+      "epoch": 3.1113320079522864,
+      "grad_norm": 0.17167581205154928,
+      "learning_rate": 3.0027665383486803e-05,
+      "loss": 0.419,
+      "step": 1565
+    },
+    {
+      "epoch": 3.113320079522863,
+      "grad_norm": 0.17616777841111306,
+      "learning_rate": 2.997389870965118e-05,
+      "loss": 0.4213,
+      "step": 1566
+    },
+    {
+      "epoch": 3.1153081510934393,
+      "grad_norm": 0.17527288324883256,
+      "learning_rate": 2.9920151358286166e-05,
+      "loss": 0.425,
+      "step": 1567
+    },
+    {
+      "epoch": 3.1172962226640157,
+      "grad_norm": 0.17865409252565498,
+      "learning_rate": 2.986642343297458e-05,
+      "loss": 0.4245,
+      "step": 1568
+    },
+    {
+      "epoch": 3.1192842942345926,
+      "grad_norm": 0.17682031138461274,
+      "learning_rate": 2.9812715037261756e-05,
+      "loss": 0.4291,
+      "step": 1569
+    },
+    {
+      "epoch": 3.121272365805169,
+      "grad_norm": 0.16631730027106617,
+      "learning_rate": 2.9759026274655433e-05,
+      "loss": 0.4301,
+      "step": 1570
+    },
+    {
+      "epoch": 3.1232604373757455,
+      "grad_norm": 0.3566716216311293,
+      "learning_rate": 2.970535724862548e-05,
+      "loss": 0.4324,
+      "step": 1571
+    },
+    {
+      "epoch": 3.125248508946322,
+      "grad_norm": 0.16423762796865618,
+      "learning_rate": 2.9651708062603763e-05,
+      "loss": 0.4223,
+      "step": 1572
+    },
+    {
+      "epoch": 3.1272365805168985,
+      "grad_norm": 0.15581656406502245,
+      "learning_rate": 2.959807881998389e-05,
+      "loss": 0.4152,
+      "step": 1573
+    },
+    {
+      "epoch": 3.1292246520874754,
+      "grad_norm": 0.17130907591898645,
+      "learning_rate": 2.9544469624121e-05,
+      "loss": 0.4206,
+      "step": 1574
+    },
+    {
+      "epoch": 3.131212723658052,
+      "grad_norm": 0.16199415820717575,
+      "learning_rate": 2.9490880578331685e-05,
+      "loss": 0.4222,
+      "step": 1575
+    },
+    {
+      "epoch": 3.1332007952286283,
+      "grad_norm": 0.16445570164173293,
+      "learning_rate": 2.9437311785893635e-05,
+      "loss": 0.419,
+      "step": 1576
+    },
+    {
+      "epoch": 3.1351888667992047,
+      "grad_norm": 0.16037847818725806,
+      "learning_rate": 2.938376335004553e-05,
+      "loss": 0.4228,
+      "step": 1577
+    },
+    {
+      "epoch": 3.137176938369781,
+      "grad_norm": 0.16786978324050583,
+      "learning_rate": 2.9330235373986803e-05,
+      "loss": 0.4247,
+      "step": 1578
+    },
+    {
+      "epoch": 3.139165009940358,
+      "grad_norm": 0.18278838840353107,
+      "learning_rate": 2.9276727960877473e-05,
+      "loss": 0.4315,
+      "step": 1579
+    },
+    {
+      "epoch": 3.1411530815109345,
+      "grad_norm": 0.1651561323450754,
+      "learning_rate": 2.9223241213837923e-05,
+      "loss": 0.4092,
+      "step": 1580
+    },
+    {
+      "epoch": 3.143141153081511,
+      "grad_norm": 0.16966254220617266,
+      "learning_rate": 2.916977523594871e-05,
+      "loss": 0.4181,
+      "step": 1581
+    },
+    {
+      "epoch": 3.1451292246520874,
+      "grad_norm": 0.17505463062318927,
+      "learning_rate": 2.9116330130250374e-05,
+      "loss": 0.4266,
+      "step": 1582
+    },
+    {
+      "epoch": 3.147117296222664,
+      "grad_norm": 0.16115989680942328,
+      "learning_rate": 2.906290599974318e-05,
+      "loss": 0.4319,
+      "step": 1583
+    },
+    {
+      "epoch": 3.1491053677932404,
+      "grad_norm": 0.18383791657309656,
+      "learning_rate": 2.9009502947387056e-05,
+      "loss": 0.4194,
+      "step": 1584
+    },
+    {
+      "epoch": 3.1510934393638173,
+      "grad_norm": 0.14571585197850928,
+      "learning_rate": 2.8956121076101244e-05,
+      "loss": 0.4231,
+      "step": 1585
+    },
+    {
+      "epoch": 3.1530815109343937,
+      "grad_norm": 0.17282383882108132,
+      "learning_rate": 2.890276048876419e-05,
+      "loss": 0.4198,
+      "step": 1586
+    },
+    {
+      "epoch": 3.15506958250497,
+      "grad_norm": 0.16960750212497086,
+      "learning_rate": 2.884942128821329e-05,
+      "loss": 0.4203,
+      "step": 1587
+    },
+    {
+      "epoch": 3.1570576540755466,
+      "grad_norm": 0.14433740341835166,
+      "learning_rate": 2.8796103577244766e-05,
+      "loss": 0.4194,
+      "step": 1588
+    },
+    {
+      "epoch": 3.159045725646123,
+      "grad_norm": 0.18244597958945752,
+      "learning_rate": 2.8742807458613393e-05,
+      "loss": 0.4205,
+      "step": 1589
+    },
+    {
+      "epoch": 3.1610337972167,
+      "grad_norm": 0.2348787312272368,
+      "learning_rate": 2.8689533035032355e-05,
+      "loss": 0.4257,
+      "step": 1590
+    },
+    {
+      "epoch": 3.1630218687872764,
+      "grad_norm": 0.1731480343868321,
+      "learning_rate": 2.8636280409173012e-05,
+      "loss": 0.4278,
+      "step": 1591
+    },
+    {
+      "epoch": 3.165009940357853,
+      "grad_norm": 0.15285486539504065,
+      "learning_rate": 2.8583049683664698e-05,
+      "loss": 0.4245,
+      "step": 1592
+    },
+    {
+      "epoch": 3.1669980119284293,
+      "grad_norm": 1.1577676169956606,
+      "learning_rate": 2.8529840961094596e-05,
+      "loss": 0.4473,
+      "step": 1593
+    },
+    {
+      "epoch": 3.168986083499006,
+      "grad_norm": 0.15314522438830153,
+      "learning_rate": 2.8476654344007438e-05,
+      "loss": 0.417,
+      "step": 1594
+    },
+    {
+      "epoch": 3.1709741550695827,
+      "grad_norm": 0.1854340474178689,
+      "learning_rate": 2.8423489934905348e-05,
+      "loss": 0.4199,
+      "step": 1595
+    },
+    {
+      "epoch": 3.172962226640159,
+      "grad_norm": 0.2189527140355646,
+      "learning_rate": 2.837034783624768e-05,
+      "loss": 0.4232,
+      "step": 1596
+    },
+    {
+      "epoch": 3.1749502982107356,
+      "grad_norm": 0.16477213116978373,
+      "learning_rate": 2.8317228150450775e-05,
+      "loss": 0.4232,
+      "step": 1597
+    },
+    {
+      "epoch": 3.176938369781312,
+      "grad_norm": 0.18766127242746655,
+      "learning_rate": 2.8264130979887786e-05,
+      "loss": 0.4189,
+      "step": 1598
+    },
+    {
+      "epoch": 3.1789264413518885,
+      "grad_norm": 0.16962240140631765,
+      "learning_rate": 2.8211056426888466e-05,
+      "loss": 0.4303,
+      "step": 1599
+    },
+    {
+      "epoch": 3.1809145129224654,
+      "grad_norm": 0.23562016505996533,
+      "learning_rate": 2.8158004593738994e-05,
+      "loss": 0.4198,
+      "step": 1600
+    },
+    {
+      "epoch": 3.182902584493042,
+      "grad_norm": 0.16522091902575367,
+      "learning_rate": 2.810497558268172e-05,
+      "loss": 0.4236,
+      "step": 1601
+    },
+    {
+      "epoch": 3.1848906560636183,
+      "grad_norm": 0.20700531289678575,
+      "learning_rate": 2.8051969495915087e-05,
+      "loss": 0.4253,
+      "step": 1602
+    },
+    {
+      "epoch": 3.1868787276341948,
+      "grad_norm": 0.15198735404558872,
+      "learning_rate": 2.7998986435593306e-05,
+      "loss": 0.4235,
+      "step": 1603
+    },
+    {
+      "epoch": 3.1888667992047712,
+      "grad_norm": 0.16179795866633792,
+      "learning_rate": 2.7946026503826203e-05,
+      "loss": 0.4223,
+      "step": 1604
+    },
+    {
+      "epoch": 3.1908548707753477,
+      "grad_norm": 0.16649993865726812,
+      "learning_rate": 2.7893089802679055e-05,
+      "loss": 0.4257,
+      "step": 1605
+    },
+    {
+      "epoch": 3.1928429423459246,
+      "grad_norm": 0.15903132930020156,
+      "learning_rate": 2.7840176434172357e-05,
+      "loss": 0.4263,
+      "step": 1606
+    },
+    {
+      "epoch": 3.194831013916501,
+      "grad_norm": 0.16054210668205585,
+      "learning_rate": 2.778728650028164e-05,
+      "loss": 0.4182,
+      "step": 1607
+    },
+    {
+      "epoch": 3.1968190854870775,
+      "grad_norm": 0.15793274003584565,
+      "learning_rate": 2.773442010293728e-05,
+      "loss": 0.4196,
+      "step": 1608
+    },
+    {
+      "epoch": 3.198807157057654,
+      "grad_norm": 0.16707398647771807,
+      "learning_rate": 2.7681577344024274e-05,
+      "loss": 0.4291,
+      "step": 1609
+    },
+    {
+      "epoch": 3.2007952286282304,
+      "grad_norm": 0.1615709751294597,
+      "learning_rate": 2.7628758325382086e-05,
+      "loss": 0.4196,
+      "step": 1610
+    },
+    {
+      "epoch": 3.2027833001988073,
+      "grad_norm": 0.18962014963322602,
+      "learning_rate": 2.757596314880441e-05,
+      "loss": 0.4249,
+      "step": 1611
+    },
+    {
+      "epoch": 3.2047713717693838,
+      "grad_norm": 0.2498774125522842,
+      "learning_rate": 2.7523191916039e-05,
+      "loss": 0.4244,
+      "step": 1612
+    },
+    {
+      "epoch": 3.20675944333996,
+      "grad_norm": 0.33194601474779295,
+      "learning_rate": 2.7470444728787457e-05,
+      "loss": 0.4236,
+      "step": 1613
+    },
+    {
+      "epoch": 3.2087475149105367,
+      "grad_norm": 0.14356716664383448,
+      "learning_rate": 2.7417721688705038e-05,
+      "loss": 0.412,
+      "step": 1614
+    },
+    {
+      "epoch": 3.210735586481113,
+      "grad_norm": 0.1605633240701923,
+      "learning_rate": 2.7365022897400474e-05,
+      "loss": 0.424,
+      "step": 1615
+    },
+    {
+      "epoch": 3.21272365805169,
+      "grad_norm": 0.15677114515341178,
+      "learning_rate": 2.7312348456435766e-05,
+      "loss": 0.4158,
+      "step": 1616
+    },
+    {
+      "epoch": 3.2147117296222665,
+      "grad_norm": 0.14935248284076827,
+      "learning_rate": 2.7259698467325977e-05,
+      "loss": 0.433,
+      "step": 1617
+    },
+    {
+      "epoch": 3.216699801192843,
+      "grad_norm": 0.1451275646279355,
+      "learning_rate": 2.720707303153904e-05,
+      "loss": 0.42,
+      "step": 1618
+    },
+    {
+      "epoch": 3.2186878727634194,
+      "grad_norm": 0.14493464992235816,
+      "learning_rate": 2.7154472250495597e-05,
+      "loss": 0.4203,
+      "step": 1619
+    },
+    {
+      "epoch": 3.220675944333996,
+      "grad_norm": 0.14380332893928113,
+      "learning_rate": 2.7101896225568744e-05,
+      "loss": 0.4295,
+      "step": 1620
+    },
+    {
+      "epoch": 3.2226640159045727,
+      "grad_norm": 0.14091520617084688,
+      "learning_rate": 2.704934505808389e-05,
+      "loss": 0.4178,
+      "step": 1621
+    },
+    {
+      "epoch": 3.224652087475149,
+      "grad_norm": 0.14513943134763385,
+      "learning_rate": 2.6996818849318508e-05,
+      "loss": 0.4183,
+      "step": 1622
+    },
+    {
+      "epoch": 3.2266401590457257,
+      "grad_norm": 0.14432815766342816,
+      "learning_rate": 2.6944317700501992e-05,
+      "loss": 0.4166,
+      "step": 1623
+    },
+    {
+      "epoch": 3.228628230616302,
+      "grad_norm": 0.16739958916754763,
+      "learning_rate": 2.6891841712815448e-05,
+      "loss": 0.4298,
+      "step": 1624
+    },
+    {
+      "epoch": 3.2306163021868786,
+      "grad_norm": 0.17632490104186138,
+      "learning_rate": 2.6839390987391475e-05,
+      "loss": 0.4151,
+      "step": 1625
+    },
+    {
+      "epoch": 3.2326043737574555,
+      "grad_norm": 0.1701061576690492,
+      "learning_rate": 2.678696562531398e-05,
+      "loss": 0.4186,
+      "step": 1626
+    },
+    {
+      "epoch": 3.234592445328032,
+      "grad_norm": 0.14923451020759781,
+      "learning_rate": 2.6734565727618016e-05,
+      "loss": 0.4226,
+      "step": 1627
+    },
+    {
+      "epoch": 3.2365805168986084,
+      "grad_norm": 0.1402483490014172,
+      "learning_rate": 2.6682191395289537e-05,
+      "loss": 0.4277,
+      "step": 1628
+    },
+    {
+      "epoch": 3.238568588469185,
+      "grad_norm": 0.21580075521102834,
+      "learning_rate": 2.6629842729265246e-05,
+      "loss": 0.4096,
+      "step": 1629
+    },
+    {
+      "epoch": 3.2405566600397613,
+      "grad_norm": 0.24086136258461305,
+      "learning_rate": 2.6577519830432365e-05,
+      "loss": 0.4206,
+      "step": 1630
+    },
+    {
+      "epoch": 3.242544731610338,
+      "grad_norm": 0.17137383827651592,
+      "learning_rate": 2.6525222799628455e-05,
+      "loss": 0.4153,
+      "step": 1631
+    },
+    {
+      "epoch": 3.2445328031809146,
+      "grad_norm": 0.27384042950995446,
+      "learning_rate": 2.647295173764123e-05,
+      "loss": 0.4216,
+      "step": 1632
+    },
+    {
+      "epoch": 3.246520874751491,
+      "grad_norm": 0.23612530448432742,
+      "learning_rate": 2.6420706745208368e-05,
+      "loss": 0.4158,
+      "step": 1633
+    },
+    {
+      "epoch": 3.2485089463220675,
+      "grad_norm": 0.19382473812414455,
+      "learning_rate": 2.6368487923017286e-05,
+      "loss": 0.4225,
+      "step": 1634
+    },
+    {
+      "epoch": 3.250497017892644,
+      "grad_norm": 0.2945482875629765,
+      "learning_rate": 2.631629537170498e-05,
+      "loss": 0.4223,
+      "step": 1635
+    },
+    {
+      "epoch": 3.252485089463221,
+      "grad_norm": 0.1557550330207868,
+      "learning_rate": 2.6264129191857805e-05,
+      "loss": 0.4145,
+      "step": 1636
+    },
+    {
+      "epoch": 3.2544731610337974,
+      "grad_norm": 0.2396368267625787,
+      "learning_rate": 2.62119894840113e-05,
+      "loss": 0.416,
+      "step": 1637
+    },
+    {
+      "epoch": 3.256461232604374,
+      "grad_norm": 0.17272951015488022,
+      "learning_rate": 2.6159876348649994e-05,
+      "loss": 0.4314,
+      "step": 1638
+    },
+    {
+      "epoch": 3.2584493041749503,
+      "grad_norm": 0.1765073891960342,
+      "learning_rate": 2.6107789886207195e-05,
+      "loss": 0.416,
+      "step": 1639
+    },
+    {
+      "epoch": 3.2604373757455267,
+      "grad_norm": 0.19928549933644896,
+      "learning_rate": 2.605573019706479e-05,
+      "loss": 0.4262,
+      "step": 1640
+    },
+    {
+      "epoch": 3.2624254473161036,
+      "grad_norm": 0.16788212491095592,
+      "learning_rate": 2.60036973815531e-05,
+      "loss": 0.4174,
+      "step": 1641
+    },
+    {
+      "epoch": 3.26441351888668,
+      "grad_norm": 0.2014485731227408,
+      "learning_rate": 2.5951691539950635e-05,
+      "loss": 0.4223,
+      "step": 1642
+    },
+    {
+      "epoch": 3.2664015904572565,
+      "grad_norm": 0.19786684925638573,
+      "learning_rate": 2.5899712772483926e-05,
+      "loss": 0.42,
+      "step": 1643
+    },
+    {
+      "epoch": 3.268389662027833,
+      "grad_norm": 0.17771485372367102,
+      "learning_rate": 2.5847761179327334e-05,
+      "loss": 0.417,
+      "step": 1644
+    },
+    {
+      "epoch": 3.2703777335984094,
+      "grad_norm": 0.18677854197656066,
+      "learning_rate": 2.5795836860602808e-05,
+      "loss": 0.4267,
+      "step": 1645
+    },
+    {
+      "epoch": 3.272365805168986,
+      "grad_norm": 0.15790603246827162,
+      "learning_rate": 2.5743939916379805e-05,
+      "loss": 0.4179,
+      "step": 1646
+    },
+    {
+      "epoch": 3.274353876739563,
+      "grad_norm": 0.1768610231628309,
+      "learning_rate": 2.569207044667498e-05,
+      "loss": 0.4275,
+      "step": 1647
+    },
+    {
+      "epoch": 3.2763419483101393,
+      "grad_norm": 0.2092056932711368,
+      "learning_rate": 2.564022855145204e-05,
+      "loss": 0.4223,
+      "step": 1648
+    },
+    {
+      "epoch": 3.2783300198807157,
+      "grad_norm": 0.1784649030856477,
+      "learning_rate": 2.558841433062155e-05,
+      "loss": 0.4217,
+      "step": 1649
+    },
+    {
+      "epoch": 3.280318091451292,
+      "grad_norm": 0.16695975013592268,
+      "learning_rate": 2.5536627884040748e-05,
+      "loss": 0.4258,
+      "step": 1650
+    },
+    {
+      "epoch": 3.2823061630218686,
+      "grad_norm": 0.16257746343716772,
+      "learning_rate": 2.548486931151334e-05,
+      "loss": 0.4146,
+      "step": 1651
+    },
+    {
+      "epoch": 3.2842942345924455,
+      "grad_norm": 0.16035482107017218,
+      "learning_rate": 2.5433138712789324e-05,
+      "loss": 0.4263,
+      "step": 1652
+    },
+    {
+      "epoch": 3.286282306163022,
+      "grad_norm": 0.1668332300916889,
+      "learning_rate": 2.5381436187564776e-05,
+      "loss": 0.4271,
+      "step": 1653
+    },
+    {
+      "epoch": 3.2882703777335984,
+      "grad_norm": 0.18765944464929687,
+      "learning_rate": 2.5329761835481647e-05,
+      "loss": 0.4305,
+      "step": 1654
+    },
+    {
+      "epoch": 3.290258449304175,
+      "grad_norm": 0.22260684778448311,
+      "learning_rate": 2.5278115756127646e-05,
+      "loss": 0.4158,
+      "step": 1655
+    },
+    {
+      "epoch": 3.2922465208747513,
+      "grad_norm": 0.175348581154186,
+      "learning_rate": 2.5226498049035958e-05,
+      "loss": 0.4187,
+      "step": 1656
+    },
+    {
+      "epoch": 3.294234592445328,
+      "grad_norm": 0.137384946554788,
+      "learning_rate": 2.5174908813685094e-05,
+      "loss": 0.4262,
+      "step": 1657
+    },
+    {
+      "epoch": 3.2962226640159047,
+      "grad_norm": 0.2289530209064395,
+      "learning_rate": 2.5123348149498684e-05,
+      "loss": 0.4176,
+      "step": 1658
+    },
+    {
+      "epoch": 3.298210735586481,
+      "grad_norm": 0.24077183815627887,
+      "learning_rate": 2.5071816155845304e-05,
+      "loss": 0.4232,
+      "step": 1659
+    },
+    {
+      "epoch": 3.3001988071570576,
+      "grad_norm": 0.1515863899105361,
+      "learning_rate": 2.5020312932038285e-05,
+      "loss": 0.4219,
+      "step": 1660
+    },
+    {
+      "epoch": 3.302186878727634,
+      "grad_norm": 0.22154638035278512,
+      "learning_rate": 2.4968838577335507e-05,
+      "loss": 0.4218,
+      "step": 1661
+    },
+    {
+      "epoch": 3.3041749502982105,
+      "grad_norm": 0.2787216556594714,
+      "learning_rate": 2.491739319093919e-05,
+      "loss": 0.4189,
+      "step": 1662
+    },
+    {
+      "epoch": 3.3061630218687874,
+      "grad_norm": 0.15341434669320073,
+      "learning_rate": 2.486597687199577e-05,
+      "loss": 0.4101,
+      "step": 1663
+    },
+    {
+      "epoch": 3.308151093439364,
+      "grad_norm": 0.18926114022594837,
+      "learning_rate": 2.481458971959563e-05,
+      "loss": 0.4279,
+      "step": 1664
+    },
+    {
+      "epoch": 3.3101391650099403,
+      "grad_norm": 0.15294495380705728,
+      "learning_rate": 2.4763231832772966e-05,
+      "loss": 0.4194,
+      "step": 1665
+    },
+    {
+      "epoch": 3.3121272365805168,
+      "grad_norm": 0.15660817811423164,
+      "learning_rate": 2.4711903310505525e-05,
+      "loss": 0.421,
+      "step": 1666
+    },
+    {
+      "epoch": 3.3141153081510932,
+      "grad_norm": 0.162337406875044,
+      "learning_rate": 2.4660604251714526e-05,
+      "loss": 0.4233,
+      "step": 1667
+    },
+    {
+      "epoch": 3.31610337972167,
+      "grad_norm": 0.14327890979444038,
+      "learning_rate": 2.460933475526436e-05,
+      "loss": 0.4118,
+      "step": 1668
+    },
+    {
+      "epoch": 3.3180914512922466,
+      "grad_norm": 0.14965507258044433,
+      "learning_rate": 2.455809491996247e-05,
+      "loss": 0.4228,
+      "step": 1669
+    },
+    {
+      "epoch": 3.320079522862823,
+      "grad_norm": 0.14560822410161192,
+      "learning_rate": 2.4506884844559124e-05,
+      "loss": 0.4148,
+      "step": 1670
+    },
+    {
+      "epoch": 3.3220675944333995,
+      "grad_norm": 0.1426400454941638,
+      "learning_rate": 2.445570462774723e-05,
+      "loss": 0.4297,
+      "step": 1671
+    },
+    {
+      "epoch": 3.324055666003976,
+      "grad_norm": 0.1569048324686495,
+      "learning_rate": 2.4404554368162177e-05,
+      "loss": 0.4142,
+      "step": 1672
+    },
+    {
+      "epoch": 3.326043737574553,
+      "grad_norm": 0.1643392865084774,
+      "learning_rate": 2.4353434164381593e-05,
+      "loss": 0.4206,
+      "step": 1673
+    },
+    {
+      "epoch": 3.3280318091451293,
+      "grad_norm": 0.1498429617187632,
+      "learning_rate": 2.4302344114925203e-05,
+      "loss": 0.4137,
+      "step": 1674
+    },
+    {
+      "epoch": 3.3300198807157058,
+      "grad_norm": 0.13721526264782666,
+      "learning_rate": 2.425128431825459e-05,
+      "loss": 0.4212,
+      "step": 1675
+    },
+    {
+      "epoch": 3.332007952286282,
+      "grad_norm": 0.1600371084528462,
+      "learning_rate": 2.4200254872773056e-05,
+      "loss": 0.4193,
+      "step": 1676
+    },
+    {
+      "epoch": 3.3339960238568587,
+      "grad_norm": 0.15972808411829598,
+      "learning_rate": 2.4149255876825413e-05,
+      "loss": 0.4204,
+      "step": 1677
+    },
+    {
+      "epoch": 3.3359840954274356,
+      "grad_norm": 0.19172287503702923,
+      "learning_rate": 2.409828742869777e-05,
+      "loss": 0.4233,
+      "step": 1678
+    },
+    {
+      "epoch": 3.337972166998012,
+      "grad_norm": 0.17154708460144463,
+      "learning_rate": 2.4047349626617383e-05,
+      "loss": 0.4269,
+      "step": 1679
+    },
+    {
+      "epoch": 3.3399602385685885,
+      "grad_norm": 0.14615085908668787,
+      "learning_rate": 2.3996442568752434e-05,
+      "loss": 0.4281,
+      "step": 1680
+    },
+    {
+      "epoch": 3.341948310139165,
+      "grad_norm": 0.16245718655783098,
+      "learning_rate": 2.3945566353211872e-05,
+      "loss": 0.4192,
+      "step": 1681
+    },
+    {
+      "epoch": 3.3439363817097414,
+      "grad_norm": 0.13505783886664183,
+      "learning_rate": 2.389472107804518e-05,
+      "loss": 0.4072,
+      "step": 1682
+    },
+    {
+      "epoch": 3.3459244532803183,
+      "grad_norm": 0.14338107591716415,
+      "learning_rate": 2.3843906841242234e-05,
+      "loss": 0.4165,
+      "step": 1683
+    },
+    {
+      "epoch": 3.3479125248508947,
+      "grad_norm": 0.13551779503546382,
+      "learning_rate": 2.3793123740733074e-05,
+      "loss": 0.419,
+      "step": 1684
+    },
+    {
+      "epoch": 3.349900596421471,
+      "grad_norm": 0.18878082132312443,
+      "learning_rate": 2.374237187438775e-05,
+      "loss": 0.4262,
+      "step": 1685
+    },
+    {
+      "epoch": 3.3518886679920477,
+      "grad_norm": 0.19720983891195784,
+      "learning_rate": 2.3691651340016106e-05,
+      "loss": 0.413,
+      "step": 1686
+    },
+    {
+      "epoch": 3.353876739562624,
+      "grad_norm": 0.15864169989169807,
+      "learning_rate": 2.364096223536761e-05,
+      "loss": 0.4127,
+      "step": 1687
+    },
+    {
+      "epoch": 3.355864811133201,
+      "grad_norm": 0.1965166357303236,
+      "learning_rate": 2.359030465813115e-05,
+      "loss": 0.4207,
+      "step": 1688
+    },
+    {
+      "epoch": 3.3578528827037775,
+      "grad_norm": 0.38613530552613956,
+      "learning_rate": 2.3539678705934854e-05,
+      "loss": 0.4317,
+      "step": 1689
+    },
+    {
+      "epoch": 3.359840954274354,
+      "grad_norm": 0.18727427096552424,
+      "learning_rate": 2.348908447634593e-05,
+      "loss": 0.417,
+      "step": 1690
+    },
+    {
+      "epoch": 3.3618290258449304,
+      "grad_norm": 0.17011695468121943,
+      "learning_rate": 2.3438522066870405e-05,
+      "loss": 0.4196,
+      "step": 1691
+    },
+    {
+      "epoch": 3.363817097415507,
+      "grad_norm": 0.18341050658376595,
+      "learning_rate": 2.338799157495302e-05,
+      "loss": 0.4202,
+      "step": 1692
+    },
+    {
+      "epoch": 3.3658051689860837,
+      "grad_norm": 0.1775335051454464,
+      "learning_rate": 2.333749309797696e-05,
+      "loss": 0.4177,
+      "step": 1693
+    },
+    {
+      "epoch": 3.36779324055666,
+      "grad_norm": 0.15602473128210675,
+      "learning_rate": 2.328702673326376e-05,
+      "loss": 0.4246,
+      "step": 1694
+    },
+    {
+      "epoch": 3.3697813121272366,
+      "grad_norm": 0.16113387149141137,
+      "learning_rate": 2.3236592578073056e-05,
+      "loss": 0.4221,
+      "step": 1695
+    },
+    {
+      "epoch": 3.371769383697813,
+      "grad_norm": 0.14998557774497948,
+      "learning_rate": 2.3186190729602364e-05,
+      "loss": 0.4279,
+      "step": 1696
+    },
+    {
+      "epoch": 3.3737574552683895,
+      "grad_norm": 0.15917274374720317,
+      "learning_rate": 2.3135821284986983e-05,
+      "loss": 0.4232,
+      "step": 1697
+    },
+    {
+      "epoch": 3.3757455268389664,
+      "grad_norm": 0.17516252336995275,
+      "learning_rate": 2.3085484341299743e-05,
+      "loss": 0.4235,
+      "step": 1698
+    },
+    {
+      "epoch": 3.377733598409543,
+      "grad_norm": 0.1515880856245295,
+      "learning_rate": 2.3035179995550876e-05,
+      "loss": 0.4208,
+      "step": 1699
+    },
+    {
+      "epoch": 3.3797216699801194,
+      "grad_norm": 0.19601830982433405,
+      "learning_rate": 2.2984908344687745e-05,
+      "loss": 0.4236,
+      "step": 1700
+    },
+    {
+      "epoch": 3.381709741550696,
+      "grad_norm": 0.18377008019381375,
+      "learning_rate": 2.2934669485594716e-05,
+      "loss": 0.4211,
+      "step": 1701
+    },
+    {
+      "epoch": 3.3836978131212723,
+      "grad_norm": 0.16561826986439118,
+      "learning_rate": 2.288446351509297e-05,
+      "loss": 0.4262,
+      "step": 1702
+    },
+    {
+      "epoch": 3.3856858846918487,
+      "grad_norm": 0.21353955552866088,
+      "learning_rate": 2.2834290529940284e-05,
+      "loss": 0.4242,
+      "step": 1703
+    },
+    {
+      "epoch": 3.3876739562624256,
+      "grad_norm": 0.14923813005518957,
+      "learning_rate": 2.2784150626830898e-05,
+      "loss": 0.4217,
+      "step": 1704
+    },
+    {
+      "epoch": 3.389662027833002,
+      "grad_norm": 0.24069279516554903,
+      "learning_rate": 2.2734043902395245e-05,
+      "loss": 0.4176,
+      "step": 1705
+    },
+    {
+      "epoch": 3.3916500994035785,
+      "grad_norm": 0.1702158291378274,
+      "learning_rate": 2.2683970453199865e-05,
+      "loss": 0.4231,
+      "step": 1706
+    },
+    {
+      "epoch": 3.393638170974155,
+      "grad_norm": 0.17350287526667996,
+      "learning_rate": 2.2633930375747118e-05,
+      "loss": 0.4183,
+      "step": 1707
+    },
+    {
+      "epoch": 3.3956262425447314,
+      "grad_norm": 0.23437538422451198,
+      "learning_rate": 2.2583923766475125e-05,
+      "loss": 0.4178,
+      "step": 1708
+    },
+    {
+      "epoch": 3.3976143141153083,
+      "grad_norm": 0.14198550996726472,
+      "learning_rate": 2.2533950721757447e-05,
+      "loss": 0.414,
+      "step": 1709
+    },
+    {
+      "epoch": 3.399602385685885,
+      "grad_norm": 0.2102399025316218,
+      "learning_rate": 2.2484011337902968e-05,
+      "loss": 0.424,
+      "step": 1710
+    },
+    {
+      "epoch": 3.4015904572564613,
+      "grad_norm": 0.1576665890073455,
+      "learning_rate": 2.2434105711155706e-05,
+      "loss": 0.4262,
+      "step": 1711
+    },
+    {
+      "epoch": 3.4035785288270377,
+      "grad_norm": 0.2942487214647327,
+      "learning_rate": 2.2384233937694626e-05,
+      "loss": 0.4138,
+      "step": 1712
+    },
+    {
+      "epoch": 3.405566600397614,
+      "grad_norm": 0.1734635276860347,
+      "learning_rate": 2.2334396113633464e-05,
+      "loss": 0.4212,
+      "step": 1713
+    },
+    {
+      "epoch": 3.4075546719681906,
+      "grad_norm": 0.20395877825842865,
+      "learning_rate": 2.2284592335020475e-05,
+      "loss": 0.4275,
+      "step": 1714
+    },
+    {
+      "epoch": 3.4095427435387675,
+      "grad_norm": 0.18126084824037503,
+      "learning_rate": 2.223482269783834e-05,
+      "loss": 0.4182,
+      "step": 1715
+    },
+    {
+      "epoch": 3.411530815109344,
+      "grad_norm": 0.15914404766557488,
+      "learning_rate": 2.2185087298003968e-05,
+      "loss": 0.4118,
+      "step": 1716
+    },
+    {
+      "epoch": 3.4135188866799204,
+      "grad_norm": 0.1590304746258962,
+      "learning_rate": 2.213538623136824e-05,
+      "loss": 0.4172,
+      "step": 1717
+    },
+    {
+      "epoch": 3.415506958250497,
+      "grad_norm": 0.17825298578092894,
+      "learning_rate": 2.2085719593715886e-05,
+      "loss": 0.4132,
+      "step": 1718
+    },
+    {
+      "epoch": 3.4174950298210733,
+      "grad_norm": 0.16302076276162225,
+      "learning_rate": 2.203608748076528e-05,
+      "loss": 0.4169,
+      "step": 1719
+    },
+    {
+      "epoch": 3.4194831013916502,
+      "grad_norm": 0.17318949422352425,
+      "learning_rate": 2.198648998816827e-05,
+      "loss": 0.4187,
+      "step": 1720
+    },
+    {
+      "epoch": 3.4214711729622267,
+      "grad_norm": 0.15012555647563922,
+      "learning_rate": 2.193692721150998e-05,
+      "loss": 0.4226,
+      "step": 1721
+    },
+    {
+      "epoch": 3.423459244532803,
+      "grad_norm": 0.18815489284348785,
+      "learning_rate": 2.1887399246308623e-05,
+      "loss": 0.413,
+      "step": 1722
+    },
+    {
+      "epoch": 3.4254473161033796,
+      "grad_norm": 0.13978612071322793,
+      "learning_rate": 2.1837906188015303e-05,
+      "loss": 0.4227,
+      "step": 1723
+    },
+    {
+      "epoch": 3.427435387673956,
+      "grad_norm": 0.19168527604634417,
+      "learning_rate": 2.1788448132013874e-05,
+      "loss": 0.4249,
+      "step": 1724
+    },
+    {
+      "epoch": 3.429423459244533,
+      "grad_norm": 0.1535627672588957,
+      "learning_rate": 2.173902517362076e-05,
+      "loss": 0.4247,
+      "step": 1725
+    },
+    {
+      "epoch": 3.4314115308151094,
+      "grad_norm": 0.13874613052414875,
+      "learning_rate": 2.1689637408084703e-05,
+      "loss": 0.418,
+      "step": 1726
+    },
+    {
+      "epoch": 3.433399602385686,
+      "grad_norm": 0.13587604761472546,
+      "learning_rate": 2.1640284930586635e-05,
+      "loss": 0.417,
+      "step": 1727
+    },
+    {
+      "epoch": 3.4353876739562623,
+      "grad_norm": 0.16274376236534926,
+      "learning_rate": 2.1590967836239464e-05,
+      "loss": 0.4091,
+      "step": 1728
+    },
+    {
+      "epoch": 3.4373757455268388,
+      "grad_norm": 0.1389554882025859,
+      "learning_rate": 2.1541686220087938e-05,
+      "loss": 0.425,
+      "step": 1729
+    },
+    {
+      "epoch": 3.4393638170974157,
+      "grad_norm": 0.15567485102640824,
+      "learning_rate": 2.1492440177108423e-05,
+      "loss": 0.4211,
+      "step": 1730
+    },
+    {
+      "epoch": 3.441351888667992,
+      "grad_norm": 0.14753218577823196,
+      "learning_rate": 2.144322980220869e-05,
+      "loss": 0.4223,
+      "step": 1731
+    },
+    {
+      "epoch": 3.4433399602385686,
+      "grad_norm": 0.1458705887737164,
+      "learning_rate": 2.1394055190227814e-05,
+      "loss": 0.4319,
+      "step": 1732
+    },
+    {
+      "epoch": 3.445328031809145,
+      "grad_norm": 0.1397058371801398,
+      "learning_rate": 2.1344916435935915e-05,
+      "loss": 0.4113,
+      "step": 1733
+    },
+    {
+      "epoch": 3.4473161033797215,
+      "grad_norm": 0.14684655321915865,
+      "learning_rate": 2.129581363403406e-05,
+      "loss": 0.4237,
+      "step": 1734
+    },
+    {
+      "epoch": 3.4493041749502984,
+      "grad_norm": 0.17704552353843994,
+      "learning_rate": 2.124674687915398e-05,
+      "loss": 0.4226,
+      "step": 1735
+    },
+    {
+      "epoch": 3.451292246520875,
+      "grad_norm": 0.14030210704985885,
+      "learning_rate": 2.1197716265857932e-05,
+      "loss": 0.4188,
+      "step": 1736
+    },
+    {
+      "epoch": 3.4532803180914513,
+      "grad_norm": 0.17008441652493428,
+      "learning_rate": 2.1148721888638556e-05,
+      "loss": 0.4168,
+      "step": 1737
+    },
+    {
+      "epoch": 3.4552683896620278,
+      "grad_norm": 0.1836350292563087,
+      "learning_rate": 2.1099763841918626e-05,
+      "loss": 0.413,
+      "step": 1738
+    },
+    {
+      "epoch": 3.457256461232604,
+      "grad_norm": 0.4358843485324826,
+      "learning_rate": 2.105084222005093e-05,
+      "loss": 0.4398,
+      "step": 1739
+    },
+    {
+      "epoch": 3.459244532803181,
+      "grad_norm": 0.19076079127469725,
+      "learning_rate": 2.1001957117318002e-05,
+      "loss": 0.4156,
+      "step": 1740
+    },
+    {
+      "epoch": 3.4612326043737576,
+      "grad_norm": 0.15919930414992362,
+      "learning_rate": 2.095310862793206e-05,
+      "loss": 0.4207,
+      "step": 1741
+    },
+    {
+      "epoch": 3.463220675944334,
+      "grad_norm": 0.20473446973024545,
+      "learning_rate": 2.0904296846034703e-05,
+      "loss": 0.4174,
+      "step": 1742
+    },
+    {
+      "epoch": 3.4652087475149105,
+      "grad_norm": 0.15137987512270168,
+      "learning_rate": 2.085552186569686e-05,
+      "loss": 0.4117,
+      "step": 1743
+    },
+    {
+      "epoch": 3.467196819085487,
+      "grad_norm": 0.18074067785558987,
+      "learning_rate": 2.0806783780918467e-05,
+      "loss": 0.4092,
+      "step": 1744
+    },
+    {
+      "epoch": 3.469184890656064,
+      "grad_norm": 0.13226943416324957,
+      "learning_rate": 2.0758082685628377e-05,
+      "loss": 0.4146,
+      "step": 1745
+    },
+    {
+      "epoch": 3.4711729622266403,
+      "grad_norm": 0.16993781310376027,
+      "learning_rate": 2.0709418673684166e-05,
+      "loss": 0.4224,
+      "step": 1746
+    },
+    {
+      "epoch": 3.4731610337972167,
+      "grad_norm": 0.15112977680745895,
+      "learning_rate": 2.066079183887193e-05,
+      "loss": 0.4038,
+      "step": 1747
+    },
+    {
+      "epoch": 3.475149105367793,
+      "grad_norm": 0.15100626586795593,
+      "learning_rate": 2.0612202274906137e-05,
+      "loss": 0.4081,
+      "step": 1748
+    },
+    {
+      "epoch": 3.4771371769383697,
+      "grad_norm": 1.9680321094660187,
+      "learning_rate": 2.0563650075429376e-05,
+      "loss": 0.4662,
+      "step": 1749
+    },
+    {
+      "epoch": 3.4791252485089466,
+      "grad_norm": 0.19286841860615947,
+      "learning_rate": 2.0515135334012283e-05,
+      "loss": 0.4157,
+      "step": 1750
+    },
+    {
+      "epoch": 3.481113320079523,
+      "grad_norm": 0.14764009814938028,
+      "learning_rate": 2.046665814415326e-05,
+      "loss": 0.4188,
+      "step": 1751
+    },
+    {
+      "epoch": 3.4831013916500995,
+      "grad_norm": 0.18319314928246067,
+      "learning_rate": 2.0418218599278388e-05,
+      "loss": 0.4331,
+      "step": 1752
+    },
+    {
+      "epoch": 3.485089463220676,
+      "grad_norm": 0.14820529438687846,
+      "learning_rate": 2.036981679274117e-05,
+      "loss": 0.4261,
+      "step": 1753
+    },
+    {
+      "epoch": 3.4870775347912524,
+      "grad_norm": 0.1497754362666427,
+      "learning_rate": 2.0321452817822363e-05,
+      "loss": 0.414,
+      "step": 1754
+    },
+    {
+      "epoch": 3.4890656063618293,
+      "grad_norm": 0.15919162663516262,
+      "learning_rate": 2.027312676772984e-05,
+      "loss": 0.4297,
+      "step": 1755
+    },
+    {
+      "epoch": 3.4910536779324057,
+      "grad_norm": 0.1758929970525375,
+      "learning_rate": 2.0224838735598374e-05,
+      "loss": 0.4187,
+      "step": 1756
+    },
+    {
+      "epoch": 3.493041749502982,
+      "grad_norm": 0.1452678866846859,
+      "learning_rate": 2.0176588814489477e-05,
+      "loss": 0.4241,
+      "step": 1757
+    },
+    {
+      "epoch": 3.4950298210735586,
+      "grad_norm": 0.1514360850159552,
+      "learning_rate": 2.0128377097391193e-05,
+      "loss": 0.4169,
+      "step": 1758
+    },
+    {
+      "epoch": 3.497017892644135,
+      "grad_norm": 0.145255134633476,
+      "learning_rate": 2.0080203677217955e-05,
+      "loss": 0.4127,
+      "step": 1759
+    },
+    {
+      "epoch": 3.4990059642147116,
+      "grad_norm": 0.1548974716711143,
+      "learning_rate": 2.0032068646810378e-05,
+      "loss": 0.4199,
+      "step": 1760
+    },
+    {
+      "epoch": 3.500994035785288,
+      "grad_norm": 0.1506452726505713,
+      "learning_rate": 1.9983972098935122e-05,
+      "loss": 0.4188,
+      "step": 1761
+    },
+    {
+      "epoch": 3.502982107355865,
+      "grad_norm": 0.14658095527849782,
+      "learning_rate": 1.9935914126284655e-05,
+      "loss": 0.414,
+      "step": 1762
+    },
+    {
+      "epoch": 3.5049701789264414,
+      "grad_norm": 0.14632438106244663,
+      "learning_rate": 1.9887894821477106e-05,
+      "loss": 0.4258,
+      "step": 1763
+    },
+    {
+      "epoch": 3.506958250497018,
+      "grad_norm": 0.14429915729010442,
+      "learning_rate": 1.983991427705608e-05,
+      "loss": 0.4144,
+      "step": 1764
+    },
+    {
+      "epoch": 3.5089463220675943,
+      "grad_norm": 0.1429848169311897,
+      "learning_rate": 1.979197258549049e-05,
+      "loss": 0.4127,
+      "step": 1765
+    },
+    {
+      "epoch": 3.5109343936381707,
+      "grad_norm": 0.1676451471694739,
+      "learning_rate": 1.9744069839174384e-05,
+      "loss": 0.4153,
+      "step": 1766
+    },
+    {
+      "epoch": 3.5129224652087476,
+      "grad_norm": 0.14066248382152913,
+      "learning_rate": 1.969620613042671e-05,
+      "loss": 0.4185,
+      "step": 1767
+    },
+    {
+      "epoch": 3.514910536779324,
+      "grad_norm": 0.16764366472616893,
+      "learning_rate": 1.9648381551491224e-05,
+      "loss": 0.4141,
+      "step": 1768
+    },
+    {
+      "epoch": 3.5168986083499005,
+      "grad_norm": 0.14003669923848128,
+      "learning_rate": 1.960059619453624e-05,
+      "loss": 0.4213,
+      "step": 1769
+    },
+    {
+      "epoch": 3.518886679920477,
+      "grad_norm": 0.14366167229047724,
+      "learning_rate": 1.955285015165453e-05,
+      "loss": 0.4206,
+      "step": 1770
+    },
+    {
+      "epoch": 3.5208747514910534,
+      "grad_norm": 0.15242720064957377,
+      "learning_rate": 1.9505143514863056e-05,
+      "loss": 0.4009,
+      "step": 1771
+    },
+    {
+      "epoch": 3.5228628230616303,
+      "grad_norm": 0.1785700156717405,
+      "learning_rate": 1.945747637610285e-05,
+      "loss": 0.4224,
+      "step": 1772
+    },
+    {
+      "epoch": 3.524850894632207,
+      "grad_norm": 0.1529136898284363,
+      "learning_rate": 1.9409848827238814e-05,
+      "loss": 0.4189,
+      "step": 1773
+    },
+    {
+      "epoch": 3.5268389662027833,
+      "grad_norm": 0.15141172921996104,
+      "learning_rate": 1.936226096005957e-05,
+      "loss": 0.4175,
+      "step": 1774
+    },
+    {
+      "epoch": 3.5288270377733597,
+      "grad_norm": 0.1440969504372814,
+      "learning_rate": 1.9314712866277264e-05,
+      "loss": 0.4084,
+      "step": 1775
+    },
+    {
+      "epoch": 3.530815109343936,
+      "grad_norm": 0.14316810507788563,
+      "learning_rate": 1.9267204637527345e-05,
+      "loss": 0.4218,
+      "step": 1776
+    },
+    {
+      "epoch": 3.532803180914513,
+      "grad_norm": 0.14564934266921534,
+      "learning_rate": 1.921973636536848e-05,
+      "loss": 0.4051,
+      "step": 1777
+    },
+    {
+      "epoch": 3.5347912524850895,
+      "grad_norm": 0.25175630532587917,
+      "learning_rate": 1.9172308141282336e-05,
+      "loss": 0.4207,
+      "step": 1778
+    },
+    {
+      "epoch": 3.536779324055666,
+      "grad_norm": 0.14020457933643266,
+      "learning_rate": 1.9124920056673374e-05,
+      "loss": 0.4193,
+      "step": 1779
+    },
+    {
+      "epoch": 3.5387673956262424,
+      "grad_norm": 0.15928874940395493,
+      "learning_rate": 1.907757220286871e-05,
+      "loss": 0.4132,
+      "step": 1780
+    },
+    {
+      "epoch": 3.540755467196819,
+      "grad_norm": 0.15522223883560438,
+      "learning_rate": 1.9030264671117914e-05,
+      "loss": 0.4199,
+      "step": 1781
+    },
+    {
+      "epoch": 3.542743538767396,
+      "grad_norm": 0.24239762715881097,
+      "learning_rate": 1.8982997552592863e-05,
+      "loss": 0.4183,
+      "step": 1782
+    },
+    {
+      "epoch": 3.5447316103379722,
+      "grad_norm": 0.1660873246133442,
+      "learning_rate": 1.893577093838753e-05,
+      "loss": 0.4242,
+      "step": 1783
+    },
+    {
+      "epoch": 3.5467196819085487,
+      "grad_norm": 0.16085140583809687,
+      "learning_rate": 1.8888584919517867e-05,
+      "loss": 0.4177,
+      "step": 1784
+    },
+    {
+      "epoch": 3.548707753479125,
+      "grad_norm": 0.16069903654575152,
+      "learning_rate": 1.8841439586921515e-05,
+      "loss": 0.4182,
+      "step": 1785
+    },
+    {
+      "epoch": 3.5506958250497016,
+      "grad_norm": 0.15442226396451475,
+      "learning_rate": 1.8794335031457763e-05,
+      "loss": 0.4168,
+      "step": 1786
+    },
+    {
+      "epoch": 3.5526838966202785,
+      "grad_norm": 0.15050396138925257,
+      "learning_rate": 1.874727134390733e-05,
+      "loss": 0.4196,
+      "step": 1787
+    },
+    {
+      "epoch": 3.554671968190855,
+      "grad_norm": 0.15316322830236687,
+      "learning_rate": 1.8700248614972122e-05,
+      "loss": 0.4261,
+      "step": 1788
+    },
+    {
+      "epoch": 3.5566600397614314,
+      "grad_norm": 0.15262385894358507,
+      "learning_rate": 1.865326693527514e-05,
+      "loss": 0.4236,
+      "step": 1789
+    },
+    {
+      "epoch": 3.558648111332008,
+      "grad_norm": 0.13953426409619946,
+      "learning_rate": 1.8606326395360254e-05,
+      "loss": 0.4278,
+      "step": 1790
+    },
+    {
+      "epoch": 3.5606361829025843,
+      "grad_norm": 0.1453249229248661,
+      "learning_rate": 1.8559427085692068e-05,
+      "loss": 0.4151,
+      "step": 1791
+    },
+    {
+      "epoch": 3.5626242544731612,
+      "grad_norm": 0.13768057292816951,
+      "learning_rate": 1.8512569096655717e-05,
+      "loss": 0.4112,
+      "step": 1792
+    },
+    {
+      "epoch": 3.5646123260437377,
+      "grad_norm": 0.15733429051521475,
+      "learning_rate": 1.8465752518556715e-05,
+      "loss": 0.4106,
+      "step": 1793
+    },
+    {
+      "epoch": 3.566600397614314,
+      "grad_norm": 0.15307708019233576,
+      "learning_rate": 1.8418977441620737e-05,
+      "loss": 0.4197,
+      "step": 1794
+    },
+    {
+      "epoch": 3.5685884691848906,
+      "grad_norm": 0.14136017029964,
+      "learning_rate": 1.8372243955993484e-05,
+      "loss": 0.4229,
+      "step": 1795
+    },
+    {
+      "epoch": 3.570576540755467,
+      "grad_norm": 0.14308762935160746,
+      "learning_rate": 1.8325552151740553e-05,
+      "loss": 0.4156,
+      "step": 1796
+    },
+    {
+      "epoch": 3.572564612326044,
+      "grad_norm": 0.14090812672312653,
+      "learning_rate": 1.827890211884715e-05,
+      "loss": 0.4139,
+      "step": 1797
+    },
+    {
+      "epoch": 3.5745526838966204,
+      "grad_norm": 0.1450916042416356,
+      "learning_rate": 1.8232293947218004e-05,
+      "loss": 0.4116,
+      "step": 1798
+    },
+    {
+      "epoch": 3.576540755467197,
+      "grad_norm": 0.14762435550768865,
+      "learning_rate": 1.8185727726677167e-05,
+      "loss": 0.4312,
+      "step": 1799
+    },
+    {
+      "epoch": 3.5785288270377733,
+      "grad_norm": 0.15042990698055161,
+      "learning_rate": 1.8139203546967835e-05,
+      "loss": 0.421,
+      "step": 1800
+    },
+    {
+      "epoch": 3.5805168986083498,
+      "grad_norm": 0.14061082501100122,
+      "learning_rate": 1.8092721497752194e-05,
+      "loss": 0.4236,
+      "step": 1801
+    },
+    {
+      "epoch": 3.5825049701789267,
+      "grad_norm": 0.14111997641843832,
+      "learning_rate": 1.8046281668611242e-05,
+      "loss": 0.4196,
+      "step": 1802
+    },
+    {
+      "epoch": 3.584493041749503,
+      "grad_norm": 0.16047052141744386,
+      "learning_rate": 1.799988414904457e-05,
+      "loss": 0.413,
+      "step": 1803
+    },
+    {
+      "epoch": 3.5864811133200796,
+      "grad_norm": 0.14324152977090976,
+      "learning_rate": 1.795352902847025e-05,
+      "loss": 0.4214,
+      "step": 1804
+    },
+    {
+      "epoch": 3.588469184890656,
+      "grad_norm": 0.1637034156499913,
+      "learning_rate": 1.790721639622468e-05,
+      "loss": 0.42,
+      "step": 1805
+    },
+    {
+      "epoch": 3.5904572564612325,
+      "grad_norm": 0.14535956431472496,
+      "learning_rate": 1.786094634156233e-05,
+      "loss": 0.4217,
+      "step": 1806
+    },
+    {
+      "epoch": 3.5924453280318094,
+      "grad_norm": 0.15532983320228866,
+      "learning_rate": 1.781471895365562e-05,
+      "loss": 0.4142,
+      "step": 1807
+    },
+    {
+      "epoch": 3.594433399602386,
+      "grad_norm": 0.1648501719516284,
+      "learning_rate": 1.7768534321594755e-05,
+      "loss": 0.422,
+      "step": 1808
+    },
+    {
+      "epoch": 3.5964214711729623,
+      "grad_norm": 0.13739363913925556,
+      "learning_rate": 1.7722392534387526e-05,
+      "loss": 0.4195,
+      "step": 1809
+    },
+    {
+      "epoch": 3.5984095427435387,
+      "grad_norm": 0.1684875509378481,
+      "learning_rate": 1.7676293680959178e-05,
+      "loss": 0.4154,
+      "step": 1810
+    },
+    {
+      "epoch": 3.600397614314115,
+      "grad_norm": 0.15049792505533202,
+      "learning_rate": 1.763023785015216e-05,
+      "loss": 0.4283,
+      "step": 1811
+    },
+    {
+      "epoch": 3.602385685884692,
+      "grad_norm": 0.14341126106705412,
+      "learning_rate": 1.7584225130726066e-05,
+      "loss": 0.4148,
+      "step": 1812
+    },
+    {
+      "epoch": 3.604373757455268,
+      "grad_norm": 0.141627160897845,
+      "learning_rate": 1.7538255611357358e-05,
+      "loss": 0.4267,
+      "step": 1813
+    },
+    {
+      "epoch": 3.606361829025845,
+      "grad_norm": 0.1574393061219025,
+      "learning_rate": 1.7492329380639298e-05,
+      "loss": 0.4216,
+      "step": 1814
+    },
+    {
+      "epoch": 3.6083499005964215,
+      "grad_norm": 0.13802470610512332,
+      "learning_rate": 1.7446446527081673e-05,
+      "loss": 0.4178,
+      "step": 1815
+    },
+    {
+      "epoch": 3.610337972166998,
+      "grad_norm": 0.15359173777670673,
+      "learning_rate": 1.7400607139110687e-05,
+      "loss": 0.4098,
+      "step": 1816
+    },
+    {
+      "epoch": 3.612326043737575,
+      "grad_norm": 0.13739662485676052,
+      "learning_rate": 1.735481130506878e-05,
+      "loss": 0.4125,
+      "step": 1817
+    },
+    {
+      "epoch": 3.614314115308151,
+      "grad_norm": 0.1671537840554596,
+      "learning_rate": 1.730905911321445e-05,
+      "loss": 0.4167,
+      "step": 1818
+    },
+    {
+      "epoch": 3.6163021868787277,
+      "grad_norm": 0.15557775501383705,
+      "learning_rate": 1.7263350651722103e-05,
+      "loss": 0.4165,
+      "step": 1819
+    },
+    {
+      "epoch": 3.618290258449304,
+      "grad_norm": 0.14372860836642484,
+      "learning_rate": 1.7217686008681825e-05,
+      "loss": 0.4157,
+      "step": 1820
+    },
+    {
+      "epoch": 3.6202783300198806,
+      "grad_norm": 0.22155274518208556,
+      "learning_rate": 1.7172065272099297e-05,
+      "loss": 0.421,
+      "step": 1821
+    },
+    {
+      "epoch": 3.622266401590457,
+      "grad_norm": 0.1305430733649789,
+      "learning_rate": 1.7126488529895548e-05,
+      "loss": 0.4057,
+      "step": 1822
+    },
+    {
+      "epoch": 3.6242544731610336,
+      "grad_norm": 0.15254680452691538,
+      "learning_rate": 1.7080955869906874e-05,
+      "loss": 0.4159,
+      "step": 1823
+    },
+    {
+      "epoch": 3.6262425447316105,
+      "grad_norm": 0.14049400385188401,
+      "learning_rate": 1.703546737988457e-05,
+      "loss": 0.4142,
+      "step": 1824
+    },
+    {
+      "epoch": 3.628230616302187,
+      "grad_norm": 0.13644463312609267,
+      "learning_rate": 1.6990023147494814e-05,
+      "loss": 0.4226,
+      "step": 1825
+    },
+    {
+      "epoch": 3.6302186878727634,
+      "grad_norm": 0.17220133648201097,
+      "learning_rate": 1.694462326031849e-05,
+      "loss": 0.4138,
+      "step": 1826
+    },
+    {
+      "epoch": 3.63220675944334,
+      "grad_norm": 0.14493215532450543,
+      "learning_rate": 1.6899267805851028e-05,
+      "loss": 0.4186,
+      "step": 1827
+    },
+    {
+      "epoch": 3.6341948310139163,
+      "grad_norm": 0.20593834230752084,
+      "learning_rate": 1.685395687150223e-05,
+      "loss": 0.4155,
+      "step": 1828
+    },
+    {
+      "epoch": 3.636182902584493,
+      "grad_norm": 0.14491486894512906,
+      "learning_rate": 1.6808690544596062e-05,
+      "loss": 0.4125,
+      "step": 1829
+    },
+    {
+      "epoch": 3.6381709741550696,
+      "grad_norm": 0.1523534114676863,
+      "learning_rate": 1.6763468912370547e-05,
+      "loss": 0.4185,
+      "step": 1830
+    },
+    {
+      "epoch": 3.640159045725646,
+      "grad_norm": 0.14504280147138135,
+      "learning_rate": 1.6718292061977604e-05,
+      "loss": 0.4283,
+      "step": 1831
+    },
+    {
+      "epoch": 3.6421471172962225,
+      "grad_norm": 0.15299410652292605,
+      "learning_rate": 1.667316008048281e-05,
+      "loss": 0.4146,
+      "step": 1832
+    },
+    {
+      "epoch": 3.644135188866799,
+      "grad_norm": 0.13635691345377363,
+      "learning_rate": 1.6628073054865262e-05,
+      "loss": 0.407,
+      "step": 1833
+    },
+    {
+      "epoch": 3.646123260437376,
+      "grad_norm": 0.14474262221855225,
+      "learning_rate": 1.658303107201744e-05,
+      "loss": 0.4165,
+      "step": 1834
+    },
+    {
+      "epoch": 3.6481113320079523,
+      "grad_norm": 0.13242000912309584,
+      "learning_rate": 1.653803421874501e-05,
+      "loss": 0.4144,
+      "step": 1835
+    },
+    {
+      "epoch": 3.650099403578529,
+      "grad_norm": 0.1417432474902151,
+      "learning_rate": 1.6493082581766664e-05,
+      "loss": 0.412,
+      "step": 1836
+    },
+    {
+      "epoch": 3.6520874751491053,
+      "grad_norm": 0.13649931195067325,
+      "learning_rate": 1.6448176247713963e-05,
+      "loss": 0.4202,
+      "step": 1837
+    },
+    {
+      "epoch": 3.6540755467196817,
+      "grad_norm": 0.13140071457176428,
+      "learning_rate": 1.6403315303131117e-05,
+      "loss": 0.4071,
+      "step": 1838
+    },
+    {
+      "epoch": 3.6560636182902586,
+      "grad_norm": 0.12785529079057562,
+      "learning_rate": 1.6358499834474888e-05,
+      "loss": 0.4238,
+      "step": 1839
+    },
+    {
+      "epoch": 3.658051689860835,
+      "grad_norm": 0.1328523406546393,
+      "learning_rate": 1.6313729928114436e-05,
+      "loss": 0.4177,
+      "step": 1840
+    },
+    {
+      "epoch": 3.6600397614314115,
+      "grad_norm": 0.12763913723943507,
+      "learning_rate": 1.6269005670331056e-05,
+      "loss": 0.4191,
+      "step": 1841
+    },
+    {
+      "epoch": 3.662027833001988,
+      "grad_norm": 0.1258915652562026,
+      "learning_rate": 1.6224327147318088e-05,
+      "loss": 0.4154,
+      "step": 1842
+    },
+    {
+      "epoch": 3.6640159045725644,
+      "grad_norm": 0.1419369252184392,
+      "learning_rate": 1.6179694445180734e-05,
+      "loss": 0.4046,
+      "step": 1843
+    },
+    {
+      "epoch": 3.6660039761431413,
+      "grad_norm": 0.11679097361725023,
+      "learning_rate": 1.613510764993589e-05,
+      "loss": 0.4105,
+      "step": 1844
+    },
+    {
+      "epoch": 3.667992047713718,
+      "grad_norm": 0.12616122435207,
+      "learning_rate": 1.6090566847511963e-05,
+      "loss": 0.424,
+      "step": 1845
+    },
+    {
+      "epoch": 3.6699801192842942,
+      "grad_norm": 0.14828367320020833,
+      "learning_rate": 1.604607212374876e-05,
+      "loss": 0.4201,
+      "step": 1846
+    },
+    {
+      "epoch": 3.6719681908548707,
+      "grad_norm": 0.13104162926966564,
+      "learning_rate": 1.600162356439723e-05,
+      "loss": 0.4128,
+      "step": 1847
+    },
+    {
+      "epoch": 3.673956262425447,
+      "grad_norm": 0.1476910246192276,
+      "learning_rate": 1.5957221255119372e-05,
+      "loss": 0.4102,
+      "step": 1848
+    },
+    {
+      "epoch": 3.675944333996024,
+      "grad_norm": 0.1368512969102732,
+      "learning_rate": 1.5912865281488102e-05,
+      "loss": 0.4191,
+      "step": 1849
+    },
+    {
+      "epoch": 3.6779324055666005,
+      "grad_norm": 0.1281089423568805,
+      "learning_rate": 1.586855572898698e-05,
+      "loss": 0.4076,
+      "step": 1850
+    },
+    {
+      "epoch": 3.679920477137177,
+      "grad_norm": 0.1395949823972034,
+      "learning_rate": 1.582429268301011e-05,
+      "loss": 0.4134,
+      "step": 1851
+    },
+    {
+      "epoch": 3.6819085487077534,
+      "grad_norm": 0.13951337217782803,
+      "learning_rate": 1.5780076228861993e-05,
+      "loss": 0.4168,
+      "step": 1852
+    },
+    {
+      "epoch": 3.68389662027833,
+      "grad_norm": 0.12829014970739838,
+      "learning_rate": 1.5735906451757307e-05,
+      "loss": 0.4198,
+      "step": 1853
+    },
+    {
+      "epoch": 3.6858846918489068,
+      "grad_norm": 0.13774909426196918,
+      "learning_rate": 1.5691783436820794e-05,
+      "loss": 0.4192,
+      "step": 1854
+    },
+    {
+      "epoch": 3.6878727634194832,
+      "grad_norm": 0.28908814343759603,
+      "learning_rate": 1.564770726908708e-05,
+      "loss": 0.4079,
+      "step": 1855
+    },
+    {
+      "epoch": 3.6898608349900597,
+      "grad_norm": 0.14430553920756006,
+      "learning_rate": 1.5603678033500466e-05,
+      "loss": 0.4152,
+      "step": 1856
+    },
+    {
+      "epoch": 3.691848906560636,
+      "grad_norm": 0.16044176639548022,
+      "learning_rate": 1.5559695814914833e-05,
+      "loss": 0.4172,
+      "step": 1857
+    },
+    {
+      "epoch": 3.6938369781312126,
+      "grad_norm": 0.14847470198565352,
+      "learning_rate": 1.5515760698093485e-05,
+      "loss": 0.4052,
+      "step": 1858
+    },
+    {
+      "epoch": 3.6958250497017895,
+      "grad_norm": 0.13066672984376299,
+      "learning_rate": 1.54718727677089e-05,
+      "loss": 0.4125,
+      "step": 1859
+    },
+    {
+      "epoch": 3.697813121272366,
+      "grad_norm": 0.16208398204896493,
+      "learning_rate": 1.542803210834263e-05,
+      "loss": 0.4109,
+      "step": 1860
+    },
+    {
+      "epoch": 3.6998011928429424,
+      "grad_norm": 0.14308392969629824,
+      "learning_rate": 1.5384238804485132e-05,
+      "loss": 0.4171,
+      "step": 1861
+    },
+    {
+      "epoch": 3.701789264413519,
+      "grad_norm": 0.13472552616176814,
+      "learning_rate": 1.5340492940535594e-05,
+      "loss": 0.4188,
+      "step": 1862
+    },
+    {
+      "epoch": 3.7037773359840953,
+      "grad_norm": 0.16245696982606775,
+      "learning_rate": 1.529679460080178e-05,
+      "loss": 0.4159,
+      "step": 1863
+    },
+    {
+      "epoch": 3.705765407554672,
+      "grad_norm": 0.12641490699261004,
+      "learning_rate": 1.5253143869499875e-05,
+      "loss": 0.4142,
+      "step": 1864
+    },
+    {
+      "epoch": 3.7077534791252487,
+      "grad_norm": 0.1407468125478082,
+      "learning_rate": 1.5209540830754282e-05,
+      "loss": 0.4081,
+      "step": 1865
+    },
+    {
+      "epoch": 3.709741550695825,
+      "grad_norm": 0.12063476957620818,
+      "learning_rate": 1.5165985568597497e-05,
+      "loss": 0.4199,
+      "step": 1866
+    },
+    {
+      "epoch": 3.7117296222664016,
+      "grad_norm": 0.1508589666989841,
+      "learning_rate": 1.5122478166969985e-05,
+      "loss": 0.4204,
+      "step": 1867
+    },
+    {
+      "epoch": 3.713717693836978,
+      "grad_norm": 0.27787074894019453,
+      "learning_rate": 1.5079018709719928e-05,
+      "loss": 0.4181,
+      "step": 1868
+    },
+    {
+      "epoch": 3.715705765407555,
+      "grad_norm": 0.12122504335335331,
+      "learning_rate": 1.503560728060312e-05,
+      "loss": 0.4049,
+      "step": 1869
+    },
+    {
+      "epoch": 3.717693836978131,
+      "grad_norm": 0.1285881332318176,
+      "learning_rate": 1.4992243963282796e-05,
+      "loss": 0.4121,
+      "step": 1870
+    },
+    {
+      "epoch": 3.719681908548708,
+      "grad_norm": 0.13056341767813032,
+      "learning_rate": 1.4948928841329466e-05,
+      "loss": 0.4129,
+      "step": 1871
+    },
+    {
+      "epoch": 3.7216699801192843,
+      "grad_norm": 0.1176335456819128,
+      "learning_rate": 1.4905661998220761e-05,
+      "loss": 0.4229,
+      "step": 1872
+    },
+    {
+      "epoch": 3.7236580516898607,
+      "grad_norm": 0.127572659098762,
+      "learning_rate": 1.4862443517341283e-05,
+      "loss": 0.4085,
+      "step": 1873
+    },
+    {
+      "epoch": 3.7256461232604376,
+      "grad_norm": 0.12448568682751915,
+      "learning_rate": 1.481927348198238e-05,
+      "loss": 0.4159,
+      "step": 1874
+    },
+    {
+      "epoch": 3.7276341948310137,
+      "grad_norm": 0.12127432985554247,
+      "learning_rate": 1.477615197534208e-05,
+      "loss": 0.4147,
+      "step": 1875
+    },
+    {
+      "epoch": 3.7296222664015906,
+      "grad_norm": 0.13778721644970837,
+      "learning_rate": 1.473307908052489e-05,
+      "loss": 0.4244,
+      "step": 1876
+    },
+    {
+      "epoch": 3.731610337972167,
+      "grad_norm": 0.12777566872269489,
+      "learning_rate": 1.4690054880541604e-05,
+      "loss": 0.4125,
+      "step": 1877
+    },
+    {
+      "epoch": 3.7335984095427435,
+      "grad_norm": 0.12730243157884172,
+      "learning_rate": 1.4647079458309189e-05,
+      "loss": 0.4035,
+      "step": 1878
+    },
+    {
+      "epoch": 3.73558648111332,
+      "grad_norm": 0.12401266431341217,
+      "learning_rate": 1.460415289665059e-05,
+      "loss": 0.4158,
+      "step": 1879
+    },
+    {
+      "epoch": 3.7375745526838964,
+      "grad_norm": 0.13340393869019873,
+      "learning_rate": 1.4561275278294602e-05,
+      "loss": 0.4122,
+      "step": 1880
+    },
+    {
+      "epoch": 3.7395626242544733,
+      "grad_norm": 0.12525013952678124,
+      "learning_rate": 1.4518446685875689e-05,
+      "loss": 0.4173,
+      "step": 1881
+    },
+    {
+      "epoch": 3.7415506958250497,
+      "grad_norm": 0.12718806203575092,
+      "learning_rate": 1.4475667201933851e-05,
+      "loss": 0.402,
+      "step": 1882
+    },
+    {
+      "epoch": 3.743538767395626,
+      "grad_norm": 0.12373186687056051,
+      "learning_rate": 1.4432936908914381e-05,
+      "loss": 0.4142,
+      "step": 1883
+    },
+    {
+      "epoch": 3.7455268389662026,
+      "grad_norm": 0.12193879482225009,
+      "learning_rate": 1.4390255889167856e-05,
+      "loss": 0.4101,
+      "step": 1884
+    },
+    {
+      "epoch": 3.747514910536779,
+      "grad_norm": 0.13219048436791064,
+      "learning_rate": 1.4347624224949845e-05,
+      "loss": 0.4219,
+      "step": 1885
+    },
+    {
+      "epoch": 3.749502982107356,
+      "grad_norm": 0.1215258991560494,
+      "learning_rate": 1.4305041998420808e-05,
+      "loss": 0.4133,
+      "step": 1886
+    },
+    {
+      "epoch": 3.7514910536779325,
+      "grad_norm": 0.13401458332037963,
+      "learning_rate": 1.4262509291645921e-05,
+      "loss": 0.413,
+      "step": 1887
+    },
+    {
+      "epoch": 3.753479125248509,
+      "grad_norm": 0.12356450775309526,
+      "learning_rate": 1.4220026186594932e-05,
+      "loss": 0.418,
+      "step": 1888
+    },
+    {
+      "epoch": 3.7554671968190854,
+      "grad_norm": 0.12577960300004168,
+      "learning_rate": 1.417759276514199e-05,
+      "loss": 0.4025,
+      "step": 1889
+    },
+    {
+      "epoch": 3.757455268389662,
+      "grad_norm": 0.1317678943066002,
+      "learning_rate": 1.4135209109065517e-05,
+      "loss": 0.409,
+      "step": 1890
+    },
+    {
+      "epoch": 3.7594433399602387,
+      "grad_norm": 0.11834469035412429,
+      "learning_rate": 1.4092875300047974e-05,
+      "loss": 0.4055,
+      "step": 1891
+    },
+    {
+      "epoch": 3.761431411530815,
+      "grad_norm": 0.12895864349124414,
+      "learning_rate": 1.4050591419675792e-05,
+      "loss": 0.4108,
+      "step": 1892
+    },
+    {
+      "epoch": 3.7634194831013916,
+      "grad_norm": 0.13329492236504692,
+      "learning_rate": 1.4008357549439202e-05,
+      "loss": 0.4257,
+      "step": 1893
+    },
+    {
+      "epoch": 3.765407554671968,
+      "grad_norm": 0.129072823887272,
+      "learning_rate": 1.3966173770732007e-05,
+      "loss": 0.4209,
+      "step": 1894
+    },
+    {
+      "epoch": 3.7673956262425445,
+      "grad_norm": 0.12921413700028228,
+      "learning_rate": 1.3924040164851497e-05,
+      "loss": 0.4117,
+      "step": 1895
+    },
+    {
+      "epoch": 3.7693836978131214,
+      "grad_norm": 0.13042817761575895,
+      "learning_rate": 1.3881956812998269e-05,
+      "loss": 0.4198,
+      "step": 1896
+    },
+    {
+      "epoch": 3.771371769383698,
+      "grad_norm": 0.12792815199593058,
+      "learning_rate": 1.3839923796276055e-05,
+      "loss": 0.413,
+      "step": 1897
+    },
+    {
+      "epoch": 3.7733598409542743,
+      "grad_norm": 0.11728201250249243,
+      "learning_rate": 1.37979411956916e-05,
+      "loss": 0.4155,
+      "step": 1898
+    },
+    {
+      "epoch": 3.775347912524851,
+      "grad_norm": 0.12476125551103852,
+      "learning_rate": 1.3756009092154482e-05,
+      "loss": 0.412,
+      "step": 1899
+    },
+    {
+      "epoch": 3.7773359840954273,
+      "grad_norm": 0.21481338746143638,
+      "learning_rate": 1.3714127566476933e-05,
+      "loss": 0.4131,
+      "step": 1900
+    },
+    {
+      "epoch": 3.779324055666004,
+      "grad_norm": 0.1470134918262448,
+      "learning_rate": 1.3672296699373729e-05,
+      "loss": 0.4161,
+      "step": 1901
+    },
+    {
+      "epoch": 3.7813121272365806,
+      "grad_norm": 0.12788644729106496,
+      "learning_rate": 1.3630516571462047e-05,
+      "loss": 0.4207,
+      "step": 1902
+    },
+    {
+      "epoch": 3.783300198807157,
+      "grad_norm": 0.1301078309175956,
+      "learning_rate": 1.3588787263261235e-05,
+      "loss": 0.4132,
+      "step": 1903
+    },
+    {
+      "epoch": 3.7852882703777335,
+      "grad_norm": 0.12211189542063601,
+      "learning_rate": 1.3547108855192717e-05,
+      "loss": 0.4143,
+      "step": 1904
+    },
+    {
+      "epoch": 3.78727634194831,
+      "grad_norm": 0.11690508523129137,
+      "learning_rate": 1.3505481427579822e-05,
+      "loss": 0.4108,
+      "step": 1905
+    },
+    {
+      "epoch": 3.789264413518887,
+      "grad_norm": 0.11402347692962254,
+      "learning_rate": 1.3463905060647627e-05,
+      "loss": 0.4115,
+      "step": 1906
+    },
+    {
+      "epoch": 3.7912524850894633,
+      "grad_norm": 0.16898178577525383,
+      "learning_rate": 1.3422379834522795e-05,
+      "loss": 0.4155,
+      "step": 1907
+    },
+    {
+      "epoch": 3.79324055666004,
+      "grad_norm": 0.12007712230669085,
+      "learning_rate": 1.3380905829233464e-05,
+      "loss": 0.4155,
+      "step": 1908
+    },
+    {
+      "epoch": 3.7952286282306162,
+      "grad_norm": 0.12168887034461463,
+      "learning_rate": 1.3339483124709003e-05,
+      "loss": 0.4164,
+      "step": 1909
+    },
+    {
+      "epoch": 3.7972166998011927,
+      "grad_norm": 0.1304638038783227,
+      "learning_rate": 1.3298111800779938e-05,
+      "loss": 0.4141,
+      "step": 1910
+    },
+    {
+      "epoch": 3.7992047713717696,
+      "grad_norm": 0.12286293611002731,
+      "learning_rate": 1.325679193717781e-05,
+      "loss": 0.4266,
+      "step": 1911
+    },
+    {
+      "epoch": 3.801192842942346,
+      "grad_norm": 0.11873178024386737,
+      "learning_rate": 1.321552361353494e-05,
+      "loss": 0.4094,
+      "step": 1912
+    },
+    {
+      "epoch": 3.8031809145129225,
+      "grad_norm": 0.11875300394160675,
+      "learning_rate": 1.3174306909384327e-05,
+      "loss": 0.405,
+      "step": 1913
+    },
+    {
+      "epoch": 3.805168986083499,
+      "grad_norm": 0.1305765918885014,
+      "learning_rate": 1.3133141904159504e-05,
+      "loss": 0.4167,
+      "step": 1914
+    },
+    {
+      "epoch": 3.8071570576540754,
+      "grad_norm": 0.1155018724842256,
+      "learning_rate": 1.3092028677194355e-05,
+      "loss": 0.4017,
+      "step": 1915
+    },
+    {
+      "epoch": 3.8091451292246523,
+      "grad_norm": 0.12128244400314282,
+      "learning_rate": 1.3050967307722981e-05,
+      "loss": 0.4207,
+      "step": 1916
+    },
+    {
+      "epoch": 3.8111332007952288,
+      "grad_norm": 0.126233600797627,
+      "learning_rate": 1.3009957874879557e-05,
+      "loss": 0.4243,
+      "step": 1917
+    },
+    {
+      "epoch": 3.8131212723658052,
+      "grad_norm": 0.11307120413410887,
+      "learning_rate": 1.2969000457698125e-05,
+      "loss": 0.41,
+      "step": 1918
+    },
+    {
+      "epoch": 3.8151093439363817,
+      "grad_norm": 0.13789982064344947,
+      "learning_rate": 1.2928095135112501e-05,
+      "loss": 0.4197,
+      "step": 1919
+    },
+    {
+      "epoch": 3.817097415506958,
+      "grad_norm": 0.12235858354606367,
+      "learning_rate": 1.288724198595614e-05,
+      "loss": 0.4176,
+      "step": 1920
+    },
+    {
+      "epoch": 3.819085487077535,
+      "grad_norm": 0.12135903666566722,
+      "learning_rate": 1.28464410889619e-05,
+      "loss": 0.4165,
+      "step": 1921
+    },
+    {
+      "epoch": 3.8210735586481115,
+      "grad_norm": 0.12771649763190396,
+      "learning_rate": 1.2805692522761955e-05,
+      "loss": 0.4116,
+      "step": 1922
+    },
+    {
+      "epoch": 3.823061630218688,
+      "grad_norm": 0.13045665895650455,
+      "learning_rate": 1.2764996365887624e-05,
+      "loss": 0.4207,
+      "step": 1923
+    },
+    {
+      "epoch": 3.8250497017892644,
+      "grad_norm": 0.12744409710040272,
+      "learning_rate": 1.2724352696769233e-05,
+      "loss": 0.4114,
+      "step": 1924
+    },
+    {
+      "epoch": 3.827037773359841,
+      "grad_norm": 0.13256940553885835,
+      "learning_rate": 1.268376159373593e-05,
+      "loss": 0.4115,
+      "step": 1925
+    },
+    {
+      "epoch": 3.8290258449304178,
+      "grad_norm": 0.1196732645649503,
+      "learning_rate": 1.264322313501559e-05,
+      "loss": 0.4182,
+      "step": 1926
+    },
+    {
+      "epoch": 3.8310139165009938,
+      "grad_norm": 0.12898140736842106,
+      "learning_rate": 1.2602737398734583e-05,
+      "loss": 0.4217,
+      "step": 1927
+    },
+    {
+      "epoch": 3.8330019880715707,
+      "grad_norm": 0.12259254324513066,
+      "learning_rate": 1.2562304462917698e-05,
+      "loss": 0.41,
+      "step": 1928
+    },
+    {
+      "epoch": 3.834990059642147,
+      "grad_norm": 0.14314796016876316,
+      "learning_rate": 1.2521924405487998e-05,
+      "loss": 0.4175,
+      "step": 1929
+    },
+    {
+      "epoch": 3.8369781312127236,
+      "grad_norm": 0.11735250306416119,
+      "learning_rate": 1.2481597304266585e-05,
+      "loss": 0.4116,
+      "step": 1930
+    },
+    {
+      "epoch": 3.8389662027833005,
+      "grad_norm": 0.1365058855329985,
+      "learning_rate": 1.2441323236972536e-05,
+      "loss": 0.4057,
+      "step": 1931
+    },
+    {
+      "epoch": 3.8409542743538765,
+      "grad_norm": 0.13339748891673117,
+      "learning_rate": 1.2401102281222705e-05,
+      "loss": 0.4229,
+      "step": 1932
+    },
+    {
+      "epoch": 3.8429423459244534,
+      "grad_norm": 0.11594781861973515,
+      "learning_rate": 1.2360934514531589e-05,
+      "loss": 0.4095,
+      "step": 1933
+    },
+    {
+      "epoch": 3.84493041749503,
+      "grad_norm": 0.13220912477680097,
+      "learning_rate": 1.2320820014311186e-05,
+      "loss": 0.4115,
+      "step": 1934
+    },
+    {
+      "epoch": 3.8469184890656063,
+      "grad_norm": 0.1313681631522006,
+      "learning_rate": 1.2280758857870842e-05,
+      "loss": 0.4192,
+      "step": 1935
+    },
+    {
+      "epoch": 3.8489065606361827,
+      "grad_norm": 0.11599884335856218,
+      "learning_rate": 1.2240751122417058e-05,
+      "loss": 0.4138,
+      "step": 1936
+    },
+    {
+      "epoch": 3.850894632206759,
+      "grad_norm": 0.12653486094497657,
+      "learning_rate": 1.2200796885053445e-05,
+      "loss": 0.4017,
+      "step": 1937
+    },
+    {
+      "epoch": 3.852882703777336,
+      "grad_norm": 0.1130518233332399,
+      "learning_rate": 1.2160896222780471e-05,
+      "loss": 0.4145,
+      "step": 1938
+    },
+    {
+      "epoch": 3.8548707753479126,
+      "grad_norm": 0.13866914719348072,
+      "learning_rate": 1.2121049212495359e-05,
+      "loss": 0.4099,
+      "step": 1939
+    },
+    {
+      "epoch": 3.856858846918489,
+      "grad_norm": 0.12066796601207383,
+      "learning_rate": 1.2081255930991942e-05,
+      "loss": 0.406,
+      "step": 1940
+    },
+    {
+      "epoch": 3.8588469184890655,
+      "grad_norm": 0.11330237818106631,
+      "learning_rate": 1.2041516454960492e-05,
+      "loss": 0.4187,
+      "step": 1941
+    },
+    {
+      "epoch": 3.860834990059642,
+      "grad_norm": 0.12538978699781372,
+      "learning_rate": 1.2001830860987598e-05,
+      "loss": 0.4091,
+      "step": 1942
+    },
+    {
+      "epoch": 3.862823061630219,
+      "grad_norm": 0.12467804244722842,
+      "learning_rate": 1.1962199225555996e-05,
+      "loss": 0.4107,
+      "step": 1943
+    },
+    {
+      "epoch": 3.8648111332007953,
+      "grad_norm": 0.12249180781767287,
+      "learning_rate": 1.1922621625044464e-05,
+      "loss": 0.4074,
+      "step": 1944
+    },
+    {
+      "epoch": 3.8667992047713717,
+      "grad_norm": 0.1306550240810388,
+      "learning_rate": 1.1883098135727571e-05,
+      "loss": 0.4207,
+      "step": 1945
+    },
+    {
+      "epoch": 3.868787276341948,
+      "grad_norm": 0.11933610478296443,
+      "learning_rate": 1.1843628833775687e-05,
+      "loss": 0.4115,
+      "step": 1946
+    },
+    {
+      "epoch": 3.8707753479125246,
+      "grad_norm": 0.1278251014710002,
+      "learning_rate": 1.1804213795254697e-05,
+      "loss": 0.4114,
+      "step": 1947
+    },
+    {
+      "epoch": 3.8727634194831015,
+      "grad_norm": 0.12541159535916258,
+      "learning_rate": 1.1764853096125935e-05,
+      "loss": 0.4188,
+      "step": 1948
+    },
+    {
+      "epoch": 3.874751491053678,
+      "grad_norm": 0.12727972434053444,
+      "learning_rate": 1.1725546812245994e-05,
+      "loss": 0.4101,
+      "step": 1949
+    },
+    {
+      "epoch": 3.8767395626242545,
+      "grad_norm": 0.21485568213880235,
+      "learning_rate": 1.1686295019366605e-05,
+      "loss": 0.4184,
+      "step": 1950
+    },
+    {
+      "epoch": 3.878727634194831,
+      "grad_norm": 0.12396718977956772,
+      "learning_rate": 1.1647097793134478e-05,
+      "loss": 0.4035,
+      "step": 1951
+    },
+    {
+      "epoch": 3.8807157057654074,
+      "grad_norm": 0.19187344793079394,
+      "learning_rate": 1.160795520909117e-05,
+      "loss": 0.4141,
+      "step": 1952
+    },
+    {
+      "epoch": 3.8827037773359843,
+      "grad_norm": 0.13972243163789239,
+      "learning_rate": 1.1568867342672939e-05,
+      "loss": 0.418,
+      "step": 1953
+    },
+    {
+      "epoch": 3.8846918489065607,
+      "grad_norm": 0.12246725141298273,
+      "learning_rate": 1.152983426921053e-05,
+      "loss": 0.4087,
+      "step": 1954
+    },
+    {
+      "epoch": 3.886679920477137,
+      "grad_norm": 0.1441970347739633,
+      "learning_rate": 1.149085606392919e-05,
+      "loss": 0.4107,
+      "step": 1955
+    },
+    {
+      "epoch": 3.8886679920477136,
+      "grad_norm": 0.11793533898722837,
+      "learning_rate": 1.1451932801948345e-05,
+      "loss": 0.4094,
+      "step": 1956
+    },
+    {
+      "epoch": 3.89065606361829,
+      "grad_norm": 0.11984657088654536,
+      "learning_rate": 1.1413064558281564e-05,
+      "loss": 0.4022,
+      "step": 1957
+    },
+    {
+      "epoch": 3.892644135188867,
+      "grad_norm": 0.13747510019212833,
+      "learning_rate": 1.1374251407836372e-05,
+      "loss": 0.4165,
+      "step": 1958
+    },
+    {
+      "epoch": 3.8946322067594434,
+      "grad_norm": 0.12283933628622347,
+      "learning_rate": 1.1335493425414135e-05,
+      "loss": 0.4251,
+      "step": 1959
+    },
+    {
+      "epoch": 3.89662027833002,
+      "grad_norm": 0.12767675489119157,
+      "learning_rate": 1.1296790685709875e-05,
+      "loss": 0.4068,
+      "step": 1960
+    },
+    {
+      "epoch": 3.8986083499005963,
+      "grad_norm": 0.11737134954171792,
+      "learning_rate": 1.1258143263312183e-05,
+      "loss": 0.414,
+      "step": 1961
+    },
+    {
+      "epoch": 3.900596421471173,
+      "grad_norm": 0.12017326193658819,
+      "learning_rate": 1.1219551232702993e-05,
+      "loss": 0.4087,
+      "step": 1962
+    },
+    {
+      "epoch": 3.9025844930417497,
+      "grad_norm": 0.1181570800422918,
+      "learning_rate": 1.1181014668257508e-05,
+      "loss": 0.4136,
+      "step": 1963
+    },
+    {
+      "epoch": 3.904572564612326,
+      "grad_norm": 0.11786433898774867,
+      "learning_rate": 1.1142533644244074e-05,
+      "loss": 0.4163,
+      "step": 1964
+    },
+    {
+      "epoch": 3.9065606361829026,
+      "grad_norm": 0.11951359580180812,
+      "learning_rate": 1.1104108234823946e-05,
+      "loss": 0.4167,
+      "step": 1965
+    },
+    {
+      "epoch": 3.908548707753479,
+      "grad_norm": 0.11430020006073105,
+      "learning_rate": 1.1065738514051225e-05,
+      "loss": 0.4108,
+      "step": 1966
+    },
+    {
+      "epoch": 3.9105367793240555,
+      "grad_norm": 0.12399755011503341,
+      "learning_rate": 1.1027424555872676e-05,
+      "loss": 0.4152,
+      "step": 1967
+    },
+    {
+      "epoch": 3.9125248508946324,
+      "grad_norm": 0.10379447658508856,
+      "learning_rate": 1.0989166434127604e-05,
+      "loss": 0.4191,
+      "step": 1968
+    },
+    {
+      "epoch": 3.914512922465209,
+      "grad_norm": 0.11841545442440568,
+      "learning_rate": 1.0950964222547706e-05,
+      "loss": 0.4074,
+      "step": 1969
+    },
+    {
+      "epoch": 3.9165009940357853,
+      "grad_norm": 0.12174811592361247,
+      "learning_rate": 1.0912817994756928e-05,
+      "loss": 0.4071,
+      "step": 1970
+    },
+    {
+      "epoch": 3.918489065606362,
+      "grad_norm": 0.11738800063725485,
+      "learning_rate": 1.0874727824271307e-05,
+      "loss": 0.4026,
+      "step": 1971
+    },
+    {
+      "epoch": 3.9204771371769382,
+      "grad_norm": 0.12683667073776073,
+      "learning_rate": 1.083669378449885e-05,
+      "loss": 0.4149,
+      "step": 1972
+    },
+    {
+      "epoch": 3.922465208747515,
+      "grad_norm": 0.23578844123586262,
+      "learning_rate": 1.0798715948739424e-05,
+      "loss": 0.4243,
+      "step": 1973
+    },
+    {
+      "epoch": 3.9244532803180916,
+      "grad_norm": 0.12470565801211261,
+      "learning_rate": 1.0760794390184542e-05,
+      "loss": 0.4084,
+      "step": 1974
+    },
+    {
+      "epoch": 3.926441351888668,
+      "grad_norm": 0.14135849783994955,
+      "learning_rate": 1.0722929181917264e-05,
+      "loss": 0.404,
+      "step": 1975
+    },
+    {
+      "epoch": 3.9284294234592445,
+      "grad_norm": 0.12266354579516658,
+      "learning_rate": 1.0685120396912057e-05,
+      "loss": 0.4111,
+      "step": 1976
+    },
+    {
+      "epoch": 3.930417495029821,
+      "grad_norm": 0.13413047404837405,
+      "learning_rate": 1.0647368108034644e-05,
+      "loss": 0.4123,
+      "step": 1977
+    },
+    {
+      "epoch": 3.932405566600398,
+      "grad_norm": 0.1253547852041438,
+      "learning_rate": 1.0609672388041864e-05,
+      "loss": 0.4062,
+      "step": 1978
+    },
+    {
+      "epoch": 3.9343936381709743,
+      "grad_norm": 0.11410535557037213,
+      "learning_rate": 1.0572033309581569e-05,
+      "loss": 0.4091,
+      "step": 1979
+    },
+    {
+      "epoch": 3.9363817097415508,
+      "grad_norm": 0.14930707729055762,
+      "learning_rate": 1.0534450945192382e-05,
+      "loss": 0.4194,
+      "step": 1980
+    },
+    {
+      "epoch": 3.9383697813121272,
+      "grad_norm": 0.11785277834457099,
+      "learning_rate": 1.0496925367303677e-05,
+      "loss": 0.4083,
+      "step": 1981
+    },
+    {
+      "epoch": 3.9403578528827037,
+      "grad_norm": 0.11621703078477161,
+      "learning_rate": 1.0459456648235395e-05,
+      "loss": 0.4265,
+      "step": 1982
+    },
+    {
+      "epoch": 3.9423459244532806,
+      "grad_norm": 0.12441075712798819,
+      "learning_rate": 1.0422044860197867e-05,
+      "loss": 0.4236,
+      "step": 1983
+    },
+    {
+      "epoch": 3.9443339960238566,
+      "grad_norm": 0.11732262831105307,
+      "learning_rate": 1.0384690075291725e-05,
+      "loss": 0.413,
+      "step": 1984
+    },
+    {
+      "epoch": 3.9463220675944335,
+      "grad_norm": 0.11971044861609616,
+      "learning_rate": 1.0347392365507734e-05,
+      "loss": 0.4064,
+      "step": 1985
+    },
+    {
+      "epoch": 3.94831013916501,
+      "grad_norm": 0.12454131199916985,
+      "learning_rate": 1.0310151802726667e-05,
+      "loss": 0.4024,
+      "step": 1986
+    },
+    {
+      "epoch": 3.9502982107355864,
+      "grad_norm": 0.1243323308512183,
+      "learning_rate": 1.0272968458719158e-05,
+      "loss": 0.4114,
+      "step": 1987
+    },
+    {
+      "epoch": 3.952286282306163,
+      "grad_norm": 0.10706316079552054,
+      "learning_rate": 1.0235842405145587e-05,
+      "loss": 0.4023,
+      "step": 1988
+    },
+    {
+      "epoch": 3.9542743538767393,
+      "grad_norm": 0.11034771625495471,
+      "learning_rate": 1.0198773713555878e-05,
+      "loss": 0.4021,
+      "step": 1989
+    },
+    {
+      "epoch": 3.956262425447316,
+      "grad_norm": 0.1114590814911575,
+      "learning_rate": 1.0161762455389445e-05,
+      "loss": 0.4144,
+      "step": 1990
+    },
+    {
+      "epoch": 3.9582504970178927,
+      "grad_norm": 0.11480269527129816,
+      "learning_rate": 1.0124808701975021e-05,
+      "loss": 0.4019,
+      "step": 1991
+    },
+    {
+      "epoch": 3.960238568588469,
+      "grad_norm": 0.10951780763117114,
+      "learning_rate": 1.0087912524530492e-05,
+      "loss": 0.4105,
+      "step": 1992
+    },
+    {
+      "epoch": 3.9622266401590456,
+      "grad_norm": 0.10730065026334104,
+      "learning_rate": 1.005107399416279e-05,
+      "loss": 0.4119,
+      "step": 1993
+    },
+    {
+      "epoch": 3.964214711729622,
+      "grad_norm": 0.1157812587471413,
+      "learning_rate": 1.0014293181867743e-05,
+      "loss": 0.4048,
+      "step": 1994
+    },
+    {
+      "epoch": 3.966202783300199,
+      "grad_norm": 0.10634464575554442,
+      "learning_rate": 9.977570158529958e-06,
+      "loss": 0.4126,
+      "step": 1995
+    },
+    {
+      "epoch": 3.9681908548707754,
+      "grad_norm": 0.14226769514993798,
+      "learning_rate": 9.940904994922658e-06,
+      "loss": 0.4027,
+      "step": 1996
+    },
+    {
+      "epoch": 3.970178926441352,
+      "grad_norm": 0.10876356513781965,
+      "learning_rate": 9.904297761707569e-06,
+      "loss": 0.4158,
+      "step": 1997
+    },
+    {
+      "epoch": 3.9721669980119283,
+      "grad_norm": 0.10968453251673606,
+      "learning_rate": 9.867748529434737e-06,
+      "loss": 0.4025,
+      "step": 1998
+    },
+    {
+      "epoch": 3.9741550695825048,
+      "grad_norm": 0.1106822867587708,
+      "learning_rate": 9.831257368542486e-06,
+      "loss": 0.4126,
+      "step": 1999
+    },
+    {
+      "epoch": 3.9761431411530817,
+      "grad_norm": 0.1458577023566835,
+      "learning_rate": 9.794824349357186e-06,
+      "loss": 0.416,
+      "step": 2000
+    },
+    {
+      "epoch": 3.978131212723658,
+      "grad_norm": 0.11447590869393695,
+      "learning_rate": 9.75844954209316e-06,
+      "loss": 0.4115,
+      "step": 2001
+    },
+    {
+      "epoch": 3.9801192842942346,
+      "grad_norm": 0.11780031913072884,
+      "learning_rate": 9.722133016852547e-06,
+      "loss": 0.4034,
+      "step": 2002
+    },
+    {
+      "epoch": 3.982107355864811,
+      "grad_norm": 0.11200150629333681,
+      "learning_rate": 9.685874843625158e-06,
+      "loss": 0.4071,
+      "step": 2003
+    },
+    {
+      "epoch": 3.9840954274353875,
+      "grad_norm": 0.11362356694063454,
+      "learning_rate": 9.649675092288366e-06,
+      "loss": 0.4086,
+      "step": 2004
+    },
+    {
+      "epoch": 3.9860834990059644,
+      "grad_norm": 0.11178076261068912,
+      "learning_rate": 9.613533832606925e-06,
+      "loss": 0.415,
+      "step": 2005
+    },
+    {
+      "epoch": 3.988071570576541,
+      "grad_norm": 0.1303616732073129,
+      "learning_rate": 9.5774511342329e-06,
+      "loss": 0.4162,
+      "step": 2006
+    },
+    {
+      "epoch": 3.9900596421471173,
+      "grad_norm": 0.11590806959347276,
+      "learning_rate": 9.54142706670543e-06,
+      "loss": 0.4104,
+      "step": 2007
+    },
+    {
+      "epoch": 3.9920477137176937,
+      "grad_norm": 0.11347675690643563,
+      "learning_rate": 9.505461699450747e-06,
+      "loss": 0.4154,
+      "step": 2008
+    },
+    {
+      "epoch": 3.99403578528827,
+      "grad_norm": 0.10616599308683848,
+      "learning_rate": 9.469555101781886e-06,
+      "loss": 0.4126,
+      "step": 2009
+    },
+    {
+      "epoch": 3.996023856858847,
+      "grad_norm": 0.11591746562675516,
+      "learning_rate": 9.433707342898647e-06,
+      "loss": 0.4065,
+      "step": 2010
+    },
+    {
+      "epoch": 3.9980119284294235,
+      "grad_norm": 0.1169087220375494,
+      "learning_rate": 9.397918491887443e-06,
+      "loss": 0.4106,
+      "step": 2011
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.11577369538103958,
+      "learning_rate": 9.362188617721135e-06,
+      "loss": 0.3946,
+      "step": 2012
+    },
+    {
+      "epoch": 4.001988071570577,
+      "grad_norm": 0.16242295220761954,
+      "learning_rate": 9.326517789258944e-06,
+      "loss": 0.3935,
+      "step": 2013
+    },
+    {
+      "epoch": 4.003976143141153,
+      "grad_norm": 0.12270671213229685,
+      "learning_rate": 9.290906075246288e-06,
+      "loss": 0.3796,
+      "step": 2014
+    },
+    {
+      "epoch": 4.00596421471173,
+      "grad_norm": 0.13856681126979406,
+      "learning_rate": 9.255353544314669e-06,
+      "loss": 0.3887,
+      "step": 2015
+    },
+    {
+      "epoch": 4.007952286282306,
+      "grad_norm": 0.14376438682763779,
+      "learning_rate": 9.219860264981486e-06,
+      "loss": 0.3893,
+      "step": 2016
+    },
+    {
+      "epoch": 4.009940357852883,
+      "grad_norm": 0.12730114055456757,
+      "learning_rate": 9.184426305650022e-06,
+      "loss": 0.3939,
+      "step": 2017
+    },
+    {
+      "epoch": 4.01192842942346,
+      "grad_norm": 0.13390495368426292,
+      "learning_rate": 9.149051734609183e-06,
+      "loss": 0.3882,
+      "step": 2018
+    },
+    {
+      "epoch": 4.013916500994036,
+      "grad_norm": 0.15428620649225006,
+      "learning_rate": 9.11373662003343e-06,
+      "loss": 0.4004,
+      "step": 2019
+    },
+    {
+      "epoch": 4.0159045725646125,
+      "grad_norm": 0.12561990727486522,
+      "learning_rate": 9.078481029982655e-06,
+      "loss": 0.3893,
+      "step": 2020
+    },
+    {
+      "epoch": 4.0178926441351885,
+      "grad_norm": 0.14373181736551474,
+      "learning_rate": 9.043285032402016e-06,
+      "loss": 0.3801,
+      "step": 2021
+    },
+    {
+      "epoch": 4.019880715705765,
+      "grad_norm": 0.12268500210699507,
+      "learning_rate": 9.008148695121832e-06,
+      "loss": 0.3992,
+      "step": 2022
+    },
+    {
+      "epoch": 4.021868787276342,
+      "grad_norm": 0.13904765421921883,
+      "learning_rate": 8.973072085857439e-06,
+      "loss": 0.3813,
+      "step": 2023
+    },
+    {
+      "epoch": 4.023856858846918,
+      "grad_norm": 0.11969320329871334,
+      "learning_rate": 8.938055272209083e-06,
+      "loss": 0.3939,
+      "step": 2024
+    },
+    {
+      "epoch": 4.025844930417495,
+      "grad_norm": 0.1274287401386006,
+      "learning_rate": 8.903098321661718e-06,
+      "loss": 0.3919,
+      "step": 2025
+    },
+    {
+      "epoch": 4.027833001988071,
+      "grad_norm": 0.12563635221967606,
+      "learning_rate": 8.868201301584997e-06,
+      "loss": 0.3895,
+      "step": 2026
+    },
+    {
+      "epoch": 4.029821073558648,
+      "grad_norm": 0.13763721797276127,
+      "learning_rate": 8.833364279233038e-06,
+      "loss": 0.388,
+      "step": 2027
+    },
+    {
+      "epoch": 4.031809145129225,
+      "grad_norm": 0.12122680857919836,
+      "learning_rate": 8.798587321744327e-06,
+      "loss": 0.3951,
+      "step": 2028
+    },
+    {
+      "epoch": 4.033797216699801,
+      "grad_norm": 0.12255329580322401,
+      "learning_rate": 8.763870496141606e-06,
+      "loss": 0.3765,
+      "step": 2029
+    },
+    {
+      "epoch": 4.035785288270378,
+      "grad_norm": 0.1335435726236594,
+      "learning_rate": 8.729213869331725e-06,
+      "loss": 0.3896,
+      "step": 2030
+    },
+    {
+      "epoch": 4.037773359840954,
+      "grad_norm": 0.11542654602014098,
+      "learning_rate": 8.69461750810551e-06,
+      "loss": 0.3823,
+      "step": 2031
+    },
+    {
+      "epoch": 4.039761431411531,
+      "grad_norm": 0.11640952234172514,
+      "learning_rate": 8.660081479137653e-06,
+      "loss": 0.3857,
+      "step": 2032
+    },
+    {
+      "epoch": 4.041749502982108,
+      "grad_norm": 0.11853107645718568,
+      "learning_rate": 8.625605848986577e-06,
+      "loss": 0.3783,
+      "step": 2033
+    },
+    {
+      "epoch": 4.043737574552684,
+      "grad_norm": 0.10880512742851284,
+      "learning_rate": 8.591190684094263e-06,
+      "loss": 0.3917,
+      "step": 2034
+    },
+    {
+      "epoch": 4.045725646123261,
+      "grad_norm": 0.12050622630816221,
+      "learning_rate": 8.556836050786228e-06,
+      "loss": 0.3812,
+      "step": 2035
+    },
+    {
+      "epoch": 4.047713717693837,
+      "grad_norm": 0.11073118657630984,
+      "learning_rate": 8.522542015271291e-06,
+      "loss": 0.3863,
+      "step": 2036
+    },
+    {
+      "epoch": 4.049701789264414,
+      "grad_norm": 0.11162845876453863,
+      "learning_rate": 8.488308643641483e-06,
+      "loss": 0.3854,
+      "step": 2037
+    },
+    {
+      "epoch": 4.0516898608349905,
+      "grad_norm": 0.10817652025454551,
+      "learning_rate": 8.45413600187194e-06,
+      "loss": 0.3773,
+      "step": 2038
+    },
+    {
+      "epoch": 4.0536779324055665,
+      "grad_norm": 0.1213300740692339,
+      "learning_rate": 8.420024155820745e-06,
+      "loss": 0.3873,
+      "step": 2039
+    },
+    {
+      "epoch": 4.055666003976143,
+      "grad_norm": 0.11152699715766383,
+      "learning_rate": 8.385973171228823e-06,
+      "loss": 0.3858,
+      "step": 2040
+    },
+    {
+      "epoch": 4.057654075546719,
+      "grad_norm": 0.11647996747982567,
+      "learning_rate": 8.351983113719812e-06,
+      "loss": 0.388,
+      "step": 2041
+    },
+    {
+      "epoch": 4.059642147117296,
+      "grad_norm": 0.11879327665095078,
+      "learning_rate": 8.318054048799903e-06,
+      "loss": 0.3853,
+      "step": 2042
+    },
+    {
+      "epoch": 4.061630218687872,
+      "grad_norm": 0.1070875071355291,
+      "learning_rate": 8.284186041857754e-06,
+      "loss": 0.3845,
+      "step": 2043
+    },
+    {
+      "epoch": 4.063618290258449,
+      "grad_norm": 0.10630385206510362,
+      "learning_rate": 8.250379158164383e-06,
+      "loss": 0.3938,
+      "step": 2044
+    },
+    {
+      "epoch": 4.065606361829026,
+      "grad_norm": 0.11597291042447759,
+      "learning_rate": 8.216633462872972e-06,
+      "loss": 0.3907,
+      "step": 2045
+    },
+    {
+      "epoch": 4.067594433399602,
+      "grad_norm": 0.7364567886782528,
+      "learning_rate": 8.182949021018784e-06,
+      "loss": 0.4164,
+      "step": 2046
+    },
+    {
+      "epoch": 4.069582504970179,
+      "grad_norm": 0.11061595816141373,
+      "learning_rate": 8.149325897519049e-06,
+      "loss": 0.3897,
+      "step": 2047
+    },
+    {
+      "epoch": 4.071570576540755,
+      "grad_norm": 0.11232959746711056,
+      "learning_rate": 8.115764157172812e-06,
+      "loss": 0.3988,
+      "step": 2048
+    },
+    {
+      "epoch": 4.073558648111332,
+      "grad_norm": 0.1016202934512297,
+      "learning_rate": 8.082263864660822e-06,
+      "loss": 0.3803,
+      "step": 2049
+    },
+    {
+      "epoch": 4.075546719681909,
+      "grad_norm": 0.11318725010275028,
+      "learning_rate": 8.048825084545413e-06,
+      "loss": 0.4008,
+      "step": 2050
+    },
+    {
+      "epoch": 4.077534791252485,
+      "grad_norm": 0.11326433626229489,
+      "learning_rate": 8.015447881270337e-06,
+      "loss": 0.4066,
+      "step": 2051
+    },
+    {
+      "epoch": 4.079522862823062,
+      "grad_norm": 0.10820812710830408,
+      "learning_rate": 7.982132319160722e-06,
+      "loss": 0.3814,
+      "step": 2052
+    },
+    {
+      "epoch": 4.081510934393638,
+      "grad_norm": 0.10664925736228041,
+      "learning_rate": 7.94887846242288e-06,
+      "loss": 0.3872,
+      "step": 2053
+    },
+    {
+      "epoch": 4.083499005964215,
+      "grad_norm": 0.10160843998696288,
+      "learning_rate": 7.915686375144193e-06,
+      "loss": 0.3936,
+      "step": 2054
+    },
+    {
+      "epoch": 4.085487077534792,
+      "grad_norm": 0.11288434735669675,
+      "learning_rate": 7.88255612129301e-06,
+      "loss": 0.3796,
+      "step": 2055
+    },
+    {
+      "epoch": 4.087475149105368,
+      "grad_norm": 0.1122952920063878,
+      "learning_rate": 7.849487764718514e-06,
+      "loss": 0.3817,
+      "step": 2056
+    },
+    {
+      "epoch": 4.0894632206759445,
+      "grad_norm": 0.11093903242708326,
+      "learning_rate": 7.816481369150594e-06,
+      "loss": 0.3888,
+      "step": 2057
+    },
+    {
+      "epoch": 4.0914512922465205,
+      "grad_norm": 0.10739815028668943,
+      "learning_rate": 7.783536998199732e-06,
+      "loss": 0.3961,
+      "step": 2058
+    },
+    {
+      "epoch": 4.093439363817097,
+      "grad_norm": 0.12452803775102535,
+      "learning_rate": 7.750654715356884e-06,
+      "loss": 0.3955,
+      "step": 2059
+    },
+    {
+      "epoch": 4.095427435387674,
+      "grad_norm": 0.11411657187569366,
+      "learning_rate": 7.717834583993298e-06,
+      "loss": 0.3933,
+      "step": 2060
+    },
+    {
+      "epoch": 4.09741550695825,
+      "grad_norm": 0.11299300329262432,
+      "learning_rate": 7.685076667360523e-06,
+      "loss": 0.3897,
+      "step": 2061
+    },
+    {
+      "epoch": 4.099403578528827,
+      "grad_norm": 0.11173542019812274,
+      "learning_rate": 7.652381028590152e-06,
+      "loss": 0.3873,
+      "step": 2062
+    },
+    {
+      "epoch": 4.101391650099403,
+      "grad_norm": 0.1174313343930306,
+      "learning_rate": 7.619747730693765e-06,
+      "loss": 0.3823,
+      "step": 2063
+    },
+    {
+      "epoch": 4.10337972166998,
+      "grad_norm": 0.11445309254950768,
+      "learning_rate": 7.587176836562799e-06,
+      "loss": 0.3882,
+      "step": 2064
+    },
+    {
+      "epoch": 4.105367793240557,
+      "grad_norm": 0.10898913985158198,
+      "learning_rate": 7.5546684089684395e-06,
+      "loss": 0.404,
+      "step": 2065
+    },
+    {
+      "epoch": 4.107355864811133,
+      "grad_norm": 0.12297485161873528,
+      "learning_rate": 7.522222510561463e-06,
+      "loss": 0.3901,
+      "step": 2066
+    },
+    {
+      "epoch": 4.10934393638171,
+      "grad_norm": 0.1250730138667789,
+      "learning_rate": 7.489839203872153e-06,
+      "loss": 0.3811,
+      "step": 2067
+    },
+    {
+      "epoch": 4.111332007952286,
+      "grad_norm": 0.11397541482738091,
+      "learning_rate": 7.45751855131017e-06,
+      "loss": 0.3944,
+      "step": 2068
+    },
+    {
+      "epoch": 4.113320079522863,
+      "grad_norm": 0.10709840073577899,
+      "learning_rate": 7.4252606151643894e-06,
+      "loss": 0.389,
+      "step": 2069
+    },
+    {
+      "epoch": 4.11530815109344,
+      "grad_norm": 0.11044500826907738,
+      "learning_rate": 7.393065457602877e-06,
+      "loss": 0.3918,
+      "step": 2070
+    },
+    {
+      "epoch": 4.117296222664016,
+      "grad_norm": 0.10317075798494911,
+      "learning_rate": 7.360933140672668e-06,
+      "loss": 0.3885,
+      "step": 2071
+    },
+    {
+      "epoch": 4.119284294234593,
+      "grad_norm": 0.10854131372035143,
+      "learning_rate": 7.328863726299707e-06,
+      "loss": 0.3919,
+      "step": 2072
+    },
+    {
+      "epoch": 4.121272365805169,
+      "grad_norm": 0.1030453432518548,
+      "learning_rate": 7.296857276288704e-06,
+      "loss": 0.3806,
+      "step": 2073
+    },
+    {
+      "epoch": 4.1232604373757455,
+      "grad_norm": 0.10895214824409698,
+      "learning_rate": 7.264913852323027e-06,
+      "loss": 0.3809,
+      "step": 2074
+    },
+    {
+      "epoch": 4.1252485089463224,
+      "grad_norm": 0.10521539908738563,
+      "learning_rate": 7.233033515964582e-06,
+      "loss": 0.3913,
+      "step": 2075
+    },
+    {
+      "epoch": 4.1272365805168985,
+      "grad_norm": 0.1046089740427435,
+      "learning_rate": 7.20121632865368e-06,
+      "loss": 0.3894,
+      "step": 2076
+    },
+    {
+      "epoch": 4.129224652087475,
+      "grad_norm": 0.11238242975442705,
+      "learning_rate": 7.169462351708958e-06,
+      "loss": 0.3765,
+      "step": 2077
+    },
+    {
+      "epoch": 4.131212723658051,
+      "grad_norm": 0.10931635695014968,
+      "learning_rate": 7.137771646327176e-06,
+      "loss": 0.3851,
+      "step": 2078
+    },
+    {
+      "epoch": 4.133200795228628,
+      "grad_norm": 0.1073786462890413,
+      "learning_rate": 7.106144273583227e-06,
+      "loss": 0.3819,
+      "step": 2079
+    },
+    {
+      "epoch": 4.135188866799205,
+      "grad_norm": 0.10454150184031392,
+      "learning_rate": 7.074580294429902e-06,
+      "loss": 0.3885,
+      "step": 2080
+    },
+    {
+      "epoch": 4.137176938369781,
+      "grad_norm": 0.11587471164722234,
+      "learning_rate": 7.043079769697833e-06,
+      "loss": 0.3968,
+      "step": 2081
+    },
+    {
+      "epoch": 4.139165009940358,
+      "grad_norm": 0.1049751719664736,
+      "learning_rate": 7.0116427600953565e-06,
+      "loss": 0.3887,
+      "step": 2082
+    },
+    {
+      "epoch": 4.141153081510934,
+      "grad_norm": 0.10651452982037905,
+      "learning_rate": 6.9802693262084106e-06,
+      "loss": 0.3897,
+      "step": 2083
+    },
+    {
+      "epoch": 4.143141153081511,
+      "grad_norm": 0.11240955066490231,
+      "learning_rate": 6.948959528500396e-06,
+      "loss": 0.3865,
+      "step": 2084
+    },
+    {
+      "epoch": 4.145129224652088,
+      "grad_norm": 0.11296598058113064,
+      "learning_rate": 6.91771342731209e-06,
+      "loss": 0.387,
+      "step": 2085
+    },
+    {
+      "epoch": 4.147117296222664,
+      "grad_norm": 0.09714300484621313,
+      "learning_rate": 6.886531082861503e-06,
+      "loss": 0.4024,
+      "step": 2086
+    },
+    {
+      "epoch": 4.149105367793241,
+      "grad_norm": 0.11486164583751605,
+      "learning_rate": 6.855412555243743e-06,
+      "loss": 0.3884,
+      "step": 2087
+    },
+    {
+      "epoch": 4.151093439363817,
+      "grad_norm": 0.10858190643310542,
+      "learning_rate": 6.824357904430998e-06,
+      "loss": 0.3943,
+      "step": 2088
+    },
+    {
+      "epoch": 4.153081510934394,
+      "grad_norm": 0.10169998242634819,
+      "learning_rate": 6.7933671902722955e-06,
+      "loss": 0.3973,
+      "step": 2089
+    },
+    {
+      "epoch": 4.155069582504971,
+      "grad_norm": 0.11225341199262792,
+      "learning_rate": 6.762440472493459e-06,
+      "loss": 0.3884,
+      "step": 2090
+    },
+    {
+      "epoch": 4.157057654075547,
+      "grad_norm": 0.10809234552066299,
+      "learning_rate": 6.731577810696981e-06,
+      "loss": 0.383,
+      "step": 2091
+    },
+    {
+      "epoch": 4.1590457256461235,
+      "grad_norm": 0.1162247680241023,
+      "learning_rate": 6.700779264361896e-06,
+      "loss": 0.3948,
+      "step": 2092
+    },
+    {
+      "epoch": 4.1610337972166995,
+      "grad_norm": 0.10331984098873934,
+      "learning_rate": 6.670044892843686e-06,
+      "loss": 0.3904,
+      "step": 2093
+    },
+    {
+      "epoch": 4.163021868787276,
+      "grad_norm": 0.11263645547156585,
+      "learning_rate": 6.6393747553741416e-06,
+      "loss": 0.3935,
+      "step": 2094
+    },
+    {
+      "epoch": 4.165009940357853,
+      "grad_norm": 0.11731311349840723,
+      "learning_rate": 6.608768911061281e-06,
+      "loss": 0.3939,
+      "step": 2095
+    },
+    {
+      "epoch": 4.166998011928429,
+      "grad_norm": 0.10993203224686413,
+      "learning_rate": 6.5782274188891696e-06,
+      "loss": 0.3956,
+      "step": 2096
+    },
+    {
+      "epoch": 4.168986083499006,
+      "grad_norm": 0.1517967414260752,
+      "learning_rate": 6.547750337717906e-06,
+      "loss": 0.3921,
+      "step": 2097
+    },
+    {
+      "epoch": 4.170974155069582,
+      "grad_norm": 0.10352035003831739,
+      "learning_rate": 6.517337726283437e-06,
+      "loss": 0.3901,
+      "step": 2098
+    },
+    {
+      "epoch": 4.172962226640159,
+      "grad_norm": 0.1019881318386552,
+      "learning_rate": 6.486989643197446e-06,
+      "loss": 0.374,
+      "step": 2099
+    },
+    {
+      "epoch": 4.174950298210735,
+      "grad_norm": 0.1058930276911533,
+      "learning_rate": 6.456706146947271e-06,
+      "loss": 0.3952,
+      "step": 2100
+    },
+    {
+      "epoch": 4.176938369781312,
+      "grad_norm": 0.10884888055550279,
+      "learning_rate": 6.426487295895772e-06,
+      "loss": 0.3755,
+      "step": 2101
+    },
+    {
+      "epoch": 4.178926441351889,
+      "grad_norm": 0.10366779903871269,
+      "learning_rate": 6.396333148281231e-06,
+      "loss": 0.387,
+      "step": 2102
+    },
+    {
+      "epoch": 4.180914512922465,
+      "grad_norm": 0.1047566988829119,
+      "learning_rate": 6.366243762217226e-06,
+      "loss": 0.3964,
+      "step": 2103
+    },
+    {
+      "epoch": 4.182902584493042,
+      "grad_norm": 0.10172953838653928,
+      "learning_rate": 6.336219195692521e-06,
+      "loss": 0.3911,
+      "step": 2104
+    },
+    {
+      "epoch": 4.184890656063618,
+      "grad_norm": 0.10614996378516255,
+      "learning_rate": 6.306259506570968e-06,
+      "loss": 0.3826,
+      "step": 2105
+    },
+    {
+      "epoch": 4.186878727634195,
+      "grad_norm": 0.11907843686153102,
+      "learning_rate": 6.276364752591386e-06,
+      "loss": 0.3804,
+      "step": 2106
+    },
+    {
+      "epoch": 4.188866799204772,
+      "grad_norm": 0.11026202432242263,
+      "learning_rate": 6.2465349913674435e-06,
+      "loss": 0.3828,
+      "step": 2107
+    },
+    {
+      "epoch": 4.190854870775348,
+      "grad_norm": 0.10861909520932465,
+      "learning_rate": 6.216770280387554e-06,
+      "loss": 0.3825,
+      "step": 2108
+    },
+    {
+      "epoch": 4.192842942345925,
+      "grad_norm": 0.11130931558291521,
+      "learning_rate": 6.18707067701477e-06,
+      "loss": 0.3889,
+      "step": 2109
+    },
+    {
+      "epoch": 4.194831013916501,
+      "grad_norm": 0.10489299799335888,
+      "learning_rate": 6.157436238486671e-06,
+      "loss": 0.3868,
+      "step": 2110
+    },
+    {
+      "epoch": 4.1968190854870775,
+      "grad_norm": 0.11564230886287205,
+      "learning_rate": 6.127867021915235e-06,
+      "loss": 0.3829,
+      "step": 2111
+    },
+    {
+      "epoch": 4.198807157057654,
+      "grad_norm": 0.10738774680568806,
+      "learning_rate": 6.098363084286765e-06,
+      "loss": 0.3902,
+      "step": 2112
+    },
+    {
+      "epoch": 4.20079522862823,
+      "grad_norm": 0.10870599960288968,
+      "learning_rate": 6.068924482461715e-06,
+      "loss": 0.389,
+      "step": 2113
+    },
+    {
+      "epoch": 4.202783300198807,
+      "grad_norm": 0.1160575426010257,
+      "learning_rate": 6.0395512731746865e-06,
+      "loss": 0.3827,
+      "step": 2114
+    },
+    {
+      "epoch": 4.204771371769383,
+      "grad_norm": 0.11306539716214972,
+      "learning_rate": 6.0102435130342085e-06,
+      "loss": 0.3926,
+      "step": 2115
+    },
+    {
+      "epoch": 4.20675944333996,
+      "grad_norm": 0.11006470554733454,
+      "learning_rate": 5.981001258522683e-06,
+      "loss": 0.3984,
+      "step": 2116
+    },
+    {
+      "epoch": 4.208747514910537,
+      "grad_norm": 0.1099261910301456,
+      "learning_rate": 5.951824565996282e-06,
+      "loss": 0.3809,
+      "step": 2117
+    },
+    {
+      "epoch": 4.210735586481113,
+      "grad_norm": 0.11092749383381181,
+      "learning_rate": 5.922713491684811e-06,
+      "loss": 0.3862,
+      "step": 2118
+    },
+    {
+      "epoch": 4.21272365805169,
+      "grad_norm": 0.09827057623284234,
+      "learning_rate": 5.893668091691625e-06,
+      "loss": 0.3935,
+      "step": 2119
+    },
+    {
+      "epoch": 4.214711729622266,
+      "grad_norm": 0.1087451514423072,
+      "learning_rate": 5.864688421993499e-06,
+      "loss": 0.3906,
+      "step": 2120
+    },
+    {
+      "epoch": 4.216699801192843,
+      "grad_norm": 0.10709955492019746,
+      "learning_rate": 5.835774538440557e-06,
+      "loss": 0.3878,
+      "step": 2121
+    },
+    {
+      "epoch": 4.21868787276342,
+      "grad_norm": 0.10181961138592265,
+      "learning_rate": 5.806926496756084e-06,
+      "loss": 0.3812,
+      "step": 2122
+    },
+    {
+      "epoch": 4.220675944333996,
+      "grad_norm": 0.1062337392337739,
+      "learning_rate": 5.778144352536545e-06,
+      "loss": 0.3836,
+      "step": 2123
+    },
+    {
+      "epoch": 4.222664015904573,
+      "grad_norm": 0.10197581797668581,
+      "learning_rate": 5.74942816125136e-06,
+      "loss": 0.3824,
+      "step": 2124
+    },
+    {
+      "epoch": 4.224652087475149,
+      "grad_norm": 0.10219575099997118,
+      "learning_rate": 5.720777978242851e-06,
+      "loss": 0.3863,
+      "step": 2125
+    },
+    {
+      "epoch": 4.226640159045726,
+      "grad_norm": 0.1040267273692468,
+      "learning_rate": 5.692193858726134e-06,
+      "loss": 0.3874,
+      "step": 2126
+    },
+    {
+      "epoch": 4.2286282306163026,
+      "grad_norm": 0.11269177444316061,
+      "learning_rate": 5.663675857789006e-06,
+      "loss": 0.3865,
+      "step": 2127
+    },
+    {
+      "epoch": 4.230616302186879,
+      "grad_norm": 0.09995425296412798,
+      "learning_rate": 5.635224030391832e-06,
+      "loss": 0.3933,
+      "step": 2128
+    },
+    {
+      "epoch": 4.2326043737574555,
+      "grad_norm": 0.10689562263264148,
+      "learning_rate": 5.606838431367454e-06,
+      "loss": 0.3921,
+      "step": 2129
+    },
+    {
+      "epoch": 4.2345924453280315,
+      "grad_norm": 0.11122546471139345,
+      "learning_rate": 5.578519115421084e-06,
+      "loss": 0.3888,
+      "step": 2130
+    },
+    {
+      "epoch": 4.236580516898608,
+      "grad_norm": 0.10344950837778637,
+      "learning_rate": 5.550266137130154e-06,
+      "loss": 0.3947,
+      "step": 2131
+    },
+    {
+      "epoch": 4.238568588469185,
+      "grad_norm": 0.10216325643865894,
+      "learning_rate": 5.5220795509443085e-06,
+      "loss": 0.3901,
+      "step": 2132
+    },
+    {
+      "epoch": 4.240556660039761,
+      "grad_norm": 0.10694524647608616,
+      "learning_rate": 5.493959411185197e-06,
+      "loss": 0.3942,
+      "step": 2133
+    },
+    {
+      "epoch": 4.242544731610338,
+      "grad_norm": 0.10609283465242796,
+      "learning_rate": 5.465905772046425e-06,
+      "loss": 0.3793,
+      "step": 2134
+    },
+    {
+      "epoch": 4.244532803180914,
+      "grad_norm": 0.1079949840251626,
+      "learning_rate": 5.437918687593433e-06,
+      "loss": 0.4008,
+      "step": 2135
+    },
+    {
+      "epoch": 4.246520874751491,
+      "grad_norm": 0.11738376878238611,
+      "learning_rate": 5.40999821176341e-06,
+      "loss": 0.3848,
+      "step": 2136
+    },
+    {
+      "epoch": 4.248508946322068,
+      "grad_norm": 0.11082300608152638,
+      "learning_rate": 5.3821443983651525e-06,
+      "loss": 0.3987,
+      "step": 2137
+    },
+    {
+      "epoch": 4.250497017892644,
+      "grad_norm": 0.10101109241687987,
+      "learning_rate": 5.3543573010790095e-06,
+      "loss": 0.394,
+      "step": 2138
+    },
+    {
+      "epoch": 4.252485089463221,
+      "grad_norm": 0.11468964097949289,
+      "learning_rate": 5.326636973456741e-06,
+      "loss": 0.4014,
+      "step": 2139
+    },
+    {
+      "epoch": 4.254473161033797,
+      "grad_norm": 0.10452528767883948,
+      "learning_rate": 5.298983468921402e-06,
+      "loss": 0.383,
+      "step": 2140
+    },
+    {
+      "epoch": 4.256461232604374,
+      "grad_norm": 0.10689689020767205,
+      "learning_rate": 5.2713968407673225e-06,
+      "loss": 0.3879,
+      "step": 2141
+    },
+    {
+      "epoch": 4.258449304174951,
+      "grad_norm": 0.10520846283587974,
+      "learning_rate": 5.243877142159899e-06,
+      "loss": 0.3919,
+      "step": 2142
+    },
+    {
+      "epoch": 4.260437375745527,
+      "grad_norm": 0.11093695400298927,
+      "learning_rate": 5.21642442613556e-06,
+      "loss": 0.378,
+      "step": 2143
+    },
+    {
+      "epoch": 4.262425447316104,
+      "grad_norm": 0.11329605962872928,
+      "learning_rate": 5.1890387456016425e-06,
+      "loss": 0.3902,
+      "step": 2144
+    },
+    {
+      "epoch": 4.26441351888668,
+      "grad_norm": 0.10901614052465748,
+      "learning_rate": 5.161720153336287e-06,
+      "loss": 0.3883,
+      "step": 2145
+    },
+    {
+      "epoch": 4.2664015904572565,
+      "grad_norm": 0.11033984670188425,
+      "learning_rate": 5.134468701988345e-06,
+      "loss": 0.3848,
+      "step": 2146
+    },
+    {
+      "epoch": 4.2683896620278325,
+      "grad_norm": 0.1112581347434242,
+      "learning_rate": 5.107284444077265e-06,
+      "loss": 0.3777,
+      "step": 2147
+    },
+    {
+      "epoch": 4.270377733598409,
+      "grad_norm": 0.11604790354287506,
+      "learning_rate": 5.080167431993018e-06,
+      "loss": 0.3936,
+      "step": 2148
+    },
+    {
+      "epoch": 4.272365805168986,
+      "grad_norm": 0.10130942223708193,
+      "learning_rate": 5.053117717995939e-06,
+      "loss": 0.3916,
+      "step": 2149
+    },
+    {
+      "epoch": 4.274353876739562,
+      "grad_norm": 0.10942647069946232,
+      "learning_rate": 5.026135354216717e-06,
+      "loss": 0.3885,
+      "step": 2150
+    },
+    {
+      "epoch": 4.276341948310139,
+      "grad_norm": 0.11395657047861649,
+      "learning_rate": 4.999220392656203e-06,
+      "loss": 0.3945,
+      "step": 2151
+    },
+    {
+      "epoch": 4.278330019880716,
+      "grad_norm": 0.10700487248942017,
+      "learning_rate": 4.972372885185368e-06,
+      "loss": 0.389,
+      "step": 2152
+    },
+    {
+      "epoch": 4.280318091451292,
+      "grad_norm": 0.10391399259588788,
+      "learning_rate": 4.945592883545179e-06,
+      "loss": 0.3934,
+      "step": 2153
+    },
+    {
+      "epoch": 4.282306163021869,
+      "grad_norm": 0.10037438756173153,
+      "learning_rate": 4.918880439346496e-06,
+      "loss": 0.381,
+      "step": 2154
+    },
+    {
+      "epoch": 4.284294234592445,
+      "grad_norm": 0.10597748791604826,
+      "learning_rate": 4.892235604069999e-06,
+      "loss": 0.3916,
+      "step": 2155
+    },
+    {
+      "epoch": 4.286282306163022,
+      "grad_norm": 0.10257742187290152,
+      "learning_rate": 4.86565842906606e-06,
+      "loss": 0.3873,
+      "step": 2156
+    },
+    {
+      "epoch": 4.288270377733598,
+      "grad_norm": 0.10673572419919322,
+      "learning_rate": 4.8391489655546495e-06,
+      "loss": 0.3734,
+      "step": 2157
+    },
+    {
+      "epoch": 4.290258449304175,
+      "grad_norm": 0.11498553199702016,
+      "learning_rate": 4.81270726462526e-06,
+      "loss": 0.3931,
+      "step": 2158
+    },
+    {
+      "epoch": 4.292246520874752,
+      "grad_norm": 0.10798065238474126,
+      "learning_rate": 4.786333377236773e-06,
+      "loss": 0.4006,
+      "step": 2159
+    },
+    {
+      "epoch": 4.294234592445328,
+      "grad_norm": 0.09816552037676422,
+      "learning_rate": 4.760027354217394e-06,
+      "loss": 0.3847,
+      "step": 2160
+    },
+    {
+      "epoch": 4.296222664015905,
+      "grad_norm": 0.10237023642683052,
+      "learning_rate": 4.733789246264526e-06,
+      "loss": 0.3985,
+      "step": 2161
+    },
+    {
+      "epoch": 4.298210735586481,
+      "grad_norm": 0.09693280463118617,
+      "learning_rate": 4.707619103944696e-06,
+      "loss": 0.3851,
+      "step": 2162
+    },
+    {
+      "epoch": 4.300198807157058,
+      "grad_norm": 0.10605120716456504,
+      "learning_rate": 4.681516977693435e-06,
+      "loss": 0.3883,
+      "step": 2163
+    },
+    {
+      "epoch": 4.3021868787276345,
+      "grad_norm": 0.09960557557894663,
+      "learning_rate": 4.6554829178152035e-06,
+      "loss": 0.3921,
+      "step": 2164
+    },
+    {
+      "epoch": 4.3041749502982105,
+      "grad_norm": 0.09498386093083336,
+      "learning_rate": 4.629516974483279e-06,
+      "loss": 0.3931,
+      "step": 2165
+    },
+    {
+      "epoch": 4.306163021868787,
+      "grad_norm": 0.09991291721444792,
+      "learning_rate": 4.603619197739662e-06,
+      "loss": 0.3997,
+      "step": 2166
+    },
+    {
+      "epoch": 4.308151093439363,
+      "grad_norm": 0.10582505708208029,
+      "learning_rate": 4.577789637494978e-06,
+      "loss": 0.3879,
+      "step": 2167
+    },
+    {
+      "epoch": 4.31013916500994,
+      "grad_norm": 0.1114427277686471,
+      "learning_rate": 4.552028343528396e-06,
+      "loss": 0.3773,
+      "step": 2168
+    },
+    {
+      "epoch": 4.312127236580517,
+      "grad_norm": 0.09989286001429934,
+      "learning_rate": 4.526335365487509e-06,
+      "loss": 0.3837,
+      "step": 2169
+    },
+    {
+      "epoch": 4.314115308151093,
+      "grad_norm": 0.10342275152788724,
+      "learning_rate": 4.500710752888253e-06,
+      "loss": 0.3883,
+      "step": 2170
+    },
+    {
+      "epoch": 4.31610337972167,
+      "grad_norm": 0.10316817695567211,
+      "learning_rate": 4.47515455511482e-06,
+      "loss": 0.4026,
+      "step": 2171
+    },
+    {
+      "epoch": 4.318091451292246,
+      "grad_norm": 0.09876129834768971,
+      "learning_rate": 4.449666821419536e-06,
+      "loss": 0.3831,
+      "step": 2172
+    },
+    {
+      "epoch": 4.320079522862823,
+      "grad_norm": 0.11477440375009836,
+      "learning_rate": 4.424247600922793e-06,
+      "loss": 0.3872,
+      "step": 2173
+    },
+    {
+      "epoch": 4.3220675944334,
+      "grad_norm": 0.10099717624479602,
+      "learning_rate": 4.398896942612934e-06,
+      "loss": 0.3947,
+      "step": 2174
+    },
+    {
+      "epoch": 4.324055666003976,
+      "grad_norm": 0.10121838635908194,
+      "learning_rate": 4.373614895346178e-06,
+      "loss": 0.3931,
+      "step": 2175
+    },
+    {
+      "epoch": 4.326043737574553,
+      "grad_norm": 0.10716998309395787,
+      "learning_rate": 4.348401507846509e-06,
+      "loss": 0.3837,
+      "step": 2176
+    },
+    {
+      "epoch": 4.328031809145129,
+      "grad_norm": 0.0976333146806361,
+      "learning_rate": 4.323256828705593e-06,
+      "loss": 0.3882,
+      "step": 2177
+    },
+    {
+      "epoch": 4.330019880715706,
+      "grad_norm": 0.10697024916266006,
+      "learning_rate": 4.2981809063826715e-06,
+      "loss": 0.3816,
+      "step": 2178
+    },
+    {
+      "epoch": 4.332007952286283,
+      "grad_norm": 0.10052526954739248,
+      "learning_rate": 4.273173789204488e-06,
+      "loss": 0.3811,
+      "step": 2179
+    },
+    {
+      "epoch": 4.333996023856859,
+      "grad_norm": 0.09928114782318379,
+      "learning_rate": 4.248235525365174e-06,
+      "loss": 0.379,
+      "step": 2180
+    },
+    {
+      "epoch": 4.335984095427436,
+      "grad_norm": 0.0999330585185468,
+      "learning_rate": 4.2233661629261746e-06,
+      "loss": 0.4004,
+      "step": 2181
+    },
+    {
+      "epoch": 4.337972166998012,
+      "grad_norm": 0.1093719529042732,
+      "learning_rate": 4.198565749816142e-06,
+      "loss": 0.4014,
+      "step": 2182
+    },
+    {
+      "epoch": 4.3399602385685885,
+      "grad_norm": 0.0990449385264406,
+      "learning_rate": 4.173834333830842e-06,
+      "loss": 0.3851,
+      "step": 2183
+    },
+    {
+      "epoch": 4.341948310139165,
+      "grad_norm": 0.09842163804658173,
+      "learning_rate": 4.149171962633083e-06,
+      "loss": 0.3771,
+      "step": 2184
+    },
+    {
+      "epoch": 4.343936381709741,
+      "grad_norm": 0.098524628734478,
+      "learning_rate": 4.124578683752596e-06,
+      "loss": 0.3851,
+      "step": 2185
+    },
+    {
+      "epoch": 4.345924453280318,
+      "grad_norm": 0.09531813272536584,
+      "learning_rate": 4.100054544585961e-06,
+      "loss": 0.3912,
+      "step": 2186
+    },
+    {
+      "epoch": 4.347912524850894,
+      "grad_norm": 0.09791263411124397,
+      "learning_rate": 4.075599592396509e-06,
+      "loss": 0.3826,
+      "step": 2187
+    },
+    {
+      "epoch": 4.349900596421471,
+      "grad_norm": 0.09273942541729,
+      "learning_rate": 4.051213874314237e-06,
+      "loss": 0.3951,
+      "step": 2188
+    },
+    {
+      "epoch": 4.351888667992048,
+      "grad_norm": 0.1050206316055863,
+      "learning_rate": 4.026897437335708e-06,
+      "loss": 0.3808,
+      "step": 2189
+    },
+    {
+      "epoch": 4.353876739562624,
+      "grad_norm": 0.10040959376895207,
+      "learning_rate": 4.002650328323969e-06,
+      "loss": 0.385,
+      "step": 2190
+    },
+    {
+      "epoch": 4.355864811133201,
+      "grad_norm": 0.1016395750899665,
+      "learning_rate": 3.978472594008454e-06,
+      "loss": 0.3944,
+      "step": 2191
+    },
+    {
+      "epoch": 4.357852882703777,
+      "grad_norm": 0.0993656184763306,
+      "learning_rate": 3.954364280984906e-06,
+      "loss": 0.3904,
+      "step": 2192
+    },
+    {
+      "epoch": 4.359840954274354,
+      "grad_norm": 0.0943218423219894,
+      "learning_rate": 3.930325435715245e-06,
+      "loss": 0.3936,
+      "step": 2193
+    },
+    {
+      "epoch": 4.361829025844931,
+      "grad_norm": 0.09681840686767862,
+      "learning_rate": 3.906356104527565e-06,
+      "loss": 0.3947,
+      "step": 2194
+    },
+    {
+      "epoch": 4.363817097415507,
+      "grad_norm": 0.09709721598226503,
+      "learning_rate": 3.882456333615952e-06,
+      "loss": 0.3878,
+      "step": 2195
+    },
+    {
+      "epoch": 4.365805168986084,
+      "grad_norm": 0.09913382152372414,
+      "learning_rate": 3.858626169040447e-06,
+      "loss": 0.3843,
+      "step": 2196
+    },
+    {
+      "epoch": 4.36779324055666,
+      "grad_norm": 0.09980848225328527,
+      "learning_rate": 3.834865656726945e-06,
+      "loss": 0.3909,
+      "step": 2197
+    },
+    {
+      "epoch": 4.369781312127237,
+      "grad_norm": 0.09420624149064247,
+      "learning_rate": 3.8111748424671e-06,
+      "loss": 0.3942,
+      "step": 2198
+    },
+    {
+      "epoch": 4.3717693836978135,
+      "grad_norm": 0.09687823459593438,
+      "learning_rate": 3.7875537719182486e-06,
+      "loss": 0.3869,
+      "step": 2199
+    },
+    {
+      "epoch": 4.3737574552683895,
+      "grad_norm": 0.10595045404187366,
+      "learning_rate": 3.7640024906033134e-06,
+      "loss": 0.3891,
+      "step": 2200
+    },
+    {
+      "epoch": 4.3757455268389664,
+      "grad_norm": 0.09612367938332525,
+      "learning_rate": 3.7405210439107254e-06,
+      "loss": 0.379,
+      "step": 2201
+    },
+    {
+      "epoch": 4.3777335984095425,
+      "grad_norm": 0.1023095190589734,
+      "learning_rate": 3.7171094770943026e-06,
+      "loss": 0.3875,
+      "step": 2202
+    },
+    {
+      "epoch": 4.379721669980119,
+      "grad_norm": 0.099151488292962,
+      "learning_rate": 3.69376783527323e-06,
+      "loss": 0.3898,
+      "step": 2203
+    },
+    {
+      "epoch": 4.381709741550695,
+      "grad_norm": 0.09958329292525986,
+      "learning_rate": 3.6704961634319046e-06,
+      "loss": 0.3851,
+      "step": 2204
+    },
+    {
+      "epoch": 4.383697813121272,
+      "grad_norm": 0.09555701123264873,
+      "learning_rate": 3.6472945064198827e-06,
+      "loss": 0.3889,
+      "step": 2205
+    },
+    {
+      "epoch": 4.385685884691849,
+      "grad_norm": 0.09980480198930511,
+      "learning_rate": 3.6241629089517873e-06,
+      "loss": 0.3804,
+      "step": 2206
+    },
+    {
+      "epoch": 4.387673956262425,
+      "grad_norm": 0.09835469378518892,
+      "learning_rate": 3.6011014156072242e-06,
+      "loss": 0.3924,
+      "step": 2207
+    },
+    {
+      "epoch": 4.389662027833002,
+      "grad_norm": 0.10440527313630382,
+      "learning_rate": 3.5781100708306827e-06,
+      "loss": 0.3905,
+      "step": 2208
+    },
+    {
+      "epoch": 4.391650099403579,
+      "grad_norm": 0.10830643367489604,
+      "learning_rate": 3.5551889189314738e-06,
+      "loss": 0.396,
+      "step": 2209
+    },
+    {
+      "epoch": 4.393638170974155,
+      "grad_norm": 0.10721642274121215,
+      "learning_rate": 3.532338004083631e-06,
+      "loss": 0.3916,
+      "step": 2210
+    },
+    {
+      "epoch": 4.395626242544732,
+      "grad_norm": 0.09987362704278649,
+      "learning_rate": 3.5095573703258023e-06,
+      "loss": 0.3914,
+      "step": 2211
+    },
+    {
+      "epoch": 4.397614314115308,
+      "grad_norm": 0.10841817698714662,
+      "learning_rate": 3.4868470615612247e-06,
+      "loss": 0.3831,
+      "step": 2212
+    },
+    {
+      "epoch": 4.399602385685885,
+      "grad_norm": 0.10340887834099978,
+      "learning_rate": 3.464207121557581e-06,
+      "loss": 0.4053,
+      "step": 2213
+    },
+    {
+      "epoch": 4.401590457256461,
+      "grad_norm": 0.09365502481903334,
+      "learning_rate": 3.4416375939469427e-06,
+      "loss": 0.3876,
+      "step": 2214
+    },
+    {
+      "epoch": 4.403578528827038,
+      "grad_norm": 0.0999030246429339,
+      "learning_rate": 3.4191385222256845e-06,
+      "loss": 0.3802,
+      "step": 2215
+    },
+    {
+      "epoch": 4.405566600397615,
+      "grad_norm": 0.1012356919744598,
+      "learning_rate": 3.396709949754393e-06,
+      "loss": 0.3889,
+      "step": 2216
+    },
+    {
+      "epoch": 4.407554671968191,
+      "grad_norm": 0.10343516451567464,
+      "learning_rate": 3.37435191975779e-06,
+      "loss": 0.3996,
+      "step": 2217
+    },
+    {
+      "epoch": 4.4095427435387675,
+      "grad_norm": 0.596392899901363,
+      "learning_rate": 3.3520644753246432e-06,
+      "loss": 0.4044,
+      "step": 2218
+    },
+    {
+      "epoch": 4.4115308151093435,
+      "grad_norm": 0.09379139186711735,
+      "learning_rate": 3.3298476594076923e-06,
+      "loss": 0.3791,
+      "step": 2219
+    },
+    {
+      "epoch": 4.41351888667992,
+      "grad_norm": 0.10187480860768795,
+      "learning_rate": 3.307701514823558e-06,
+      "loss": 0.3834,
+      "step": 2220
+    },
+    {
+      "epoch": 4.415506958250497,
+      "grad_norm": 0.10571085909393454,
+      "learning_rate": 3.28562608425266e-06,
+      "loss": 0.39,
+      "step": 2221
+    },
+    {
+      "epoch": 4.417495029821073,
+      "grad_norm": 0.10014337973432315,
+      "learning_rate": 3.263621410239144e-06,
+      "loss": 0.3913,
+      "step": 2222
+    },
+    {
+      "epoch": 4.41948310139165,
+      "grad_norm": 0.1092625763710919,
+      "learning_rate": 3.241687535190776e-06,
+      "loss": 0.4004,
+      "step": 2223
+    },
+    {
+      "epoch": 4.421471172962226,
+      "grad_norm": 0.09879099002909277,
+      "learning_rate": 3.2198245013788943e-06,
+      "loss": 0.393,
+      "step": 2224
+    },
+    {
+      "epoch": 4.423459244532803,
+      "grad_norm": 0.36421280599677913,
+      "learning_rate": 3.198032350938309e-06,
+      "loss": 0.3872,
+      "step": 2225
+    },
+    {
+      "epoch": 4.42544731610338,
+      "grad_norm": 0.10043741715437189,
+      "learning_rate": 3.1763111258672084e-06,
+      "loss": 0.3811,
+      "step": 2226
+    },
+    {
+      "epoch": 4.427435387673956,
+      "grad_norm": 0.10483503269426613,
+      "learning_rate": 3.1546608680271105e-06,
+      "loss": 0.3925,
+      "step": 2227
+    },
+    {
+      "epoch": 4.429423459244533,
+      "grad_norm": 0.09693474001090108,
+      "learning_rate": 3.1330816191427505e-06,
+      "loss": 0.3928,
+      "step": 2228
+    },
+    {
+      "epoch": 4.431411530815109,
+      "grad_norm": 0.09737601508794358,
+      "learning_rate": 3.111573420802021e-06,
+      "loss": 0.3834,
+      "step": 2229
+    },
+    {
+      "epoch": 4.433399602385686,
+      "grad_norm": 0.1010936127284044,
+      "learning_rate": 3.0901363144558804e-06,
+      "loss": 0.3856,
+      "step": 2230
+    },
+    {
+      "epoch": 4.435387673956263,
+      "grad_norm": 0.1048068301583792,
+      "learning_rate": 3.0687703414182814e-06,
+      "loss": 0.3873,
+      "step": 2231
+    },
+    {
+      "epoch": 4.437375745526839,
+      "grad_norm": 0.09218979311028978,
+      "learning_rate": 3.0474755428660853e-06,
+      "loss": 0.3869,
+      "step": 2232
+    },
+    {
+      "epoch": 4.439363817097416,
+      "grad_norm": 0.09891825070474362,
+      "learning_rate": 3.0262519598389883e-06,
+      "loss": 0.392,
+      "step": 2233
+    },
+    {
+      "epoch": 4.441351888667992,
+      "grad_norm": 0.09824471049330011,
+      "learning_rate": 3.005099633239428e-06,
+      "loss": 0.3878,
+      "step": 2234
+    },
+    {
+      "epoch": 4.443339960238569,
+      "grad_norm": 0.10111959954987662,
+      "learning_rate": 2.9840186038325326e-06,
+      "loss": 0.3925,
+      "step": 2235
+    },
+    {
+      "epoch": 4.4453280318091455,
+      "grad_norm": 0.09410522943055849,
+      "learning_rate": 2.963008912246008e-06,
+      "loss": 0.391,
+      "step": 2236
+    },
+    {
+      "epoch": 4.4473161033797215,
+      "grad_norm": 0.1025495604201037,
+      "learning_rate": 2.9420705989700882e-06,
+      "loss": 0.3953,
+      "step": 2237
+    },
+    {
+      "epoch": 4.449304174950298,
+      "grad_norm": 0.09987461410476814,
+      "learning_rate": 2.9212037043574402e-06,
+      "loss": 0.384,
+      "step": 2238
+    },
+    {
+      "epoch": 4.451292246520874,
+      "grad_norm": 0.0926074391620245,
+      "learning_rate": 2.9004082686230916e-06,
+      "loss": 0.3895,
+      "step": 2239
+    },
+    {
+      "epoch": 4.453280318091451,
+      "grad_norm": 0.09482564866145016,
+      "learning_rate": 2.879684331844361e-06,
+      "loss": 0.3762,
+      "step": 2240
+    },
+    {
+      "epoch": 4.455268389662028,
+      "grad_norm": 0.13865518028486007,
+      "learning_rate": 2.859031933960763e-06,
+      "loss": 0.3906,
+      "step": 2241
+    },
+    {
+      "epoch": 4.457256461232604,
+      "grad_norm": 0.09264050462975766,
+      "learning_rate": 2.8384511147739437e-06,
+      "loss": 0.4027,
+      "step": 2242
+    },
+    {
+      "epoch": 4.459244532803181,
+      "grad_norm": 0.10005832449719292,
+      "learning_rate": 2.817941913947606e-06,
+      "loss": 0.3805,
+      "step": 2243
+    },
+    {
+      "epoch": 4.461232604373757,
+      "grad_norm": 0.09821994260465117,
+      "learning_rate": 2.797504371007427e-06,
+      "loss": 0.3937,
+      "step": 2244
+    },
+    {
+      "epoch": 4.463220675944334,
+      "grad_norm": 0.09631788391400439,
+      "learning_rate": 2.777138525340979e-06,
+      "loss": 0.3827,
+      "step": 2245
+    },
+    {
+      "epoch": 4.465208747514911,
+      "grad_norm": 0.09039255877447051,
+      "learning_rate": 2.756844416197666e-06,
+      "loss": 0.3825,
+      "step": 2246
+    },
+    {
+      "epoch": 4.467196819085487,
+      "grad_norm": 0.09465359257625011,
+      "learning_rate": 2.7366220826886294e-06,
+      "loss": 0.3809,
+      "step": 2247
+    },
+    {
+      "epoch": 4.469184890656064,
+      "grad_norm": 0.09254465134150146,
+      "learning_rate": 2.716471563786689e-06,
+      "loss": 0.3979,
+      "step": 2248
+    },
+    {
+      "epoch": 4.47117296222664,
+      "grad_norm": 0.09612944852010466,
+      "learning_rate": 2.6963928983262745e-06,
+      "loss": 0.3848,
+      "step": 2249
+    },
+    {
+      "epoch": 4.473161033797217,
+      "grad_norm": 0.105893288150653,
+      "learning_rate": 2.6763861250033118e-06,
+      "loss": 0.3916,
+      "step": 2250
+    },
+    {
+      "epoch": 4.475149105367794,
+      "grad_norm": 0.09300696972063668,
+      "learning_rate": 2.6564512823751987e-06,
+      "loss": 0.3994,
+      "step": 2251
+    },
+    {
+      "epoch": 4.47713717693837,
+      "grad_norm": 0.0930437296362057,
+      "learning_rate": 2.636588408860701e-06,
+      "loss": 0.3853,
+      "step": 2252
+    },
+    {
+      "epoch": 4.4791252485089466,
+      "grad_norm": 0.09325528669971057,
+      "learning_rate": 2.616797542739873e-06,
+      "loss": 0.3775,
+      "step": 2253
+    },
+    {
+      "epoch": 4.481113320079523,
+      "grad_norm": 0.09465774340633075,
+      "learning_rate": 2.597078722154014e-06,
+      "loss": 0.387,
+      "step": 2254
+    },
+    {
+      "epoch": 4.4831013916500995,
+      "grad_norm": 0.13768709356858644,
+      "learning_rate": 2.5774319851055564e-06,
+      "loss": 0.383,
+      "step": 2255
+    },
+    {
+      "epoch": 4.485089463220676,
+      "grad_norm": 0.09645823543283955,
+      "learning_rate": 2.5578573694580254e-06,
+      "loss": 0.3807,
+      "step": 2256
+    },
+    {
+      "epoch": 4.487077534791252,
+      "grad_norm": 0.09758883625704344,
+      "learning_rate": 2.5383549129359473e-06,
+      "loss": 0.3853,
+      "step": 2257
+    },
+    {
+      "epoch": 4.489065606361829,
+      "grad_norm": 0.09776875136116465,
+      "learning_rate": 2.51892465312479e-06,
+      "loss": 0.3795,
+      "step": 2258
+    },
+    {
+      "epoch": 4.491053677932405,
+      "grad_norm": 0.09531997029739311,
+      "learning_rate": 2.4995666274708664e-06,
+      "loss": 0.3859,
+      "step": 2259
+    },
+    {
+      "epoch": 4.493041749502982,
+      "grad_norm": 0.09613584567556226,
+      "learning_rate": 2.4802808732812935e-06,
+      "loss": 0.3963,
+      "step": 2260
+    },
+    {
+      "epoch": 4.495029821073558,
+      "grad_norm": 0.10310822349287414,
+      "learning_rate": 2.4610674277239046e-06,
+      "loss": 0.3944,
+      "step": 2261
+    },
+    {
+      "epoch": 4.497017892644135,
+      "grad_norm": 0.09144966609722156,
+      "learning_rate": 2.4419263278271643e-06,
+      "loss": 0.3956,
+      "step": 2262
+    },
+    {
+      "epoch": 4.499005964214712,
+      "grad_norm": 0.09492031385454708,
+      "learning_rate": 2.4228576104801295e-06,
+      "loss": 0.3877,
+      "step": 2263
+    },
+    {
+      "epoch": 4.500994035785288,
+      "grad_norm": 0.09327970875824858,
+      "learning_rate": 2.4038613124323496e-06,
+      "loss": 0.4014,
+      "step": 2264
+    },
+    {
+      "epoch": 4.502982107355865,
+      "grad_norm": 0.09558401559285043,
+      "learning_rate": 2.3849374702938065e-06,
+      "loss": 0.397,
+      "step": 2265
+    },
+    {
+      "epoch": 4.504970178926442,
+      "grad_norm": 0.09747152770568815,
+      "learning_rate": 2.366086120534852e-06,
+      "loss": 0.3801,
+      "step": 2266
+    },
+    {
+      "epoch": 4.506958250497018,
+      "grad_norm": 0.09409003277053672,
+      "learning_rate": 2.3473072994861167e-06,
+      "loss": 0.3935,
+      "step": 2267
+    },
+    {
+      "epoch": 4.508946322067595,
+      "grad_norm": 0.09512871106275056,
+      "learning_rate": 2.3286010433384656e-06,
+      "loss": 0.3903,
+      "step": 2268
+    },
+    {
+      "epoch": 4.510934393638171,
+      "grad_norm": 0.0937640062202609,
+      "learning_rate": 2.3099673881429086e-06,
+      "loss": 0.3839,
+      "step": 2269
+    },
+    {
+      "epoch": 4.512922465208748,
+      "grad_norm": 0.09330775039486519,
+      "learning_rate": 2.2914063698105338e-06,
+      "loss": 0.3978,
+      "step": 2270
+    },
+    {
+      "epoch": 4.514910536779324,
+      "grad_norm": 0.09712706610467936,
+      "learning_rate": 2.2729180241124517e-06,
+      "loss": 0.3844,
+      "step": 2271
+    },
+    {
+      "epoch": 4.5168986083499005,
+      "grad_norm": 0.09403036480214051,
+      "learning_rate": 2.254502386679711e-06,
+      "loss": 0.3943,
+      "step": 2272
+    },
+    {
+      "epoch": 4.518886679920477,
+      "grad_norm": 0.10067963044125913,
+      "learning_rate": 2.2361594930032338e-06,
+      "loss": 0.4011,
+      "step": 2273
+    },
+    {
+      "epoch": 4.5208747514910534,
+      "grad_norm": 0.09169835381585156,
+      "learning_rate": 2.2178893784337594e-06,
+      "loss": 0.3937,
+      "step": 2274
+    },
+    {
+      "epoch": 4.52286282306163,
+      "grad_norm": 0.08850262851562063,
+      "learning_rate": 2.1996920781817544e-06,
+      "loss": 0.3802,
+      "step": 2275
+    },
+    {
+      "epoch": 4.524850894632207,
+      "grad_norm": 0.093917345011886,
+      "learning_rate": 2.1815676273173604e-06,
+      "loss": 0.3927,
+      "step": 2276
+    },
+    {
+      "epoch": 4.526838966202783,
+      "grad_norm": 0.09501270473791533,
+      "learning_rate": 2.1635160607703256e-06,
+      "loss": 0.3828,
+      "step": 2277
+    },
+    {
+      "epoch": 4.52882703777336,
+      "grad_norm": 0.09135115060354339,
+      "learning_rate": 2.145537413329932e-06,
+      "loss": 0.3835,
+      "step": 2278
+    },
+    {
+      "epoch": 4.530815109343936,
+      "grad_norm": 0.08990517715951528,
+      "learning_rate": 2.1276317196449315e-06,
+      "loss": 0.3955,
+      "step": 2279
+    },
+    {
+      "epoch": 4.532803180914513,
+      "grad_norm": 0.30857334159915706,
+      "learning_rate": 2.109799014223475e-06,
+      "loss": 0.3999,
+      "step": 2280
+    },
+    {
+      "epoch": 4.534791252485089,
+      "grad_norm": 0.09985728888122387,
+      "learning_rate": 2.092039331433049e-06,
+      "loss": 0.388,
+      "step": 2281
+    },
+    {
+      "epoch": 4.536779324055666,
+      "grad_norm": 0.0938150353363061,
+      "learning_rate": 2.0743527055004176e-06,
+      "loss": 0.3816,
+      "step": 2282
+    },
+    {
+      "epoch": 4.538767395626243,
+      "grad_norm": 0.09823097175682717,
+      "learning_rate": 2.0567391705115416e-06,
+      "loss": 0.3965,
+      "step": 2283
+    },
+    {
+      "epoch": 4.540755467196819,
+      "grad_norm": 0.086397808420212,
+      "learning_rate": 2.0391987604115204e-06,
+      "loss": 0.3896,
+      "step": 2284
+    },
+    {
+      "epoch": 4.542743538767396,
+      "grad_norm": 0.09653654555914719,
+      "learning_rate": 2.021731509004523e-06,
+      "loss": 0.4017,
+      "step": 2285
+    },
+    {
+      "epoch": 4.544731610337972,
+      "grad_norm": 0.09598593518878275,
+      "learning_rate": 2.0043374499537373e-06,
+      "loss": 0.3805,
+      "step": 2286
+    },
+    {
+      "epoch": 4.546719681908549,
+      "grad_norm": 0.09535832154199035,
+      "learning_rate": 1.987016616781281e-06,
+      "loss": 0.389,
+      "step": 2287
+    },
+    {
+      "epoch": 4.548707753479126,
+      "grad_norm": 0.09030336197873803,
+      "learning_rate": 1.969769042868155e-06,
+      "loss": 0.3782,
+      "step": 2288
+    },
+    {
+      "epoch": 4.550695825049702,
+      "grad_norm": 0.0900881624984173,
+      "learning_rate": 1.9525947614541694e-06,
+      "loss": 0.3825,
+      "step": 2289
+    },
+    {
+      "epoch": 4.5526838966202785,
+      "grad_norm": 0.0932574491444472,
+      "learning_rate": 1.9354938056378935e-06,
+      "loss": 0.3779,
+      "step": 2290
+    },
+    {
+      "epoch": 4.5546719681908545,
+      "grad_norm": 0.09410751821503942,
+      "learning_rate": 1.918466208376573e-06,
+      "loss": 0.3943,
+      "step": 2291
+    },
+    {
+      "epoch": 4.556660039761431,
+      "grad_norm": 0.09098872687261256,
+      "learning_rate": 1.9015120024860811e-06,
+      "loss": 0.3785,
+      "step": 2292
+    },
+    {
+      "epoch": 4.558648111332008,
+      "grad_norm": 0.08926273695926924,
+      "learning_rate": 1.8846312206408468e-06,
+      "loss": 0.3871,
+      "step": 2293
+    },
+    {
+      "epoch": 4.560636182902584,
+      "grad_norm": 0.08923403092076015,
+      "learning_rate": 1.8678238953737926e-06,
+      "loss": 0.392,
+      "step": 2294
+    },
+    {
+      "epoch": 4.562624254473161,
+      "grad_norm": 0.08970320411457725,
+      "learning_rate": 1.8510900590762838e-06,
+      "loss": 0.3769,
+      "step": 2295
+    },
+    {
+      "epoch": 4.564612326043737,
+      "grad_norm": 0.0928633771471017,
+      "learning_rate": 1.8344297439980475e-06,
+      "loss": 0.4015,
+      "step": 2296
+    },
+    {
+      "epoch": 4.566600397614314,
+      "grad_norm": 0.09530337273078132,
+      "learning_rate": 1.81784298224712e-06,
+      "loss": 0.393,
+      "step": 2297
+    },
+    {
+      "epoch": 4.568588469184891,
+      "grad_norm": 0.09252969398882711,
+      "learning_rate": 1.8013298057897932e-06,
+      "loss": 0.3896,
+      "step": 2298
+    },
+    {
+      "epoch": 4.570576540755467,
+      "grad_norm": 0.09219676491217316,
+      "learning_rate": 1.7848902464505303e-06,
+      "loss": 0.3884,
+      "step": 2299
+    },
+    {
+      "epoch": 4.572564612326044,
+      "grad_norm": 0.09242046744560646,
+      "learning_rate": 1.7685243359119298e-06,
+      "loss": 0.3792,
+      "step": 2300
+    },
+    {
+      "epoch": 4.57455268389662,
+      "grad_norm": 0.08935728010398887,
+      "learning_rate": 1.7522321057146508e-06,
+      "loss": 0.3864,
+      "step": 2301
+    },
+    {
+      "epoch": 4.576540755467197,
+      "grad_norm": 0.0931849223946967,
+      "learning_rate": 1.7360135872573502e-06,
+      "loss": 0.3801,
+      "step": 2302
+    },
+    {
+      "epoch": 4.578528827037774,
+      "grad_norm": 0.09107346999398934,
+      "learning_rate": 1.71986881179663e-06,
+      "loss": 0.3974,
+      "step": 2303
+    },
+    {
+      "epoch": 4.58051689860835,
+      "grad_norm": 0.09210780909523932,
+      "learning_rate": 1.70379781044697e-06,
+      "loss": 0.3834,
+      "step": 2304
+    },
+    {
+      "epoch": 4.582504970178927,
+      "grad_norm": 0.0884785884728689,
+      "learning_rate": 1.6878006141806747e-06,
+      "loss": 0.397,
+      "step": 2305
+    },
+    {
+      "epoch": 4.584493041749503,
+      "grad_norm": 0.09331749408939408,
+      "learning_rate": 1.6718772538278117e-06,
+      "loss": 0.3915,
+      "step": 2306
+    },
+    {
+      "epoch": 4.58648111332008,
+      "grad_norm": 0.09223920546338787,
+      "learning_rate": 1.6560277600761487e-06,
+      "loss": 0.3816,
+      "step": 2307
+    },
+    {
+      "epoch": 4.588469184890656,
+      "grad_norm": 0.09681340642252376,
+      "learning_rate": 1.6402521634710877e-06,
+      "loss": 0.3785,
+      "step": 2308
+    },
+    {
+      "epoch": 4.5904572564612325,
+      "grad_norm": 0.09629143650589808,
+      "learning_rate": 1.6245504944156332e-06,
+      "loss": 0.3916,
+      "step": 2309
+    },
+    {
+      "epoch": 4.592445328031809,
+      "grad_norm": 0.09627909480747703,
+      "learning_rate": 1.608922783170299e-06,
+      "loss": 0.3833,
+      "step": 2310
+    },
+    {
+      "epoch": 4.594433399602385,
+      "grad_norm": 0.08841171190126924,
+      "learning_rate": 1.5933690598530737e-06,
+      "loss": 0.3752,
+      "step": 2311
+    },
+    {
+      "epoch": 4.596421471172962,
+      "grad_norm": 0.09050689164750318,
+      "learning_rate": 1.577889354439357e-06,
+      "loss": 0.3826,
+      "step": 2312
+    },
+    {
+      "epoch": 4.598409542743539,
+      "grad_norm": 0.08629910372587375,
+      "learning_rate": 1.5624836967618939e-06,
+      "loss": 0.3769,
+      "step": 2313
+    },
+    {
+      "epoch": 4.600397614314115,
+      "grad_norm": 0.09722776788968594,
+      "learning_rate": 1.547152116510726e-06,
+      "loss": 0.3923,
+      "step": 2314
+    },
+    {
+      "epoch": 4.602385685884692,
+      "grad_norm": 0.09237377126717078,
+      "learning_rate": 1.5318946432331338e-06,
+      "loss": 0.3878,
+      "step": 2315
+    },
+    {
+      "epoch": 4.604373757455268,
+      "grad_norm": 0.08858713383649396,
+      "learning_rate": 1.5167113063335736e-06,
+      "loss": 0.3945,
+      "step": 2316
+    },
+    {
+      "epoch": 4.606361829025845,
+      "grad_norm": 0.10149378277319007,
+      "learning_rate": 1.5016021350736253e-06,
+      "loss": 0.394,
+      "step": 2317
+    },
+    {
+      "epoch": 4.608349900596421,
+      "grad_norm": 0.09370047111339964,
+      "learning_rate": 1.4865671585719477e-06,
+      "loss": 0.3791,
+      "step": 2318
+    },
+    {
+      "epoch": 4.610337972166998,
+      "grad_norm": 0.10062037398190488,
+      "learning_rate": 1.4716064058041846e-06,
+      "loss": 0.3834,
+      "step": 2319
+    },
+    {
+      "epoch": 4.612326043737575,
+      "grad_norm": 0.08806754588403382,
+      "learning_rate": 1.4567199056029613e-06,
+      "loss": 0.3972,
+      "step": 2320
+    },
+    {
+      "epoch": 4.614314115308151,
+      "grad_norm": 0.08442134067610135,
+      "learning_rate": 1.4419076866577864e-06,
+      "loss": 0.3748,
+      "step": 2321
+    },
+    {
+      "epoch": 4.616302186878728,
+      "grad_norm": 0.08796926416731948,
+      "learning_rate": 1.4271697775150162e-06,
+      "loss": 0.3834,
+      "step": 2322
+    },
+    {
+      "epoch": 4.618290258449305,
+      "grad_norm": 0.0910818428968802,
+      "learning_rate": 1.4125062065777973e-06,
+      "loss": 0.3827,
+      "step": 2323
+    },
+    {
+      "epoch": 4.620278330019881,
+      "grad_norm": 0.09065119233786946,
+      "learning_rate": 1.397917002106013e-06,
+      "loss": 0.3881,
+      "step": 2324
+    },
+    {
+      "epoch": 4.6222664015904575,
+      "grad_norm": 0.09104282484132259,
+      "learning_rate": 1.3834021922162167e-06,
+      "loss": 0.4014,
+      "step": 2325
+    },
+    {
+      "epoch": 4.6242544731610336,
+      "grad_norm": 0.09429251418317382,
+      "learning_rate": 1.3689618048815966e-06,
+      "loss": 0.3883,
+      "step": 2326
+    },
+    {
+      "epoch": 4.6262425447316105,
+      "grad_norm": 0.09749806535855796,
+      "learning_rate": 1.3545958679319093e-06,
+      "loss": 0.3879,
+      "step": 2327
+    },
+    {
+      "epoch": 4.6282306163021865,
+      "grad_norm": 0.09057348522000637,
+      "learning_rate": 1.3403044090534301e-06,
+      "loss": 0.3911,
+      "step": 2328
+    },
+    {
+      "epoch": 4.630218687872763,
+      "grad_norm": 0.08893250766924893,
+      "learning_rate": 1.3260874557889002e-06,
+      "loss": 0.4004,
+      "step": 2329
+    },
+    {
+      "epoch": 4.63220675944334,
+      "grad_norm": 0.08812327485932064,
+      "learning_rate": 1.3119450355374651e-06,
+      "loss": 0.3871,
+      "step": 2330
+    },
+    {
+      "epoch": 4.634194831013916,
+      "grad_norm": 0.08984430375875535,
+      "learning_rate": 1.2978771755546382e-06,
+      "loss": 0.389,
+      "step": 2331
+    },
+    {
+      "epoch": 4.636182902584493,
+      "grad_norm": 0.11238909857144332,
+      "learning_rate": 1.2838839029522344e-06,
+      "loss": 0.3882,
+      "step": 2332
+    },
+    {
+      "epoch": 4.63817097415507,
+      "grad_norm": 0.08542870167083381,
+      "learning_rate": 1.2699652446983213e-06,
+      "loss": 0.3998,
+      "step": 2333
+    },
+    {
+      "epoch": 4.640159045725646,
+      "grad_norm": 0.09072455538699621,
+      "learning_rate": 1.2561212276171753e-06,
+      "loss": 0.3906,
+      "step": 2334
+    },
+    {
+      "epoch": 4.642147117296223,
+      "grad_norm": 0.09077176680951171,
+      "learning_rate": 1.242351878389214e-06,
+      "loss": 0.399,
+      "step": 2335
+    },
+    {
+      "epoch": 4.644135188866799,
+      "grad_norm": 0.09087377061392632,
+      "learning_rate": 1.228657223550962e-06,
+      "loss": 0.3775,
+      "step": 2336
+    },
+    {
+      "epoch": 4.646123260437376,
+      "grad_norm": 0.08493200594804953,
+      "learning_rate": 1.2150372894949869e-06,
+      "loss": 0.3984,
+      "step": 2337
+    },
+    {
+      "epoch": 4.648111332007952,
+      "grad_norm": 0.08817667865233096,
+      "learning_rate": 1.2014921024698523e-06,
+      "loss": 0.3922,
+      "step": 2338
+    },
+    {
+      "epoch": 4.650099403578529,
+      "grad_norm": 0.08927232121946307,
+      "learning_rate": 1.1880216885800676e-06,
+      "loss": 0.3971,
+      "step": 2339
+    },
+    {
+      "epoch": 4.652087475149106,
+      "grad_norm": 0.08613839924431113,
+      "learning_rate": 1.1746260737860449e-06,
+      "loss": 0.3983,
+      "step": 2340
+    },
+    {
+      "epoch": 4.654075546719682,
+      "grad_norm": 0.0918930802800742,
+      "learning_rate": 1.1613052839040395e-06,
+      "loss": 0.3911,
+      "step": 2341
+    },
+    {
+      "epoch": 4.656063618290259,
+      "grad_norm": 0.08905521353173235,
+      "learning_rate": 1.148059344606094e-06,
+      "loss": 0.3769,
+      "step": 2342
+    },
+    {
+      "epoch": 4.658051689860835,
+      "grad_norm": 0.08840020767025152,
+      "learning_rate": 1.1348882814200103e-06,
+      "loss": 0.3869,
+      "step": 2343
+    },
+    {
+      "epoch": 4.6600397614314115,
+      "grad_norm": 0.09020713218255047,
+      "learning_rate": 1.1217921197292836e-06,
+      "loss": 0.3808,
+      "step": 2344
+    },
+    {
+      "epoch": 4.662027833001988,
+      "grad_norm": 0.08584989452622944,
+      "learning_rate": 1.1087708847730494e-06,
+      "loss": 0.387,
+      "step": 2345
+    },
+    {
+      "epoch": 4.664015904572564,
+      "grad_norm": 0.08782265773096049,
+      "learning_rate": 1.095824601646056e-06,
+      "loss": 0.3805,
+      "step": 2346
+    },
+    {
+      "epoch": 4.666003976143141,
+      "grad_norm": 0.0856685513359666,
+      "learning_rate": 1.0829532952985987e-06,
+      "loss": 0.3816,
+      "step": 2347
+    },
+    {
+      "epoch": 4.667992047713717,
+      "grad_norm": 0.09328575113682944,
+      "learning_rate": 1.0701569905364706e-06,
+      "loss": 0.3886,
+      "step": 2348
+    },
+    {
+      "epoch": 4.669980119284294,
+      "grad_norm": 0.09334769831602094,
+      "learning_rate": 1.0574357120209266e-06,
+      "loss": 0.388,
+      "step": 2349
+    },
+    {
+      "epoch": 4.671968190854871,
+      "grad_norm": 0.09183604939307251,
+      "learning_rate": 1.0447894842686269e-06,
+      "loss": 0.3854,
+      "step": 2350
+    },
+    {
+      "epoch": 4.673956262425447,
+      "grad_norm": 0.08849811025639472,
+      "learning_rate": 1.0322183316515955e-06,
+      "loss": 0.3928,
+      "step": 2351
+    },
+    {
+      "epoch": 4.675944333996024,
+      "grad_norm": 0.0861408438176817,
+      "learning_rate": 1.0197222783971728e-06,
+      "loss": 0.382,
+      "step": 2352
+    },
+    {
+      "epoch": 4.6779324055666,
+      "grad_norm": 0.09179102020068529,
+      "learning_rate": 1.0073013485879523e-06,
+      "loss": 0.3924,
+      "step": 2353
+    },
+    {
+      "epoch": 4.679920477137177,
+      "grad_norm": 0.08985169454264477,
+      "learning_rate": 9.949555661617638e-07,
+      "loss": 0.3911,
+      "step": 2354
+    },
+    {
+      "epoch": 4.681908548707754,
+      "grad_norm": 0.08712350247200025,
+      "learning_rate": 9.826849549116103e-07,
+      "loss": 0.3851,
+      "step": 2355
+    },
+    {
+      "epoch": 4.68389662027833,
+      "grad_norm": 0.08603921934702671,
+      "learning_rate": 9.704895384856106e-07,
+      "loss": 0.39,
+      "step": 2356
+    },
+    {
+      "epoch": 4.685884691848907,
+      "grad_norm": 0.08904593449849948,
+      "learning_rate": 9.583693403869864e-07,
+      "loss": 0.3821,
+      "step": 2357
+    },
+    {
+      "epoch": 4.687872763419483,
+      "grad_norm": 0.09273629581237089,
+      "learning_rate": 9.463243839739778e-07,
+      "loss": 0.382,
+      "step": 2358
+    },
+    {
+      "epoch": 4.68986083499006,
+      "grad_norm": 0.0926055652891825,
+      "learning_rate": 9.343546924598291e-07,
+      "loss": 0.3652,
+      "step": 2359
+    },
+    {
+      "epoch": 4.691848906560637,
+      "grad_norm": 0.08902119464039318,
+      "learning_rate": 9.224602889127321e-07,
+      "loss": 0.3839,
+      "step": 2360
+    },
+    {
+      "epoch": 4.693836978131213,
+      "grad_norm": 0.09461665383225337,
+      "learning_rate": 9.106411962557816e-07,
+      "loss": 0.3815,
+      "step": 2361
+    },
+    {
+      "epoch": 4.6958250497017895,
+      "grad_norm": 0.08978190990006801,
+      "learning_rate": 8.988974372669302e-07,
+      "loss": 0.3893,
+      "step": 2362
+    },
+    {
+      "epoch": 4.6978131212723655,
+      "grad_norm": 0.08480912747580577,
+      "learning_rate": 8.872290345789492e-07,
+      "loss": 0.3763,
+      "step": 2363
+    },
+    {
+      "epoch": 4.699801192842942,
+      "grad_norm": 0.08888694711926962,
+      "learning_rate": 8.756360106793793e-07,
+      "loss": 0.3827,
+      "step": 2364
+    },
+    {
+      "epoch": 4.701789264413518,
+      "grad_norm": 0.08768867258285498,
+      "learning_rate": 8.641183879104864e-07,
+      "loss": 0.3948,
+      "step": 2365
+    },
+    {
+      "epoch": 4.703777335984095,
+      "grad_norm": 0.0862141848794459,
+      "learning_rate": 8.526761884692259e-07,
+      "loss": 0.3924,
+      "step": 2366
+    },
+    {
+      "epoch": 4.705765407554672,
+      "grad_norm": 0.0892669179569011,
+      "learning_rate": 8.41309434407207e-07,
+      "loss": 0.389,
+      "step": 2367
+    },
+    {
+      "epoch": 4.707753479125248,
+      "grad_norm": 0.09566520976823947,
+      "learning_rate": 8.300181476306179e-07,
+      "loss": 0.3919,
+      "step": 2368
+    },
+    {
+      "epoch": 4.709741550695825,
+      "grad_norm": 0.08459337487960555,
+      "learning_rate": 8.188023499002206e-07,
+      "loss": 0.3838,
+      "step": 2369
+    },
+    {
+      "epoch": 4.711729622266402,
+      "grad_norm": 0.08665109130706619,
+      "learning_rate": 8.076620628312892e-07,
+      "loss": 0.3822,
+      "step": 2370
+    },
+    {
+      "epoch": 4.713717693836978,
+      "grad_norm": 0.09024375966314252,
+      "learning_rate": 7.965973078935696e-07,
+      "loss": 0.3813,
+      "step": 2371
+    },
+    {
+      "epoch": 4.715705765407555,
+      "grad_norm": 0.09285780165532485,
+      "learning_rate": 7.856081064112441e-07,
+      "loss": 0.3912,
+      "step": 2372
+    },
+    {
+      "epoch": 4.717693836978131,
+      "grad_norm": 0.0943616237445835,
+      "learning_rate": 7.746944795628874e-07,
+      "loss": 0.3862,
+      "step": 2373
+    },
+    {
+      "epoch": 4.719681908548708,
+      "grad_norm": 0.08785063174101017,
+      "learning_rate": 7.638564483814215e-07,
+      "loss": 0.3947,
+      "step": 2374
+    },
+    {
+      "epoch": 4.721669980119284,
+      "grad_norm": 0.09095838243764175,
+      "learning_rate": 7.530940337540848e-07,
+      "loss": 0.3871,
+      "step": 2375
+    },
+    {
+      "epoch": 4.723658051689861,
+      "grad_norm": 0.09266236841123172,
+      "learning_rate": 7.424072564223795e-07,
+      "loss": 0.3792,
+      "step": 2376
+    },
+    {
+      "epoch": 4.725646123260438,
+      "grad_norm": 0.08856421790524949,
+      "learning_rate": 7.317961369820481e-07,
+      "loss": 0.3984,
+      "step": 2377
+    },
+    {
+      "epoch": 4.727634194831014,
+      "grad_norm": 0.09046046436239621,
+      "learning_rate": 7.212606958830126e-07,
+      "loss": 0.388,
+      "step": 2378
+    },
+    {
+      "epoch": 4.729622266401591,
+      "grad_norm": 0.08810710235320644,
+      "learning_rate": 7.108009534293514e-07,
+      "loss": 0.3926,
+      "step": 2379
+    },
+    {
+      "epoch": 4.7316103379721675,
+      "grad_norm": 0.08776085645147964,
+      "learning_rate": 7.00416929779264e-07,
+      "loss": 0.3835,
+      "step": 2380
+    },
+    {
+      "epoch": 4.7335984095427435,
+      "grad_norm": 0.09018333553170645,
+      "learning_rate": 6.901086449450045e-07,
+      "loss": 0.3926,
+      "step": 2381
+    },
+    {
+      "epoch": 4.73558648111332,
+      "grad_norm": 0.08976241821965435,
+      "learning_rate": 6.798761187928815e-07,
+      "loss": 0.3824,
+      "step": 2382
+    },
+    {
+      "epoch": 4.737574552683896,
+      "grad_norm": 0.08687984662867637,
+      "learning_rate": 6.697193710431827e-07,
+      "loss": 0.3894,
+      "step": 2383
+    },
+    {
+      "epoch": 4.739562624254473,
+      "grad_norm": 0.08513429550866758,
+      "learning_rate": 6.596384212701656e-07,
+      "loss": 0.3889,
+      "step": 2384
+    },
+    {
+      "epoch": 4.741550695825049,
+      "grad_norm": 0.09283958366235047,
+      "learning_rate": 6.496332889020096e-07,
+      "loss": 0.3917,
+      "step": 2385
+    },
+    {
+      "epoch": 4.743538767395626,
+      "grad_norm": 0.08978086123224327,
+      "learning_rate": 6.397039932207705e-07,
+      "loss": 0.3874,
+      "step": 2386
+    },
+    {
+      "epoch": 4.745526838966203,
+      "grad_norm": 0.09028630878201256,
+      "learning_rate": 6.298505533623544e-07,
+      "loss": 0.3979,
+      "step": 2387
+    },
+    {
+      "epoch": 4.747514910536779,
+      "grad_norm": 0.08511974572102049,
+      "learning_rate": 6.200729883164736e-07,
+      "loss": 0.3868,
+      "step": 2388
+    },
+    {
+      "epoch": 4.749502982107356,
+      "grad_norm": 0.09014999836299889,
+      "learning_rate": 6.10371316926619e-07,
+      "loss": 0.3832,
+      "step": 2389
+    },
+    {
+      "epoch": 4.751491053677933,
+      "grad_norm": 0.08562371944420073,
+      "learning_rate": 6.007455578900079e-07,
+      "loss": 0.3835,
+      "step": 2390
+    },
+    {
+      "epoch": 4.753479125248509,
+      "grad_norm": 0.08437078505374138,
+      "learning_rate": 5.911957297575743e-07,
+      "loss": 0.3995,
+      "step": 2391
+    },
+    {
+      "epoch": 4.755467196819086,
+      "grad_norm": 0.08509023606898676,
+      "learning_rate": 5.817218509339029e-07,
+      "loss": 0.382,
+      "step": 2392
+    },
+    {
+      "epoch": 4.757455268389662,
+      "grad_norm": 0.0862939742400007,
+      "learning_rate": 5.723239396772107e-07,
+      "loss": 0.3751,
+      "step": 2393
+    },
+    {
+      "epoch": 4.759443339960239,
+      "grad_norm": 0.08427683497565448,
+      "learning_rate": 5.630020140993076e-07,
+      "loss": 0.3906,
+      "step": 2394
+    },
+    {
+      "epoch": 4.761431411530815,
+      "grad_norm": 0.08659925642974421,
+      "learning_rate": 5.537560921655694e-07,
+      "loss": 0.3739,
+      "step": 2395
+    },
+    {
+      "epoch": 4.763419483101392,
+      "grad_norm": 0.08684889514283894,
+      "learning_rate": 5.445861916948891e-07,
+      "loss": 0.3883,
+      "step": 2396
+    },
+    {
+      "epoch": 4.7654075546719685,
+      "grad_norm": 0.08586227981413348,
+      "learning_rate": 5.354923303596593e-07,
+      "loss": 0.3882,
+      "step": 2397
+    },
+    {
+      "epoch": 4.7673956262425445,
+      "grad_norm": 0.08933801071830935,
+      "learning_rate": 5.26474525685714e-07,
+      "loss": 0.3921,
+      "step": 2398
+    },
+    {
+      "epoch": 4.769383697813121,
+      "grad_norm": 0.08393331580601653,
+      "learning_rate": 5.175327950523201e-07,
+      "loss": 0.3823,
+      "step": 2399
+    },
+    {
+      "epoch": 4.7713717693836974,
+      "grad_norm": 0.08312263829480795,
+      "learning_rate": 5.086671556921286e-07,
+      "loss": 0.3914,
+      "step": 2400
+    },
+    {
+      "epoch": 4.773359840954274,
+      "grad_norm": 0.08583116775976779,
+      "learning_rate": 4.998776246911563e-07,
+      "loss": 0.3879,
+      "step": 2401
+    },
+    {
+      "epoch": 4.775347912524851,
+      "grad_norm": 0.08780350297111986,
+      "learning_rate": 4.911642189887289e-07,
+      "loss": 0.3822,
+      "step": 2402
+    },
+    {
+      "epoch": 4.777335984095427,
+      "grad_norm": 0.08554313989570071,
+      "learning_rate": 4.82526955377467e-07,
+      "loss": 0.3767,
+      "step": 2403
+    },
+    {
+      "epoch": 4.779324055666004,
+      "grad_norm": 0.08765985124847109,
+      "learning_rate": 4.7396585050325074e-07,
+      "loss": 0.3847,
+      "step": 2404
+    },
+    {
+      "epoch": 4.78131212723658,
+      "grad_norm": 0.08780793267342793,
+      "learning_rate": 4.654809208651889e-07,
+      "loss": 0.3844,
+      "step": 2405
+    },
+    {
+      "epoch": 4.783300198807157,
+      "grad_norm": 0.08604341225517562,
+      "learning_rate": 4.5707218281557883e-07,
+      "loss": 0.393,
+      "step": 2406
+    },
+    {
+      "epoch": 4.785288270377734,
+      "grad_norm": 0.08308372988525528,
+      "learning_rate": 4.487396525598797e-07,
+      "loss": 0.3913,
+      "step": 2407
+    },
+    {
+      "epoch": 4.78727634194831,
+      "grad_norm": 0.08542634110771802,
+      "learning_rate": 4.4048334615668156e-07,
+      "loss": 0.4001,
+      "step": 2408
+    },
+    {
+      "epoch": 4.789264413518887,
+      "grad_norm": 0.08706213681011674,
+      "learning_rate": 4.323032795176785e-07,
+      "loss": 0.382,
+      "step": 2409
+    },
+    {
+      "epoch": 4.791252485089463,
+      "grad_norm": 0.0859838895047276,
+      "learning_rate": 4.241994684076378e-07,
+      "loss": 0.3961,
+      "step": 2410
+    },
+    {
+      "epoch": 4.79324055666004,
+      "grad_norm": 0.08596617254705141,
+      "learning_rate": 4.1617192844435105e-07,
+      "loss": 0.3814,
+      "step": 2411
+    },
+    {
+      "epoch": 4.795228628230617,
+      "grad_norm": 0.09187239171244084,
+      "learning_rate": 4.0822067509863397e-07,
+      "loss": 0.3964,
+      "step": 2412
+    },
+    {
+      "epoch": 4.797216699801193,
+      "grad_norm": 0.09151958557517886,
+      "learning_rate": 4.0034572369427315e-07,
+      "loss": 0.3825,
+      "step": 2413
+    },
+    {
+      "epoch": 4.79920477137177,
+      "grad_norm": 0.08886486817433736,
+      "learning_rate": 3.9254708940800855e-07,
+      "loss": 0.3905,
+      "step": 2414
+    },
+    {
+      "epoch": 4.801192842942346,
+      "grad_norm": 0.08799982211453036,
+      "learning_rate": 3.848247872694977e-07,
+      "loss": 0.3881,
+      "step": 2415
+    },
+    {
+      "epoch": 4.8031809145129225,
+      "grad_norm": 0.0902921148546499,
+      "learning_rate": 3.7717883216128904e-07,
+      "loss": 0.3855,
+      "step": 2416
+    },
+    {
+      "epoch": 4.805168986083499,
+      "grad_norm": 0.08950875599619121,
+      "learning_rate": 3.6960923881879994e-07,
+      "loss": 0.3997,
+      "step": 2417
+    },
+    {
+      "epoch": 4.807157057654075,
+      "grad_norm": 0.08405639164843014,
+      "learning_rate": 3.6211602183027216e-07,
+      "loss": 0.3827,
+      "step": 2418
+    },
+    {
+      "epoch": 4.809145129224652,
+      "grad_norm": 0.08385315321218918,
+      "learning_rate": 3.5469919563676735e-07,
+      "loss": 0.3853,
+      "step": 2419
+    },
+    {
+      "epoch": 4.811133200795228,
+      "grad_norm": 0.08505559287027027,
+      "learning_rate": 3.4735877453210944e-07,
+      "loss": 0.3839,
+      "step": 2420
+    },
+    {
+      "epoch": 4.813121272365805,
+      "grad_norm": 0.0877458325572006,
+      "learning_rate": 3.400947726628845e-07,
+      "loss": 0.3889,
+      "step": 2421
+    },
+    {
+      "epoch": 4.815109343936381,
+      "grad_norm": 0.08379392354212882,
+      "learning_rate": 3.329072040284009e-07,
+      "loss": 0.3813,
+      "step": 2422
+    },
+    {
+      "epoch": 4.817097415506958,
+      "grad_norm": 0.08663727574036406,
+      "learning_rate": 3.257960824806583e-07,
+      "loss": 0.3724,
+      "step": 2423
+    },
+    {
+      "epoch": 4.819085487077535,
+      "grad_norm": 0.08972288136722377,
+      "learning_rate": 3.187614217243296e-07,
+      "loss": 0.3904,
+      "step": 2424
+    },
+    {
+      "epoch": 4.821073558648111,
+      "grad_norm": 0.08914702376767991,
+      "learning_rate": 3.1180323531673e-07,
+      "loss": 0.3902,
+      "step": 2425
+    },
+    {
+      "epoch": 4.823061630218688,
+      "grad_norm": 0.08226333159404381,
+      "learning_rate": 3.049215366677949e-07,
+      "loss": 0.3919,
+      "step": 2426
+    },
+    {
+      "epoch": 4.825049701789265,
+      "grad_norm": 0.08530573163134422,
+      "learning_rate": 2.9811633904004876e-07,
+      "loss": 0.3813,
+      "step": 2427
+    },
+    {
+      "epoch": 4.827037773359841,
+      "grad_norm": 0.08624288085838655,
+      "learning_rate": 2.913876555485873e-07,
+      "loss": 0.3863,
+      "step": 2428
+    },
+    {
+      "epoch": 4.829025844930418,
+      "grad_norm": 0.08814860299045851,
+      "learning_rate": 2.8473549916103295e-07,
+      "loss": 0.3923,
+      "step": 2429
+    },
+    {
+      "epoch": 4.831013916500994,
+      "grad_norm": 0.0834424995172425,
+      "learning_rate": 2.7815988269753957e-07,
+      "loss": 0.3816,
+      "step": 2430
+    },
+    {
+      "epoch": 4.833001988071571,
+      "grad_norm": 0.0875951945007766,
+      "learning_rate": 2.7166081883074793e-07,
+      "loss": 0.3831,
+      "step": 2431
+    },
+    {
+      "epoch": 4.834990059642147,
+      "grad_norm": 0.08573228313795739,
+      "learning_rate": 2.6523832008575447e-07,
+      "loss": 0.3797,
+      "step": 2432
+    },
+    {
+      "epoch": 4.836978131212724,
+      "grad_norm": 0.08471682476679154,
+      "learning_rate": 2.5889239884011594e-07,
+      "loss": 0.386,
+      "step": 2433
+    },
+    {
+      "epoch": 4.8389662027833005,
+      "grad_norm": 0.08533182226317534,
+      "learning_rate": 2.526230673237873e-07,
+      "loss": 0.3911,
+      "step": 2434
+    },
+    {
+      "epoch": 4.8409542743538765,
+      "grad_norm": 0.08729608897177847,
+      "learning_rate": 2.46430337619139e-07,
+      "loss": 0.3933,
+      "step": 2435
+    },
+    {
+      "epoch": 4.842942345924453,
+      "grad_norm": 0.08464771353563329,
+      "learning_rate": 2.4031422166090447e-07,
+      "loss": 0.3877,
+      "step": 2436
+    },
+    {
+      "epoch": 4.84493041749503,
+      "grad_norm": 0.08529987456863733,
+      "learning_rate": 2.3427473123615263e-07,
+      "loss": 0.3872,
+      "step": 2437
+    },
+    {
+      "epoch": 4.846918489065606,
+      "grad_norm": 0.08251198483736942,
+      "learning_rate": 2.2831187798430633e-07,
+      "loss": 0.3851,
+      "step": 2438
+    },
+    {
+      "epoch": 4.848906560636183,
+      "grad_norm": 0.08697926770346306,
+      "learning_rate": 2.2242567339706645e-07,
+      "loss": 0.3891,
+      "step": 2439
+    },
+    {
+      "epoch": 4.850894632206759,
+      "grad_norm": 0.08389709541673776,
+      "learning_rate": 2.1661612881842542e-07,
+      "loss": 0.3974,
+      "step": 2440
+    },
+    {
+      "epoch": 4.852882703777336,
+      "grad_norm": 0.08176299785081795,
+      "learning_rate": 2.108832554446405e-07,
+      "loss": 0.3865,
+      "step": 2441
+    },
+    {
+      "epoch": 4.854870775347912,
+      "grad_norm": 0.0859453841790713,
+      "learning_rate": 2.0522706432419382e-07,
+      "loss": 0.3894,
+      "step": 2442
+    },
+    {
+      "epoch": 4.856858846918489,
+      "grad_norm": 0.09171848286661113,
+      "learning_rate": 1.9964756635780125e-07,
+      "loss": 0.388,
+      "step": 2443
+    },
+    {
+      "epoch": 4.858846918489066,
+      "grad_norm": 0.08974957511489433,
+      "learning_rate": 1.9414477229835472e-07,
+      "loss": 0.3798,
+      "step": 2444
+    },
+    {
+      "epoch": 4.860834990059642,
+      "grad_norm": 0.08912058067697651,
+      "learning_rate": 1.887186927509399e-07,
+      "loss": 0.3844,
+      "step": 2445
+    },
+    {
+      "epoch": 4.862823061630219,
+      "grad_norm": 0.08776737055560599,
+      "learning_rate": 1.8336933817278746e-07,
+      "loss": 0.3827,
+      "step": 2446
+    },
+    {
+      "epoch": 4.864811133200796,
+      "grad_norm": 0.08734208552490895,
+      "learning_rate": 1.780967188732552e-07,
+      "loss": 0.3988,
+      "step": 2447
+    },
+    {
+      "epoch": 4.866799204771372,
+      "grad_norm": 0.08473457345009698,
+      "learning_rate": 1.7290084501383254e-07,
+      "loss": 0.39,
+      "step": 2448
+    },
+    {
+      "epoch": 4.868787276341949,
+      "grad_norm": 0.08742974713015055,
+      "learning_rate": 1.6778172660809167e-07,
+      "loss": 0.3841,
+      "step": 2449
+    },
+    {
+      "epoch": 4.870775347912525,
+      "grad_norm": 0.08420030904113016,
+      "learning_rate": 1.627393735216787e-07,
+      "loss": 0.3871,
+      "step": 2450
+    },
+    {
+      "epoch": 4.8727634194831015,
+      "grad_norm": 0.0862672824125834,
+      "learning_rate": 1.5777379547230908e-07,
+      "loss": 0.3875,
+      "step": 2451
+    },
+    {
+      "epoch": 4.8747514910536776,
+      "grad_norm": 0.08692886838458265,
+      "learning_rate": 1.5288500202972346e-07,
+      "loss": 0.3726,
+      "step": 2452
+    },
+    {
+      "epoch": 4.8767395626242545,
+      "grad_norm": 0.08351633165920348,
+      "learning_rate": 1.4807300261568735e-07,
+      "loss": 0.3844,
+      "step": 2453
+    },
+    {
+      "epoch": 4.878727634194831,
+      "grad_norm": 0.08207322266636974,
+      "learning_rate": 1.4333780650396922e-07,
+      "loss": 0.3836,
+      "step": 2454
+    },
+    {
+      "epoch": 4.880715705765407,
+      "grad_norm": 0.08636151753971841,
+      "learning_rate": 1.3867942282031365e-07,
+      "loss": 0.3814,
+      "step": 2455
+    },
+    {
+      "epoch": 4.882703777335984,
+      "grad_norm": 0.08372475407996112,
+      "learning_rate": 1.3409786054244144e-07,
+      "loss": 0.3864,
+      "step": 2456
+    },
+    {
+      "epoch": 4.88469184890656,
+      "grad_norm": 0.08656783681493549,
+      "learning_rate": 1.295931285000096e-07,
+      "loss": 0.3936,
+      "step": 2457
+    },
+    {
+      "epoch": 4.886679920477137,
+      "grad_norm": 0.0865359135074539,
+      "learning_rate": 1.2516523537462023e-07,
+      "loss": 0.3854,
+      "step": 2458
+    },
+    {
+      "epoch": 4.888667992047714,
+      "grad_norm": 0.08370382158458198,
+      "learning_rate": 1.208141896997761e-07,
+      "loss": 0.3874,
+      "step": 2459
+    },
+    {
+      "epoch": 4.89065606361829,
+      "grad_norm": 0.08898094046346416,
+      "learning_rate": 1.1653999986089404e-07,
+      "loss": 0.3816,
+      "step": 2460
+    },
+    {
+      "epoch": 4.892644135188867,
+      "grad_norm": 0.08630171969853209,
+      "learning_rate": 1.1234267409525601e-07,
+      "loss": 0.3923,
+      "step": 2461
+    },
+    {
+      "epoch": 4.894632206759443,
+      "grad_norm": 0.0822988617920718,
+      "learning_rate": 1.0822222049202247e-07,
+      "loss": 0.3764,
+      "step": 2462
+    },
+    {
+      "epoch": 4.89662027833002,
+      "grad_norm": 0.08320129070038762,
+      "learning_rate": 1.0417864699220126e-07,
+      "loss": 0.3761,
+      "step": 2463
+    },
+    {
+      "epoch": 4.898608349900597,
+      "grad_norm": 0.0842454627956477,
+      "learning_rate": 1.0021196138863432e-07,
+      "loss": 0.3837,
+      "step": 2464
+    },
+    {
+      "epoch": 4.900596421471173,
+      "grad_norm": 0.08427608839565273,
+      "learning_rate": 9.632217132598431e-08,
+      "loss": 0.3962,
+      "step": 2465
+    },
+    {
+      "epoch": 4.90258449304175,
+      "grad_norm": 0.08425550035154374,
+      "learning_rate": 9.250928430072138e-08,
+      "loss": 0.3826,
+      "step": 2466
+    },
+    {
+      "epoch": 4.904572564612326,
+      "grad_norm": 0.0870394245800322,
+      "learning_rate": 8.877330766110526e-08,
+      "loss": 0.3899,
+      "step": 2467
+    },
+    {
+      "epoch": 4.906560636182903,
+      "grad_norm": 0.08676508699928365,
+      "learning_rate": 8.51142486071721e-08,
+      "loss": 0.3924,
+      "step": 2468
+    },
+    {
+      "epoch": 4.9085487077534795,
+      "grad_norm": 0.08646978372665866,
+      "learning_rate": 8.153211419072104e-08,
+      "loss": 0.3844,
+      "step": 2469
+    },
+    {
+      "epoch": 4.9105367793240555,
+      "grad_norm": 0.08355087037223606,
+      "learning_rate": 7.802691131530093e-08,
+      "loss": 0.3936,
+      "step": 2470
+    },
+    {
+      "epoch": 4.912524850894632,
+      "grad_norm": 0.0825099428809249,
+      "learning_rate": 7.459864673619698e-08,
+      "loss": 0.3844,
+      "step": 2471
+    },
+    {
+      "epoch": 4.914512922465208,
+      "grad_norm": 0.0874220543050281,
+      "learning_rate": 7.124732706042636e-08,
+      "loss": 0.3839,
+      "step": 2472
+    },
+    {
+      "epoch": 4.916500994035785,
+      "grad_norm": 0.08504844158831366,
+      "learning_rate": 6.797295874669818e-08,
+      "loss": 0.3775,
+      "step": 2473
+    },
+    {
+      "epoch": 4.918489065606362,
+      "grad_norm": 0.08380430885759332,
+      "learning_rate": 6.47755481054313e-08,
+      "loss": 0.3864,
+      "step": 2474
+    },
+    {
+      "epoch": 4.920477137176938,
+      "grad_norm": 0.08542485101299789,
+      "learning_rate": 6.165510129873653e-08,
+      "loss": 0.3831,
+      "step": 2475
+    },
+    {
+      "epoch": 4.922465208747515,
+      "grad_norm": 0.08407973206991785,
+      "learning_rate": 5.8611624340385585e-08,
+      "loss": 0.3918,
+      "step": 2476
+    },
+    {
+      "epoch": 4.924453280318091,
+      "grad_norm": 0.08941652662215563,
+      "learning_rate": 5.564512309581993e-08,
+      "loss": 0.3857,
+      "step": 2477
+    },
+    {
+      "epoch": 4.926441351888668,
+      "grad_norm": 0.08476705011477939,
+      "learning_rate": 5.2755603282128585e-08,
+      "loss": 0.3996,
+      "step": 2478
+    },
+    {
+      "epoch": 4.928429423459244,
+      "grad_norm": 0.08424740037805077,
+      "learning_rate": 4.994307046804814e-08,
+      "loss": 0.4005,
+      "step": 2479
+    },
+    {
+      "epoch": 4.930417495029821,
+      "grad_norm": 0.08089692105587475,
+      "learning_rate": 4.720753007393164e-08,
+      "loss": 0.3856,
+      "step": 2480
+    },
+    {
+      "epoch": 4.932405566600398,
+      "grad_norm": 0.0828129028116285,
+      "learning_rate": 4.454898737176194e-08,
+      "loss": 0.3896,
+      "step": 2481
+    },
+    {
+      "epoch": 4.934393638170974,
+      "grad_norm": 0.08581140480052221,
+      "learning_rate": 4.1967447485129486e-08,
+      "loss": 0.3877,
+      "step": 2482
+    },
+    {
+      "epoch": 4.936381709741551,
+      "grad_norm": 0.0841501559497828,
+      "learning_rate": 3.946291538921454e-08,
+      "loss": 0.3932,
+      "step": 2483
+    },
+    {
+      "epoch": 4.938369781312128,
+      "grad_norm": 0.084945831920411,
+      "learning_rate": 3.703539591080052e-08,
+      "loss": 0.3893,
+      "step": 2484
+    },
+    {
+      "epoch": 4.940357852882704,
+      "grad_norm": 0.08602665475516638,
+      "learning_rate": 3.468489372823847e-08,
+      "loss": 0.3857,
+      "step": 2485
+    },
+    {
+      "epoch": 4.942345924453281,
+      "grad_norm": 0.0843351047913454,
+      "learning_rate": 3.2411413371460364e-08,
+      "loss": 0.3944,
+      "step": 2486
+    },
+    {
+      "epoch": 4.944333996023857,
+      "grad_norm": 0.09458575400648904,
+      "learning_rate": 3.02149592219525e-08,
+      "loss": 0.371,
+      "step": 2487
+    },
+    {
+      "epoch": 4.9463220675944335,
+      "grad_norm": 0.09060041810386756,
+      "learning_rate": 2.8095535512759898e-08,
+      "loss": 0.3823,
+      "step": 2488
+    },
+    {
+      "epoch": 4.9483101391650095,
+      "grad_norm": 0.08243819150909268,
+      "learning_rate": 2.605314632846856e-08,
+      "loss": 0.3857,
+      "step": 2489
+    },
+    {
+      "epoch": 4.950298210735586,
+      "grad_norm": 0.08673906431689675,
+      "learning_rate": 2.4087795605209906e-08,
+      "loss": 0.3857,
+      "step": 2490
+    },
+    {
+      "epoch": 4.952286282306163,
+      "grad_norm": 0.18285784263478053,
+      "learning_rate": 2.2199487130638575e-08,
+      "loss": 0.3808,
+      "step": 2491
+    },
+    {
+      "epoch": 4.954274353876739,
+      "grad_norm": 0.08505699269511605,
+      "learning_rate": 2.038822454393241e-08,
+      "loss": 0.3919,
+      "step": 2492
+    },
+    {
+      "epoch": 4.956262425447316,
+      "grad_norm": 0.08105600682700584,
+      "learning_rate": 1.8654011335788037e-08,
+      "loss": 0.388,
+      "step": 2493
+    },
+    {
+      "epoch": 4.958250497017893,
+      "grad_norm": 0.08527795126915254,
+      "learning_rate": 1.699685084840752e-08,
+      "loss": 0.384,
+      "step": 2494
+    },
+    {
+      "epoch": 4.960238568588469,
+      "grad_norm": 0.08653563160420075,
+      "learning_rate": 1.5416746275507267e-08,
+      "loss": 0.3962,
+      "step": 2495
+    },
+    {
+      "epoch": 4.962226640159046,
+      "grad_norm": 0.08979740263573663,
+      "learning_rate": 1.3913700662282482e-08,
+      "loss": 0.3914,
+      "step": 2496
+    },
+    {
+      "epoch": 4.964214711729622,
+      "grad_norm": 0.08523863622899382,
+      "learning_rate": 1.2487716905429382e-08,
+      "loss": 0.3787,
+      "step": 2497
+    },
+    {
+      "epoch": 4.966202783300199,
+      "grad_norm": 0.08508629391165508,
+      "learning_rate": 1.1138797753131868e-08,
+      "loss": 0.3885,
+      "step": 2498
+    },
+    {
+      "epoch": 4.968190854870775,
+      "grad_norm": 0.08532887882999854,
+      "learning_rate": 9.866945805048211e-09,
+      "loss": 0.3769,
+      "step": 2499
+    },
+    {
+      "epoch": 4.970178926441352,
+      "grad_norm": 0.08378156608125406,
+      "learning_rate": 8.672163512311039e-09,
+      "loss": 0.3835,
+      "step": 2500
+    },
+    {
+      "epoch": 4.972166998011929,
+      "grad_norm": 0.08684442546437508,
+      "learning_rate": 7.554453177522902e-09,
+      "loss": 0.389,
+      "step": 2501
+    },
+    {
+      "epoch": 4.974155069582505,
+      "grad_norm": 0.0813736466741287,
+      "learning_rate": 6.513816954760721e-09,
+      "loss": 0.3744,
+      "step": 2502
+    },
+    {
+      "epoch": 4.976143141153082,
+      "grad_norm": 0.08714909210347142,
+      "learning_rate": 5.5502568495491295e-09,
+      "loss": 0.3766,
+      "step": 2503
+    },
+    {
+      "epoch": 4.9781312127236585,
+      "grad_norm": 0.08760221313835241,
+      "learning_rate": 4.663774718878245e-09,
+      "loss": 0.3885,
+      "step": 2504
+    },
+    {
+      "epoch": 4.980119284294235,
+      "grad_norm": 0.08699109801070153,
+      "learning_rate": 3.854372271194784e-09,
+      "loss": 0.3888,
+      "step": 2505
+    },
+    {
+      "epoch": 4.9821073558648115,
+      "grad_norm": 0.08377737770104218,
+      "learning_rate": 3.1220510663931836e-09,
+      "loss": 0.3841,
+      "step": 2506
+    },
+    {
+      "epoch": 4.9840954274353875,
+      "grad_norm": 0.0860928281202727,
+      "learning_rate": 2.4668125158111567e-09,
+      "loss": 0.3874,
+      "step": 2507
+    },
+    {
+      "epoch": 4.986083499005964,
+      "grad_norm": 0.08466195525302847,
+      "learning_rate": 1.8886578822385758e-09,
+      "loss": 0.3968,
+      "step": 2508
+    },
+    {
+      "epoch": 4.98807157057654,
+      "grad_norm": 0.08539612831774646,
+      "learning_rate": 1.3875882799041506e-09,
+      "loss": 0.389,
+      "step": 2509
+    },
+    {
+      "epoch": 4.990059642147117,
+      "grad_norm": 0.0871537888144173,
+      "learning_rate": 9.636046744754268e-10,
+      "loss": 0.3887,
+      "step": 2510
+    },
+    {
+      "epoch": 4.992047713717694,
+      "grad_norm": 0.09005357737681634,
+      "learning_rate": 6.167078830587869e-10,
+      "loss": 0.3828,
+      "step": 2511
+    },
+    {
+      "epoch": 4.99403578528827,
+      "grad_norm": 0.08342821477503674,
+      "learning_rate": 3.4689857420833193e-10,
+      "loss": 0.3765,
+      "step": 2512
+    },
+    {
+      "epoch": 4.996023856858847,
+      "grad_norm": 0.08551466138197758,
+      "learning_rate": 1.5417726789923594e-10,
+      "loss": 0.399,
+      "step": 2513
+    },
+    {
+      "epoch": 4.998011928429423,
+      "grad_norm": 0.08572188617520322,
+      "learning_rate": 3.8544335545509515e-11,
+      "loss": 0.3885,
+      "step": 2514
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.08448061468786694,
+      "learning_rate": 0.0,
+      "loss": 0.3759,
+      "step": 2515
+    },
+    {
+      "epoch": 5.0,
+      "step": 2515,
+      "total_flos": 4.219199777931264e+16,
+      "train_loss": 0.48938420348565576,
+      "train_runtime": 39399.5763,
+      "train_samples_per_second": 32.661,
+      "train_steps_per_second": 0.064
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2515,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.219199777931264e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}