diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,137685 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 19665,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.557192575617696,
+      "learning_rate": 3.3898305084745764e-08,
+      "loss": 0.8805,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.06872712778614,
+      "learning_rate": 6.779661016949153e-08,
+      "loss": 0.8498,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.14569355426566,
+      "learning_rate": 1.0169491525423729e-07,
+      "loss": 0.8802,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.2812101013113,
+      "learning_rate": 1.3559322033898305e-07,
+      "loss": 0.8447,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.22501739492148,
+      "learning_rate": 1.6949152542372883e-07,
+      "loss": 0.8661,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.107731386918358,
+      "learning_rate": 2.0338983050847458e-07,
+      "loss": 0.9148,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.537106942847725,
+      "learning_rate": 2.3728813559322036e-07,
+      "loss": 0.7615,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.780092872367305,
+      "learning_rate": 2.711864406779661e-07,
+      "loss": 0.8196,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.790315313064749,
+      "learning_rate": 3.050847457627119e-07,
+      "loss": 0.859,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.841389331094158,
+      "learning_rate": 3.3898305084745766e-07,
+      "loss": 0.8597,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.054485774470677,
+      "learning_rate": 3.7288135593220347e-07,
+      "loss": 0.8975,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.348866446220944,
+      "learning_rate": 4.0677966101694916e-07,
+      "loss": 0.8716,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.375820124871165,
+      "learning_rate": 4.4067796610169497e-07,
+      "loss": 0.8262,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.489161661750288,
+      "learning_rate": 4.745762711864407e-07,
+      "loss": 0.8317,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.647864966680701,
+      "learning_rate": 5.084745762711865e-07,
+      "loss": 0.8068,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.25567460340579,
+      "learning_rate": 5.423728813559322e-07,
+      "loss": 0.8784,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.243574981713614,
+      "learning_rate": 5.76271186440678e-07,
+      "loss": 0.7368,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.459065444090269,
+      "learning_rate": 6.101694915254238e-07,
+      "loss": 0.7662,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.014917635106802,
+      "learning_rate": 6.440677966101695e-07,
+      "loss": 0.7592,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.860940141421972,
+      "learning_rate": 6.779661016949153e-07,
+      "loss": 0.731,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.185099928998005,
+      "learning_rate": 7.118644067796611e-07,
+      "loss": 0.6707,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.23904808480235,
+      "learning_rate": 7.457627118644069e-07,
+      "loss": 0.7006,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.327549403384544,
+      "learning_rate": 7.796610169491527e-07,
+      "loss": 0.6307,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.097537794801515,
+      "learning_rate": 8.135593220338983e-07,
+      "loss": 0.6203,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.377204730201115,
+      "learning_rate": 8.474576271186441e-07,
+      "loss": 0.5821,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.358088694702184,
+      "learning_rate": 8.813559322033899e-07,
+      "loss": 0.602,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.197286363634574,
+      "learning_rate": 9.152542372881357e-07,
+      "loss": 0.6424,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.799809603049265,
+      "learning_rate": 9.491525423728814e-07,
+      "loss": 0.5751,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.383671404077906,
+      "learning_rate": 9.830508474576272e-07,
+      "loss": 0.5296,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.216187266889182,
+      "learning_rate": 1.016949152542373e-06,
+      "loss": 0.536,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.3106453196134393,
+      "learning_rate": 1.0508474576271187e-06,
+      "loss": 0.5286,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.618350813260054,
+      "learning_rate": 1.0847457627118644e-06,
+      "loss": 0.5057,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.5217817090570738,
+      "learning_rate": 1.1186440677966102e-06,
+      "loss": 0.4745,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3533426940286937,
+      "learning_rate": 1.152542372881356e-06,
+      "loss": 0.4211,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.61696040904103,
+      "learning_rate": 1.186440677966102e-06,
+      "loss": 0.4285,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.456578510501274,
+      "learning_rate": 1.2203389830508477e-06,
+      "loss": 0.4165,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.6810354384934776,
+      "learning_rate": 1.2542372881355932e-06,
+      "loss": 0.4464,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.387681024180993,
+      "learning_rate": 1.288135593220339e-06,
+      "loss": 0.4644,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4467785580460735,
+      "learning_rate": 1.322033898305085e-06,
+      "loss": 0.4539,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4529315649495667,
+      "learning_rate": 1.3559322033898307e-06,
+      "loss": 0.4166,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.212819968943549,
+      "learning_rate": 1.3898305084745764e-06,
+      "loss": 0.3938,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.863910204540751,
+      "learning_rate": 1.4237288135593222e-06,
+      "loss": 0.3991,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6764048559205706,
+      "learning_rate": 1.457627118644068e-06,
+      "loss": 0.372,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.799219722876445,
+      "learning_rate": 1.4915254237288139e-06,
+      "loss": 0.3803,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9642819860180136,
+      "learning_rate": 1.5254237288135596e-06,
+      "loss": 0.39,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.102560995531875,
+      "learning_rate": 1.5593220338983054e-06,
+      "loss": 0.399,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.2717682014897997,
+      "learning_rate": 1.593220338983051e-06,
+      "loss": 0.3334,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8451916643138015,
+      "learning_rate": 1.6271186440677967e-06,
+      "loss": 0.3806,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.791231995618392,
+      "learning_rate": 1.6610169491525424e-06,
+      "loss": 0.3786,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8661867054992236,
+      "learning_rate": 1.6949152542372882e-06,
+      "loss": 0.3763,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9794021888687487,
+      "learning_rate": 1.728813559322034e-06,
+      "loss": 0.3371,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.224468293642754,
+      "learning_rate": 1.7627118644067799e-06,
+      "loss": 0.3475,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.31298307443876,
+      "learning_rate": 1.7966101694915256e-06,
+      "loss": 0.3692,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8429593134990803,
+      "learning_rate": 1.8305084745762714e-06,
+      "loss": 0.336,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8510671554362415,
+      "learning_rate": 1.8644067796610171e-06,
+      "loss": 0.3369,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0884605066914403,
+      "learning_rate": 1.8983050847457629e-06,
+      "loss": 0.3531,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7766166145814422,
+      "learning_rate": 1.932203389830509e-06,
+      "loss": 0.3224,
+      "step": 57
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.00091805426027,
+      "learning_rate": 1.9661016949152544e-06,
+      "loss": 0.3207,
+      "step": 58
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.671445434483612,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.3473,
+      "step": 59
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8089600378647877,
+      "learning_rate": 2.033898305084746e-06,
+      "loss": 0.3158,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.857053304370097,
+      "learning_rate": 2.0677966101694914e-06,
+      "loss": 0.3353,
+      "step": 61
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8869217170364412,
+      "learning_rate": 2.1016949152542374e-06,
+      "loss": 0.2722,
+      "step": 62
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6325695751917972,
+      "learning_rate": 2.1355932203389833e-06,
+      "loss": 0.3661,
+      "step": 63
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6938194400308952,
+      "learning_rate": 2.169491525423729e-06,
+      "loss": 0.3112,
+      "step": 64
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0006172337912416,
+      "learning_rate": 2.203389830508475e-06,
+      "loss": 0.3639,
+      "step": 65
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0199725555296335,
+      "learning_rate": 2.2372881355932204e-06,
+      "loss": 0.3086,
+      "step": 66
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8730176329500001,
+      "learning_rate": 2.2711864406779663e-06,
+      "loss": 0.323,
+      "step": 67
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7440535132509414,
+      "learning_rate": 2.305084745762712e-06,
+      "loss": 0.343,
+      "step": 68
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.112486078314485,
+      "learning_rate": 2.338983050847458e-06,
+      "loss": 0.3391,
+      "step": 69
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7751924675393387,
+      "learning_rate": 2.372881355932204e-06,
+      "loss": 0.3234,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.964742351336944,
+      "learning_rate": 2.4067796610169493e-06,
+      "loss": 0.2933,
+      "step": 71
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3092146299206104,
+      "learning_rate": 2.4406779661016953e-06,
+      "loss": 0.2857,
+      "step": 72
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7904622412420848,
+      "learning_rate": 2.474576271186441e-06,
+      "loss": 0.3443,
+      "step": 73
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7863495982308861,
+      "learning_rate": 2.5084745762711864e-06,
+      "loss": 0.3031,
+      "step": 74
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5671252584011932,
+      "learning_rate": 2.5423728813559323e-06,
+      "loss": 0.2966,
+      "step": 75
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.656662139777409,
+      "learning_rate": 2.576271186440678e-06,
+      "loss": 0.3134,
+      "step": 76
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.041825731687294,
+      "learning_rate": 2.610169491525424e-06,
+      "loss": 0.3457,
+      "step": 77
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.516375034415401,
+      "learning_rate": 2.64406779661017e-06,
+      "loss": 0.2944,
+      "step": 78
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9869017421478328,
+      "learning_rate": 2.6779661016949153e-06,
+      "loss": 0.2837,
+      "step": 79
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6175841066321706,
+      "learning_rate": 2.7118644067796613e-06,
+      "loss": 0.2864,
+      "step": 80
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7170519798903372,
+      "learning_rate": 2.745762711864407e-06,
+      "loss": 0.3284,
+      "step": 81
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7972083881181034,
+      "learning_rate": 2.779661016949153e-06,
+      "loss": 0.299,
+      "step": 82
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.8307414003842113,
+      "learning_rate": 2.8135593220338988e-06,
+      "loss": 0.3023,
+      "step": 83
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.613381220629313,
+      "learning_rate": 2.8474576271186443e-06,
+      "loss": 0.3124,
+      "step": 84
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7950742462204192,
+      "learning_rate": 2.8813559322033903e-06,
+      "loss": 0.3006,
+      "step": 85
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.221539092274775,
+      "learning_rate": 2.915254237288136e-06,
+      "loss": 0.3048,
+      "step": 86
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.259845817762709,
+      "learning_rate": 2.9491525423728818e-06,
+      "loss": 0.3044,
+      "step": 87
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.541938397353296,
+      "learning_rate": 2.9830508474576277e-06,
+      "loss": 0.3434,
+      "step": 88
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1376130907622923,
+      "learning_rate": 3.0169491525423733e-06,
+      "loss": 0.3282,
+      "step": 89
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0565278408576604,
+      "learning_rate": 3.0508474576271192e-06,
+      "loss": 0.2927,
+      "step": 90
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.926909173349644,
+      "learning_rate": 3.0847457627118648e-06,
+      "loss": 0.3109,
+      "step": 91
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.6185375769384565,
+      "learning_rate": 3.1186440677966107e-06,
+      "loss": 0.3094,
+      "step": 92
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8249433332855955,
+      "learning_rate": 3.1525423728813563e-06,
+      "loss": 0.3225,
+      "step": 93
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7226923086848356,
+      "learning_rate": 3.186440677966102e-06,
+      "loss": 0.293,
+      "step": 94
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8579049970780483,
+      "learning_rate": 3.2203389830508473e-06,
+      "loss": 0.297,
+      "step": 95
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8217865526593613,
+      "learning_rate": 3.2542372881355933e-06,
+      "loss": 0.299,
+      "step": 96
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.9857064112941214,
+      "learning_rate": 3.288135593220339e-06,
+      "loss": 0.2983,
+      "step": 97
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0821655536527746,
+      "learning_rate": 3.322033898305085e-06,
+      "loss": 0.3261,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1380580706567773,
+      "learning_rate": 3.3559322033898308e-06,
+      "loss": 0.2775,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0160814105257923,
+      "learning_rate": 3.3898305084745763e-06,
+      "loss": 0.3317,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5909755454757093,
+      "learning_rate": 3.4237288135593223e-06,
+      "loss": 0.303,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.191902880650215,
+      "learning_rate": 3.457627118644068e-06,
+      "loss": 0.3183,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9238805550148785,
+      "learning_rate": 3.4915254237288138e-06,
+      "loss": 0.3072,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2333502172845163,
+      "learning_rate": 3.5254237288135597e-06,
+      "loss": 0.3586,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.589718562654825,
+      "learning_rate": 3.5593220338983053e-06,
+      "loss": 0.3147,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6143252679477402,
+      "learning_rate": 3.5932203389830512e-06,
+      "loss": 0.2904,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2836235321900076,
+      "learning_rate": 3.6271186440677968e-06,
+      "loss": 0.3066,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6830983780860997,
+      "learning_rate": 3.6610169491525427e-06,
+      "loss": 0.3336,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7804936986415487,
+      "learning_rate": 3.6949152542372883e-06,
+      "loss": 0.3005,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.065589112282441,
+      "learning_rate": 3.7288135593220342e-06,
+      "loss": 0.2842,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.350341154005084,
+      "learning_rate": 3.76271186440678e-06,
+      "loss": 0.3179,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4814230337002219,
+      "learning_rate": 3.7966101694915257e-06,
+      "loss": 0.2744,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.733113410194878,
+      "learning_rate": 3.830508474576271e-06,
+      "loss": 0.3027,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5967154788237727,
+      "learning_rate": 3.864406779661018e-06,
+      "loss": 0.3258,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3370723281642563,
+      "learning_rate": 3.898305084745763e-06,
+      "loss": 0.3114,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6445205745938416,
+      "learning_rate": 3.932203389830509e-06,
+      "loss": 0.3116,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0862236046831435,
+      "learning_rate": 3.966101694915255e-06,
+      "loss": 0.2998,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5820347545216342,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.3014,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9780553225075865,
+      "learning_rate": 4.033898305084746e-06,
+      "loss": 0.321,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0824896309123964,
+      "learning_rate": 4.067796610169492e-06,
+      "loss": 0.2436,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.235099816002165,
+      "learning_rate": 4.101694915254237e-06,
+      "loss": 0.2895,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9274858752777464,
+      "learning_rate": 4.135593220338983e-06,
+      "loss": 0.2712,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0017257398500234,
+      "learning_rate": 4.169491525423729e-06,
+      "loss": 0.2674,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3276432711309596,
+      "learning_rate": 4.203389830508475e-06,
+      "loss": 0.3218,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1175621752154283,
+      "learning_rate": 4.23728813559322e-06,
+      "loss": 0.3057,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8693993867856755,
+      "learning_rate": 4.271186440677967e-06,
+      "loss": 0.2679,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6694157796882232,
+      "learning_rate": 4.305084745762712e-06,
+      "loss": 0.2739,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.249653153315445,
+      "learning_rate": 4.338983050847458e-06,
+      "loss": 0.2827,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6424596204092072,
+      "learning_rate": 4.372881355932203e-06,
+      "loss": 0.2826,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7334690566735274,
+      "learning_rate": 4.40677966101695e-06,
+      "loss": 0.2932,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9839422437397973,
+      "learning_rate": 4.440677966101695e-06,
+      "loss": 0.3081,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.953781353665557,
+      "learning_rate": 4.474576271186441e-06,
+      "loss": 0.3027,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8911646090025422,
+      "learning_rate": 4.508474576271187e-06,
+      "loss": 0.3037,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6962571768979062,
+      "learning_rate": 4.542372881355933e-06,
+      "loss": 0.2561,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.634664008844036,
+      "learning_rate": 4.576271186440678e-06,
+      "loss": 0.2879,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.777790158029909,
+      "learning_rate": 4.610169491525424e-06,
+      "loss": 0.2721,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7148761472284506,
+      "learning_rate": 4.64406779661017e-06,
+      "loss": 0.2717,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7208081889064668,
+      "learning_rate": 4.677966101694916e-06,
+      "loss": 0.2838,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.89025248616027,
+      "learning_rate": 4.711864406779661e-06,
+      "loss": 0.3075,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7365633256432669,
+      "learning_rate": 4.745762711864408e-06,
+      "loss": 0.2909,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6009133709225887,
+      "learning_rate": 4.779661016949153e-06,
+      "loss": 0.2538,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0332412939427944,
+      "learning_rate": 4.813559322033899e-06,
+      "loss": 0.3089,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.775911485217657,
+      "learning_rate": 4.847457627118645e-06,
+      "loss": 0.3002,
+      "step": 143
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4574169659822291,
+      "learning_rate": 4.881355932203391e-06,
+      "loss": 0.2843,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7122297188914908,
+      "learning_rate": 4.915254237288136e-06,
+      "loss": 0.2923,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8661430535697405,
+      "learning_rate": 4.949152542372882e-06,
+      "loss": 0.2933,
+      "step": 146
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5280893595273282,
+      "learning_rate": 4.983050847457628e-06,
+      "loss": 0.3097,
+      "step": 147
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5414815172107994,
+      "learning_rate": 5.016949152542373e-06,
+      "loss": 0.2894,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6979336511591085,
+      "learning_rate": 5.050847457627119e-06,
+      "loss": 0.3058,
+      "step": 149
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.399367962381608,
+      "learning_rate": 5.084745762711865e-06,
+      "loss": 0.2826,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.666309227933981,
+      "learning_rate": 5.118644067796611e-06,
+      "loss": 0.2729,
+      "step": 151
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.479961432900143,
+      "learning_rate": 5.152542372881356e-06,
+      "loss": 0.2805,
+      "step": 152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7384952021588505,
+      "learning_rate": 5.186440677966102e-06,
+      "loss": 0.2759,
+      "step": 153
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8596650431273607,
+      "learning_rate": 5.220338983050848e-06,
+      "loss": 0.2758,
+      "step": 154
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.857375302476961,
+      "learning_rate": 5.254237288135594e-06,
+      "loss": 0.3048,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7591182469694164,
+      "learning_rate": 5.28813559322034e-06,
+      "loss": 0.2762,
+      "step": 156
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2948961387152282,
+      "learning_rate": 5.322033898305086e-06,
+      "loss": 0.2911,
+      "step": 157
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.261579664075019,
+      "learning_rate": 5.355932203389831e-06,
+      "loss": 0.2845,
+      "step": 158
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0906674064521806,
+      "learning_rate": 5.389830508474577e-06,
+      "loss": 0.2798,
+      "step": 159
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.689900338950711,
+      "learning_rate": 5.423728813559323e-06,
+      "loss": 0.3181,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5558610571402536,
+      "learning_rate": 5.457627118644067e-06,
+      "loss": 0.2835,
+      "step": 161
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.231530175438594,
+      "learning_rate": 5.491525423728814e-06,
+      "loss": 0.3044,
+      "step": 162
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6429093181423005,
+      "learning_rate": 5.525423728813559e-06,
+      "loss": 0.3068,
+      "step": 163
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7844549696996412,
+      "learning_rate": 5.559322033898306e-06,
+      "loss": 0.2853,
+      "step": 164
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0082259106645943,
+      "learning_rate": 5.593220338983051e-06,
+      "loss": 0.3477,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.960115573884396,
+      "learning_rate": 5.6271186440677975e-06,
+      "loss": 0.2772,
+      "step": 166
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.068249027081921,
+      "learning_rate": 5.661016949152542e-06,
+      "loss": 0.2967,
+      "step": 167
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7051358404672956,
+      "learning_rate": 5.694915254237289e-06,
+      "loss": 0.2871,
+      "step": 168
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.709105980487174,
+      "learning_rate": 5.728813559322034e-06,
+      "loss": 0.2613,
+      "step": 169
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.950311162930796,
+      "learning_rate": 5.7627118644067805e-06,
+      "loss": 0.2923,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.651513874408586,
+      "learning_rate": 5.796610169491525e-06,
+      "loss": 0.3087,
+      "step": 171
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.710559983293689,
+      "learning_rate": 5.830508474576272e-06,
+      "loss": 0.2794,
+      "step": 172
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7771722763697293,
+      "learning_rate": 5.864406779661017e-06,
+      "loss": 0.2783,
+      "step": 173
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.160139050072537,
+      "learning_rate": 5.8983050847457635e-06,
+      "loss": 0.2966,
+      "step": 174
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5002256496386404,
+      "learning_rate": 5.932203389830509e-06,
+      "loss": 0.2879,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8875259490488177,
+      "learning_rate": 5.9661016949152555e-06,
+      "loss": 0.2937,
+      "step": 176
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7701374784962995,
+      "learning_rate": 6e-06,
+      "loss": 0.2667,
+      "step": 177
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7400487952240289,
+      "learning_rate": 6.0338983050847465e-06,
+      "loss": 0.317,
+      "step": 178
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6262926424592759,
+      "learning_rate": 6.067796610169492e-06,
+      "loss": 0.2988,
+      "step": 179
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7533113610007194,
+      "learning_rate": 6.1016949152542385e-06,
+      "loss": 0.2743,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1487191473952074,
+      "learning_rate": 6.135593220338983e-06,
+      "loss": 0.2731,
+      "step": 181
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.837373024422903,
+      "learning_rate": 6.1694915254237295e-06,
+      "loss": 0.2474,
+      "step": 182
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.384431938898038,
+      "learning_rate": 6.203389830508475e-06,
+      "loss": 0.2928,
+      "step": 183
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5271044696646137,
+      "learning_rate": 6.2372881355932215e-06,
+      "loss": 0.2544,
+      "step": 184
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3710907064781839,
+      "learning_rate": 6.271186440677966e-06,
+      "loss": 0.2716,
+      "step": 185
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7978500758050704,
+      "learning_rate": 6.3050847457627125e-06,
+      "loss": 0.2629,
+      "step": 186
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.266493908137502,
+      "learning_rate": 6.338983050847458e-06,
+      "loss": 0.2636,
+      "step": 187
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6865435211762627,
+      "learning_rate": 6.372881355932204e-06,
+      "loss": 0.2919,
+      "step": 188
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5177106818253951,
+      "learning_rate": 6.40677966101695e-06,
+      "loss": 0.2696,
+      "step": 189
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.598312898809967,
+      "learning_rate": 6.440677966101695e-06,
+      "loss": 0.2941,
+      "step": 190
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7020330476074053,
+      "learning_rate": 6.474576271186441e-06,
+      "loss": 0.2566,
+      "step": 191
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6452188115531354,
+      "learning_rate": 6.508474576271187e-06,
+      "loss": 0.2627,
+      "step": 192
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5337855227059187,
+      "learning_rate": 6.542372881355933e-06,
+      "loss": 0.2749,
+      "step": 193
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5885009119467683,
+      "learning_rate": 6.576271186440678e-06,
+      "loss": 0.2871,
+      "step": 194
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9656764093443815,
+      "learning_rate": 6.610169491525424e-06,
+      "loss": 0.3027,
+      "step": 195
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0483602782105494,
+      "learning_rate": 6.64406779661017e-06,
+      "loss": 0.3092,
+      "step": 196
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8565023223461687,
+      "learning_rate": 6.677966101694916e-06,
+      "loss": 0.3084,
+      "step": 197
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7646513939378161,
+      "learning_rate": 6.7118644067796615e-06,
+      "loss": 0.2953,
+      "step": 198
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6957118175929442,
+      "learning_rate": 6.745762711864408e-06,
+      "loss": 0.2939,
+      "step": 199
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6271626182864425,
+      "learning_rate": 6.779661016949153e-06,
+      "loss": 0.2732,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.665492075397036,
+      "learning_rate": 6.813559322033899e-06,
+      "loss": 0.2643,
+      "step": 201
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7849521054857391,
+      "learning_rate": 6.8474576271186445e-06,
+      "loss": 0.2967,
+      "step": 202
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6961191232197494,
+      "learning_rate": 6.881355932203391e-06,
+      "loss": 0.3109,
+      "step": 203
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.725277238030002,
+      "learning_rate": 6.915254237288136e-06,
+      "loss": 0.2882,
+      "step": 204
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7040601722570679,
+      "learning_rate": 6.949152542372882e-06,
+      "loss": 0.261,
+      "step": 205
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7790750902245551,
+      "learning_rate": 6.9830508474576275e-06,
+      "loss": 0.2928,
+      "step": 206
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8954542367183922,
+      "learning_rate": 7.016949152542374e-06,
+      "loss": 0.2961,
+      "step": 207
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0100427678525317,
+      "learning_rate": 7.0508474576271195e-06,
+      "loss": 0.271,
+      "step": 208
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.635043641028692,
+      "learning_rate": 7.084745762711865e-06,
+      "loss": 0.3017,
+      "step": 209
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5721374879331982,
+      "learning_rate": 7.1186440677966106e-06,
+      "loss": 0.2586,
+      "step": 210
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7623509032189595,
+      "learning_rate": 7.152542372881357e-06,
+      "loss": 0.2852,
+      "step": 211
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8084559334984929,
+      "learning_rate": 7.1864406779661025e-06,
+      "loss": 0.2502,
+      "step": 212
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5700471733368953,
+      "learning_rate": 7.220338983050849e-06,
+      "loss": 0.2658,
+      "step": 213
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7884143221055777,
+      "learning_rate": 7.2542372881355936e-06,
+      "loss": 0.2813,
+      "step": 214
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8403028285709235,
+      "learning_rate": 7.288135593220339e-06,
+      "loss": 0.2509,
+      "step": 215
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6358262298861554,
+      "learning_rate": 7.3220338983050855e-06,
+      "loss": 0.2608,
+      "step": 216
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.065627300198091,
+      "learning_rate": 7.355932203389831e-06,
+      "loss": 0.2775,
+      "step": 217
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5288052157129923,
+      "learning_rate": 7.3898305084745766e-06,
+      "loss": 0.2694,
+      "step": 218
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6187358313701743,
+      "learning_rate": 7.423728813559322e-06,
+      "loss": 0.2611,
+      "step": 219
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.500589738935865,
+      "learning_rate": 7.4576271186440685e-06,
+      "loss": 0.2611,
+      "step": 220
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6846488257425314,
+      "learning_rate": 7.491525423728814e-06,
+      "loss": 0.2625,
+      "step": 221
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5149283412876415,
+      "learning_rate": 7.52542372881356e-06,
+      "loss": 0.2698,
+      "step": 222
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.520702365266604,
+      "learning_rate": 7.559322033898305e-06,
+      "loss": 0.2756,
+      "step": 223
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6031162343212535,
+      "learning_rate": 7.5932203389830515e-06,
+      "loss": 0.2617,
+      "step": 224
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6952860116491963,
+      "learning_rate": 7.627118644067797e-06,
+      "loss": 0.2818,
+      "step": 225
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.477255037322905,
+      "learning_rate": 7.661016949152543e-06,
+      "loss": 0.2496,
+      "step": 226
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6660271734854857,
+      "learning_rate": 7.694915254237289e-06,
+      "loss": 0.2768,
+      "step": 227
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9775731672940686,
+      "learning_rate": 7.728813559322035e-06,
+      "loss": 0.2535,
+      "step": 228
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7011300558433908,
+      "learning_rate": 7.76271186440678e-06,
+      "loss": 0.2851,
+      "step": 229
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.646079153476468,
+      "learning_rate": 7.796610169491526e-06,
+      "loss": 0.3166,
+      "step": 230
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4393049865391425,
+      "learning_rate": 7.830508474576271e-06,
+      "loss": 0.3049,
+      "step": 231
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6556807137319012,
+      "learning_rate": 7.864406779661017e-06,
+      "loss": 0.2875,
+      "step": 232
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8731866145038525,
+      "learning_rate": 7.898305084745764e-06,
+      "loss": 0.2701,
+      "step": 233
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.62571796111418,
+      "learning_rate": 7.93220338983051e-06,
+      "loss": 0.2728,
+      "step": 234
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.487202975941064,
+      "learning_rate": 7.966101694915255e-06,
+      "loss": 0.2719,
+      "step": 235
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.535136361003896,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.2615,
+      "step": 236
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6670845697053458,
+      "learning_rate": 8.033898305084746e-06,
+      "loss": 0.2603,
+      "step": 237
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2513572304096718,
+      "learning_rate": 8.067796610169492e-06,
+      "loss": 0.2842,
+      "step": 238
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6010215262108967,
+      "learning_rate": 8.101694915254237e-06,
+      "loss": 0.2575,
+      "step": 239
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.264897174430261,
+      "learning_rate": 8.135593220338983e-06,
+      "loss": 0.28,
+      "step": 240
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5210659548113379,
+      "learning_rate": 8.16949152542373e-06,
+      "loss": 0.2957,
+      "step": 241
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5993802470049059,
+      "learning_rate": 8.203389830508475e-06,
+      "loss": 0.284,
+      "step": 242
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4987229304363077,
+      "learning_rate": 8.237288135593221e-06,
+      "loss": 0.2617,
+      "step": 243
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.45585250607327,
+      "learning_rate": 8.271186440677966e-06,
+      "loss": 0.2637,
+      "step": 244
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.538155435073543,
+      "learning_rate": 8.305084745762712e-06,
+      "loss": 0.2741,
+      "step": 245
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1391423571076142,
+      "learning_rate": 8.338983050847458e-06,
+      "loss": 0.2631,
+      "step": 246
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7253417394805781,
+      "learning_rate": 8.372881355932205e-06,
+      "loss": 0.2565,
+      "step": 247
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6275070709595192,
+      "learning_rate": 8.40677966101695e-06,
+      "loss": 0.2644,
+      "step": 248
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6116776111875923,
+      "learning_rate": 8.440677966101696e-06,
+      "loss": 0.2802,
+      "step": 249
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.570689718867318,
+      "learning_rate": 8.47457627118644e-06,
+      "loss": 0.2606,
+      "step": 250
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5257535084444283,
+      "learning_rate": 8.508474576271187e-06,
+      "loss": 0.2729,
+      "step": 251
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4065841492399755,
+      "learning_rate": 8.542372881355933e-06,
+      "loss": 0.2552,
+      "step": 252
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.580942735174244,
+      "learning_rate": 8.57627118644068e-06,
+      "loss": 0.2869,
+      "step": 253
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7823741192008924,
+      "learning_rate": 8.610169491525424e-06,
+      "loss": 0.2458,
+      "step": 254
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6882930628464303,
+      "learning_rate": 8.64406779661017e-06,
+      "loss": 0.2797,
+      "step": 255
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5710000159998525,
+      "learning_rate": 8.677966101694915e-06,
+      "loss": 0.2596,
+      "step": 256
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5117969977132588,
+      "learning_rate": 8.711864406779662e-06,
+      "loss": 0.2915,
+      "step": 257
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0879634715923228,
+      "learning_rate": 8.745762711864407e-06,
+      "loss": 0.2928,
+      "step": 258
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4763672894599624,
+      "learning_rate": 8.779661016949153e-06,
+      "loss": 0.2661,
+      "step": 259
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5887813347318072,
+      "learning_rate": 8.8135593220339e-06,
+      "loss": 0.2644,
+      "step": 260
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1284805411680634,
+      "learning_rate": 8.847457627118646e-06,
+      "loss": 0.2896,
+      "step": 261
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6764740798105202,
+      "learning_rate": 8.88135593220339e-06,
+      "loss": 0.2679,
+      "step": 262
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5530926451627123,
+      "learning_rate": 8.915254237288137e-06,
+      "loss": 0.2618,
+      "step": 263
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5762836341877522,
+      "learning_rate": 8.949152542372881e-06,
+      "loss": 0.278,
+      "step": 264
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.229849002174887,
+      "learning_rate": 8.983050847457628e-06,
+      "loss": 0.2623,
+      "step": 265
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.360225880617487,
+      "learning_rate": 9.016949152542374e-06,
+      "loss": 0.2769,
+      "step": 266
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.578613608691057,
+      "learning_rate": 9.05084745762712e-06,
+      "loss": 0.2935,
+      "step": 267
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5343613642802507,
+      "learning_rate": 9.084745762711865e-06,
+      "loss": 0.2827,
+      "step": 268
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9651340500582921,
+      "learning_rate": 9.11864406779661e-06,
+      "loss": 0.3065,
+      "step": 269
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.189584326247112,
+      "learning_rate": 9.152542372881356e-06,
+      "loss": 0.2452,
+      "step": 270
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.568110390610819,
+      "learning_rate": 9.186440677966101e-06,
+      "loss": 0.2642,
+      "step": 271
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.490918665724689,
+      "learning_rate": 9.220338983050847e-06,
+      "loss": 0.2484,
+      "step": 272
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5178780940378127,
+      "learning_rate": 9.254237288135594e-06,
+      "loss": 0.2615,
+      "step": 273
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4243185626593395,
+      "learning_rate": 9.28813559322034e-06,
+      "loss": 0.2706,
+      "step": 274
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3943699378544865,
+      "learning_rate": 9.322033898305085e-06,
+      "loss": 0.2846,
+      "step": 275
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.39958325547155,
+      "learning_rate": 9.355932203389831e-06,
+      "loss": 0.2644,
+      "step": 276
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5566516750657318,
+      "learning_rate": 9.389830508474576e-06,
+      "loss": 0.2709,
+      "step": 277
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5807818263236182,
+      "learning_rate": 9.423728813559322e-06,
+      "loss": 0.2904,
+      "step": 278
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6657073622633627,
+      "learning_rate": 9.457627118644069e-06,
+      "loss": 0.2831,
+      "step": 279
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5508860676326361,
+      "learning_rate": 9.491525423728815e-06,
+      "loss": 0.3093,
+      "step": 280
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4241612321960757,
+      "learning_rate": 9.52542372881356e-06,
+      "loss": 0.2769,
+      "step": 281
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5006924170004527,
+      "learning_rate": 9.559322033898306e-06,
+      "loss": 0.2415,
+      "step": 282
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5294543822090856,
+      "learning_rate": 9.593220338983051e-06,
+      "loss": 0.2515,
+      "step": 283
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.533886353920441,
+      "learning_rate": 9.627118644067797e-06,
+      "loss": 0.2829,
+      "step": 284
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.069418788425172,
+      "learning_rate": 9.661016949152544e-06,
+      "loss": 0.2887,
+      "step": 285
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5889620658480574,
+      "learning_rate": 9.69491525423729e-06,
+      "loss": 0.2991,
+      "step": 286
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4597563573302703,
+      "learning_rate": 9.728813559322035e-06,
+      "loss": 0.2624,
+      "step": 287
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.740665680466997,
+      "learning_rate": 9.762711864406781e-06,
+      "loss": 0.2881,
+      "step": 288
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3885281952656445,
+      "learning_rate": 9.796610169491526e-06,
+      "loss": 0.2958,
+      "step": 289
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4980986665212326,
+      "learning_rate": 9.830508474576272e-06,
+      "loss": 0.2894,
+      "step": 290
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6970976017778991,
+      "learning_rate": 9.864406779661017e-06,
+      "loss": 0.277,
+      "step": 291
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5465766596610715,
+      "learning_rate": 9.898305084745763e-06,
+      "loss": 0.2723,
+      "step": 292
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7023966032743651,
+      "learning_rate": 9.93220338983051e-06,
+      "loss": 0.2977,
+      "step": 293
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3390528257847034,
+      "learning_rate": 9.966101694915256e-06,
+      "loss": 0.2537,
+      "step": 294
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1821890114626603,
+      "learning_rate": 1e-05,
+      "loss": 0.2668,
+      "step": 295
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7286845032215452,
+      "learning_rate": 1.0033898305084746e-05,
+      "loss": 0.2744,
+      "step": 296
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4950792801793105,
+      "learning_rate": 1.0067796610169492e-05,
+      "loss": 0.272,
+      "step": 297
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4700432391190763,
+      "learning_rate": 1.0101694915254238e-05,
+      "loss": 0.2662,
+      "step": 298
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3967794895014582,
+      "learning_rate": 1.0135593220338985e-05,
+      "loss": 0.2802,
+      "step": 299
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.391964862237865,
+      "learning_rate": 1.016949152542373e-05,
+      "loss": 0.273,
+      "step": 300
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4563931631110092,
+      "learning_rate": 1.0203389830508474e-05,
+      "loss": 0.2905,
+      "step": 301
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.284413744656201,
+      "learning_rate": 1.0237288135593222e-05,
+      "loss": 0.2749,
+      "step": 302
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7292121695347487,
+      "learning_rate": 1.0271186440677967e-05,
+      "loss": 0.243,
+      "step": 303
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8232969594205204,
+      "learning_rate": 1.0305084745762712e-05,
+      "loss": 0.2722,
+      "step": 304
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7932246344435712,
+      "learning_rate": 1.0338983050847458e-05,
+      "loss": 0.2684,
+      "step": 305
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.533073576390442,
+      "learning_rate": 1.0372881355932204e-05,
+      "loss": 0.2745,
+      "step": 306
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8333910248684946,
+      "learning_rate": 1.040677966101695e-05,
+      "loss": 0.293,
+      "step": 307
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1995055633474583,
+      "learning_rate": 1.0440677966101695e-05,
+      "loss": 0.2629,
+      "step": 308
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4645465210433868,
+      "learning_rate": 1.047457627118644e-05,
+      "loss": 0.2698,
+      "step": 309
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.346835487540612,
+      "learning_rate": 1.0508474576271188e-05,
+      "loss": 0.2704,
+      "step": 310
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4321145393072345,
+      "learning_rate": 1.0542372881355933e-05,
+      "loss": 0.2677,
+      "step": 311
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5265056300536548,
+      "learning_rate": 1.057627118644068e-05,
+      "loss": 0.2803,
+      "step": 312
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3311183401050715,
+      "learning_rate": 1.0610169491525424e-05,
+      "loss": 0.2662,
+      "step": 313
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4977642596654024,
+      "learning_rate": 1.0644067796610172e-05,
+      "loss": 0.2729,
+      "step": 314
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5866201311518897,
+      "learning_rate": 1.0677966101694917e-05,
+      "loss": 0.255,
+      "step": 315
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5568209460381872,
+      "learning_rate": 1.0711864406779661e-05,
+      "loss": 0.2771,
+      "step": 316
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3484579322691495,
+      "learning_rate": 1.0745762711864408e-05,
+      "loss": 0.254,
+      "step": 317
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4060019872392868,
+      "learning_rate": 1.0779661016949154e-05,
+      "loss": 0.2847,
+      "step": 318
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6215123876307729,
+      "learning_rate": 1.08135593220339e-05,
+      "loss": 0.2582,
+      "step": 319
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1960240824180124,
+      "learning_rate": 1.0847457627118645e-05,
+      "loss": 0.2757,
+      "step": 320
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6496098507964483,
+      "learning_rate": 1.088135593220339e-05,
+      "loss": 0.2946,
+      "step": 321
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3862470565731742,
+      "learning_rate": 1.0915254237288135e-05,
+      "loss": 0.2768,
+      "step": 322
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4544121582161837,
+      "learning_rate": 1.0949152542372883e-05,
+      "loss": 0.3032,
+      "step": 323
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4175535736163114,
+      "learning_rate": 1.0983050847457627e-05,
+      "loss": 0.2686,
+      "step": 324
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3191834379151826,
+      "learning_rate": 1.1016949152542374e-05,
+      "loss": 0.252,
+      "step": 325
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.33272323985383,
+      "learning_rate": 1.1050847457627118e-05,
+      "loss": 0.2779,
+      "step": 326
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.660681621213467,
+      "learning_rate": 1.1084745762711867e-05,
+      "loss": 0.2794,
+      "step": 327
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.431182692042468,
+      "learning_rate": 1.1118644067796611e-05,
+      "loss": 0.2699,
+      "step": 328
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4489972397114494,
+      "learning_rate": 1.1152542372881356e-05,
+      "loss": 0.2682,
+      "step": 329
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6541146528927058,
+      "learning_rate": 1.1186440677966102e-05,
+      "loss": 0.3242,
+      "step": 330
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.331208350127372,
+      "learning_rate": 1.1220338983050849e-05,
+      "loss": 0.266,
+      "step": 331
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4591943463955406,
+      "learning_rate": 1.1254237288135595e-05,
+      "loss": 0.2462,
+      "step": 332
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4940367024477517,
+      "learning_rate": 1.128813559322034e-05,
+      "loss": 0.2634,
+      "step": 333
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6020801701154066,
+      "learning_rate": 1.1322033898305084e-05,
+      "loss": 0.258,
+      "step": 334
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.567603002600629,
+      "learning_rate": 1.1355932203389833e-05,
+      "loss": 0.2893,
+      "step": 335
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4512891656223712,
+      "learning_rate": 1.1389830508474577e-05,
+      "loss": 0.2592,
+      "step": 336
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9013614792745004,
+      "learning_rate": 1.1423728813559322e-05,
+      "loss": 0.3183,
+      "step": 337
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4816696067269588,
+      "learning_rate": 1.1457627118644068e-05,
+      "loss": 0.2521,
+      "step": 338
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4699744125682022,
+      "learning_rate": 1.1491525423728815e-05,
+      "loss": 0.2435,
+      "step": 339
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5292539331602504,
+      "learning_rate": 1.1525423728813561e-05,
+      "loss": 0.2773,
+      "step": 340
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.493805068160698,
+      "learning_rate": 1.1559322033898306e-05,
+      "loss": 0.2776,
+      "step": 341
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3320441267535226,
+      "learning_rate": 1.159322033898305e-05,
+      "loss": 0.3148,
+      "step": 342
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5768942640637522,
+      "learning_rate": 1.1627118644067799e-05,
+      "loss": 0.2882,
+      "step": 343
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6327174925876067,
+      "learning_rate": 1.1661016949152543e-05,
+      "loss": 0.2643,
+      "step": 344
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5979901584661973,
+      "learning_rate": 1.169491525423729e-05,
+      "loss": 0.2793,
+      "step": 345
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.490949422523808,
+      "learning_rate": 1.1728813559322034e-05,
+      "loss": 0.2797,
+      "step": 346
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6373314185665537,
+      "learning_rate": 1.1762711864406782e-05,
+      "loss": 0.2702,
+      "step": 347
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.494597348638564,
+      "learning_rate": 1.1796610169491527e-05,
+      "loss": 0.256,
+      "step": 348
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.268989549770058,
+      "learning_rate": 1.1830508474576272e-05,
+      "loss": 0.275,
+      "step": 349
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5728102611683135,
+      "learning_rate": 1.1864406779661018e-05,
+      "loss": 0.2747,
+      "step": 350
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4554919008312315,
+      "learning_rate": 1.1898305084745763e-05,
+      "loss": 0.2656,
+      "step": 351
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6882217452125865,
+      "learning_rate": 1.1932203389830511e-05,
+      "loss": 0.2724,
+      "step": 352
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5458553639090058,
+      "learning_rate": 1.1966101694915256e-05,
+      "loss": 0.2574,
+      "step": 353
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9734651946977806,
+      "learning_rate": 1.2e-05,
+      "loss": 0.2682,
+      "step": 354
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4674881583289108,
+      "learning_rate": 1.2033898305084745e-05,
+      "loss": 0.2624,
+      "step": 355
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6725844123567122,
+      "learning_rate": 1.2067796610169493e-05,
+      "loss": 0.2693,
+      "step": 356
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3657325069194421,
+      "learning_rate": 1.2101694915254238e-05,
+      "loss": 0.2422,
+      "step": 357
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4560212112782298,
+      "learning_rate": 1.2135593220338984e-05,
+      "loss": 0.2519,
+      "step": 358
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3112148572383935,
+      "learning_rate": 1.2169491525423729e-05,
+      "loss": 0.2769,
+      "step": 359
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.491608195320627,
+      "learning_rate": 1.2203389830508477e-05,
+      "loss": 0.2526,
+      "step": 360
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3012264572970316,
+      "learning_rate": 1.2237288135593222e-05,
+      "loss": 0.2841,
+      "step": 361
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.42262697227907,
+      "learning_rate": 1.2271186440677966e-05,
+      "loss": 0.264,
+      "step": 362
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.376381527131282,
+      "learning_rate": 1.2305084745762713e-05,
+      "loss": 0.2895,
+      "step": 363
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3621092285439624,
+      "learning_rate": 1.2338983050847459e-05,
+      "loss": 0.2818,
+      "step": 364
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6416902061672696,
+      "learning_rate": 1.2372881355932205e-05,
+      "loss": 0.242,
+      "step": 365
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8070792328678054,
+      "learning_rate": 1.240677966101695e-05,
+      "loss": 0.2513,
+      "step": 366
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4230650127004605,
+      "learning_rate": 1.2440677966101695e-05,
+      "loss": 0.2663,
+      "step": 367
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5876663584225577,
+      "learning_rate": 1.2474576271186443e-05,
+      "loss": 0.2619,
+      "step": 368
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4967545863007254,
+      "learning_rate": 1.2508474576271188e-05,
+      "loss": 0.2822,
+      "step": 369
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9039535303784283,
+      "learning_rate": 1.2542372881355932e-05,
+      "loss": 0.284,
+      "step": 370
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.428940565663837,
+      "learning_rate": 1.2576271186440679e-05,
+      "loss": 0.2601,
+      "step": 371
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5662155563814324,
+      "learning_rate": 1.2610169491525425e-05,
+      "loss": 0.2873,
+      "step": 372
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2971094059613077,
+      "learning_rate": 1.2644067796610171e-05,
+      "loss": 0.2626,
+      "step": 373
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0351725300274326,
+      "learning_rate": 1.2677966101694916e-05,
+      "loss": 0.2743,
+      "step": 374
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3553284978125044,
+      "learning_rate": 1.2711864406779661e-05,
+      "loss": 0.2709,
+      "step": 375
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5882208330471586,
+      "learning_rate": 1.2745762711864407e-05,
+      "loss": 0.2708,
+      "step": 376
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8462534147511642,
+      "learning_rate": 1.2779661016949154e-05,
+      "loss": 0.2699,
+      "step": 377
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.426504473136271,
+      "learning_rate": 1.28135593220339e-05,
+      "loss": 0.281,
+      "step": 378
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2340057175968697,
+      "learning_rate": 1.2847457627118645e-05,
+      "loss": 0.2483,
+      "step": 379
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.835330825736432,
+      "learning_rate": 1.288135593220339e-05,
+      "loss": 0.308,
+      "step": 380
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3981944760568665,
+      "learning_rate": 1.2915254237288137e-05,
+      "loss": 0.2507,
+      "step": 381
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3749165361391003,
+      "learning_rate": 1.2949152542372882e-05,
+      "loss": 0.2647,
+      "step": 382
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0292717133598073,
+      "learning_rate": 1.2983050847457629e-05,
+      "loss": 0.2718,
+      "step": 383
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4857167490857963,
+      "learning_rate": 1.3016949152542373e-05,
+      "loss": 0.2635,
+      "step": 384
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.747307923320907,
+      "learning_rate": 1.305084745762712e-05,
+      "loss": 0.2706,
+      "step": 385
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3277925058690871,
+      "learning_rate": 1.3084745762711866e-05,
+      "loss": 0.2432,
+      "step": 386
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5142700170780168,
+      "learning_rate": 1.311864406779661e-05,
+      "loss": 0.2811,
+      "step": 387
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3779037473413898,
+      "learning_rate": 1.3152542372881355e-05,
+      "loss": 0.2839,
+      "step": 388
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5766348414501605,
+      "learning_rate": 1.3186440677966103e-05,
+      "loss": 0.2622,
+      "step": 389
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2036696780959797,
+      "learning_rate": 1.3220338983050848e-05,
+      "loss": 0.2799,
+      "step": 390
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3338172319722734,
+      "learning_rate": 1.3254237288135595e-05,
+      "loss": 0.2516,
+      "step": 391
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5466043637609173,
+      "learning_rate": 1.328813559322034e-05,
+      "loss": 0.2892,
+      "step": 392
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.375902595553232,
+      "learning_rate": 1.3322033898305087e-05,
+      "loss": 0.3,
+      "step": 393
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.223183507661743,
+      "learning_rate": 1.3355932203389832e-05,
+      "loss": 0.2864,
+      "step": 394
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4529473301965807,
+      "learning_rate": 1.3389830508474577e-05,
+      "loss": 0.2643,
+      "step": 395
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3608254167875427,
+      "learning_rate": 1.3423728813559323e-05,
+      "loss": 0.2635,
+      "step": 396
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2513147886137037,
+      "learning_rate": 1.345762711864407e-05,
+      "loss": 0.2513,
+      "step": 397
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5112671514145013,
+      "learning_rate": 1.3491525423728816e-05,
+      "loss": 0.2861,
+      "step": 398
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6150957465634022,
+      "learning_rate": 1.352542372881356e-05,
+      "loss": 0.2701,
+      "step": 399
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7792619826658456,
+      "learning_rate": 1.3559322033898305e-05,
+      "loss": 0.2726,
+      "step": 400
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3186392092611758,
+      "learning_rate": 1.3593220338983053e-05,
+      "loss": 0.2749,
+      "step": 401
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.255587345530338,
+      "learning_rate": 1.3627118644067798e-05,
+      "loss": 0.2524,
+      "step": 402
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2400872274349326,
+      "learning_rate": 1.3661016949152543e-05,
+      "loss": 0.2778,
+      "step": 403
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.482742495732322,
+      "learning_rate": 1.3694915254237289e-05,
+      "loss": 0.2698,
+      "step": 404
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9013554292200026,
+      "learning_rate": 1.3728813559322034e-05,
+      "loss": 0.2835,
+      "step": 405
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.401766043993974,
+      "learning_rate": 1.3762711864406782e-05,
+      "loss": 0.2991,
+      "step": 406
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.881122545071624,
+      "learning_rate": 1.3796610169491527e-05,
+      "loss": 0.3155,
+      "step": 407
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.644242035858314,
+      "learning_rate": 1.3830508474576271e-05,
+      "loss": 0.2766,
+      "step": 408
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4160651170704013,
+      "learning_rate": 1.3864406779661018e-05,
+      "loss": 0.2864,
+      "step": 409
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1874236164670124,
+      "learning_rate": 1.3898305084745764e-05,
+      "loss": 0.275,
+      "step": 410
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2883313224686657,
+      "learning_rate": 1.393220338983051e-05,
+      "loss": 0.2647,
+      "step": 411
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2956543891841616,
+      "learning_rate": 1.3966101694915255e-05,
+      "loss": 0.2598,
+      "step": 412
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4091672402968367,
+      "learning_rate": 1.4e-05,
+      "loss": 0.2744,
+      "step": 413
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3024750495831694,
+      "learning_rate": 1.4033898305084748e-05,
+      "loss": 0.2525,
+      "step": 414
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7724406208457262,
+      "learning_rate": 1.4067796610169493e-05,
+      "loss": 0.2916,
+      "step": 415
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.306539549408704,
+      "learning_rate": 1.4101694915254239e-05,
+      "loss": 0.2697,
+      "step": 416
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4086019817090085,
+      "learning_rate": 1.4135593220338984e-05,
+      "loss": 0.2597,
+      "step": 417
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5154347103816932,
+      "learning_rate": 1.416949152542373e-05,
+      "loss": 0.2622,
+      "step": 418
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2824331064750205,
+      "learning_rate": 1.4203389830508476e-05,
+      "loss": 0.295,
+      "step": 419
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1915074345746457,
+      "learning_rate": 1.4237288135593221e-05,
+      "loss": 0.2642,
+      "step": 420
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1801999544960764,
+      "learning_rate": 1.4271186440677966e-05,
+      "loss": 0.2695,
+      "step": 421
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2427987470169837,
+      "learning_rate": 1.4305084745762714e-05,
+      "loss": 0.2693,
+      "step": 422
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.416370988934776,
+      "learning_rate": 1.4338983050847459e-05,
+      "loss": 0.2566,
+      "step": 423
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4784492398159457,
+      "learning_rate": 1.4372881355932205e-05,
+      "loss": 0.3128,
+      "step": 424
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7103590658489138,
+      "learning_rate": 1.440677966101695e-05,
+      "loss": 0.2916,
+      "step": 425
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.177582810375724,
+      "learning_rate": 1.4440677966101698e-05,
+      "loss": 0.2787,
+      "step": 426
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2187682328759377,
+      "learning_rate": 1.4474576271186442e-05,
+      "loss": 0.2766,
+      "step": 427
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1786396709929452,
+      "learning_rate": 1.4508474576271187e-05,
+      "loss": 0.2958,
+      "step": 428
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1586043727922222,
+      "learning_rate": 1.4542372881355933e-05,
+      "loss": 0.2629,
+      "step": 429
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.374274393866389,
+      "learning_rate": 1.4576271186440678e-05,
+      "loss": 0.2725,
+      "step": 430
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1422580383010243,
+      "learning_rate": 1.4610169491525426e-05,
+      "loss": 0.2608,
+      "step": 431
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.362018091483847,
+      "learning_rate": 1.4644067796610171e-05,
+      "loss": 0.2686,
+      "step": 432
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3772004572389647,
+      "learning_rate": 1.4677966101694916e-05,
+      "loss": 0.2848,
+      "step": 433
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3697625449391713,
+      "learning_rate": 1.4711864406779662e-05,
+      "loss": 0.2618,
+      "step": 434
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2817523722761615,
+      "learning_rate": 1.4745762711864408e-05,
+      "loss": 0.281,
+      "step": 435
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5667550162037722,
+      "learning_rate": 1.4779661016949153e-05,
+      "loss": 0.3107,
+      "step": 436
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2154704540265984,
+      "learning_rate": 1.48135593220339e-05,
+      "loss": 0.2641,
+      "step": 437
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.601548252501335,
+      "learning_rate": 1.4847457627118644e-05,
+      "loss": 0.2749,
+      "step": 438
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3412392808454243,
+      "learning_rate": 1.4881355932203392e-05,
+      "loss": 0.2591,
+      "step": 439
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8745181492381693,
+      "learning_rate": 1.4915254237288137e-05,
+      "loss": 0.293,
+      "step": 440
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2032286439232478,
+      "learning_rate": 1.4949152542372882e-05,
+      "loss": 0.2488,
+      "step": 441
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8027795794557806,
+      "learning_rate": 1.4983050847457628e-05,
+      "loss": 0.2828,
+      "step": 442
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1544444706346058,
+      "learning_rate": 1.5016949152542374e-05,
+      "loss": 0.2894,
+      "step": 443
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.28630079163346,
+      "learning_rate": 1.505084745762712e-05,
+      "loss": 0.2826,
+      "step": 444
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3571363572134902,
+      "learning_rate": 1.5084745762711865e-05,
+      "loss": 0.2838,
+      "step": 445
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4725896993155925,
+      "learning_rate": 1.511864406779661e-05,
+      "loss": 0.2623,
+      "step": 446
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1861507973687793,
+      "learning_rate": 1.5152542372881358e-05,
+      "loss": 0.2835,
+      "step": 447
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8819167655356706,
+      "learning_rate": 1.5186440677966103e-05,
+      "loss": 0.2677,
+      "step": 448
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1347425370954585,
+      "learning_rate": 1.522033898305085e-05,
+      "loss": 0.2794,
+      "step": 449
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1153160184489386,
+      "learning_rate": 1.5254237288135594e-05,
+      "loss": 0.2658,
+      "step": 450
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1449747260362184,
+      "learning_rate": 1.528813559322034e-05,
+      "loss": 0.2524,
+      "step": 451
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.441613690575214,
+      "learning_rate": 1.5322033898305085e-05,
+      "loss": 0.2901,
+      "step": 452
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1305687396821262,
+      "learning_rate": 1.5355932203389833e-05,
+      "loss": 0.2702,
+      "step": 453
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.627737800353934,
+      "learning_rate": 1.5389830508474578e-05,
+      "loss": 0.2708,
+      "step": 454
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1744672296713026,
+      "learning_rate": 1.5423728813559326e-05,
+      "loss": 0.2516,
+      "step": 455
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2999579170148163,
+      "learning_rate": 1.545762711864407e-05,
+      "loss": 0.2595,
+      "step": 456
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.407224076389098,
+      "learning_rate": 1.5491525423728815e-05,
+      "loss": 0.2781,
+      "step": 457
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7356557602560885,
+      "learning_rate": 1.552542372881356e-05,
+      "loss": 0.274,
+      "step": 458
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7821775541345888,
+      "learning_rate": 1.5559322033898305e-05,
+      "loss": 0.2563,
+      "step": 459
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3388894848421589,
+      "learning_rate": 1.5593220338983053e-05,
+      "loss": 0.2848,
+      "step": 460
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.42437970488922,
+      "learning_rate": 1.5627118644067798e-05,
+      "loss": 0.261,
+      "step": 461
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2321061249146887,
+      "learning_rate": 1.5661016949152542e-05,
+      "loss": 0.2685,
+      "step": 462
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6009091932685735,
+      "learning_rate": 1.5694915254237287e-05,
+      "loss": 0.2772,
+      "step": 463
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2297059855573542,
+      "learning_rate": 1.5728813559322035e-05,
+      "loss": 0.2537,
+      "step": 464
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3453190050509531,
+      "learning_rate": 1.576271186440678e-05,
+      "loss": 0.2881,
+      "step": 465
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9238560454482667,
+      "learning_rate": 1.5796610169491528e-05,
+      "loss": 0.2704,
+      "step": 466
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5999053962529295,
+      "learning_rate": 1.5830508474576272e-05,
+      "loss": 0.2626,
+      "step": 467
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3191503928216304,
+      "learning_rate": 1.586440677966102e-05,
+      "loss": 0.277,
+      "step": 468
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5589440578069296,
+      "learning_rate": 1.5898305084745765e-05,
+      "loss": 0.2668,
+      "step": 469
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0608630567816568,
+      "learning_rate": 1.593220338983051e-05,
+      "loss": 0.2587,
+      "step": 470
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.278828909895488,
+      "learning_rate": 1.5966101694915255e-05,
+      "loss": 0.2757,
+      "step": 471
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2073777948507576,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.3151,
+      "step": 472
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3296212003545202,
+      "learning_rate": 1.6033898305084747e-05,
+      "loss": 0.289,
+      "step": 473
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1975562289235413,
+      "learning_rate": 1.6067796610169492e-05,
+      "loss": 0.269,
+      "step": 474
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.890410123709992,
+      "learning_rate": 1.6101694915254237e-05,
+      "loss": 0.248,
+      "step": 475
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2316296356173326,
+      "learning_rate": 1.6135593220338985e-05,
+      "loss": 0.2764,
+      "step": 476
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.158451224340742,
+      "learning_rate": 1.616949152542373e-05,
+      "loss": 0.2515,
+      "step": 477
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.386768235161273,
+      "learning_rate": 1.6203389830508474e-05,
+      "loss": 0.2766,
+      "step": 478
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9916586427606384,
+      "learning_rate": 1.6237288135593222e-05,
+      "loss": 0.2425,
+      "step": 479
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0663820855084396,
+      "learning_rate": 1.6271186440677967e-05,
+      "loss": 0.2849,
+      "step": 480
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2016145341855438,
+      "learning_rate": 1.6305084745762715e-05,
+      "loss": 0.2686,
+      "step": 481
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1723921344826131,
+      "learning_rate": 1.633898305084746e-05,
+      "loss": 0.2573,
+      "step": 482
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2111557499904375,
+      "learning_rate": 1.6372881355932204e-05,
+      "loss": 0.2532,
+      "step": 483
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.699199017647184,
+      "learning_rate": 1.640677966101695e-05,
+      "loss": 0.2902,
+      "step": 484
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2909577938024068,
+      "learning_rate": 1.6440677966101697e-05,
+      "loss": 0.2663,
+      "step": 485
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5571826032846792,
+      "learning_rate": 1.6474576271186442e-05,
+      "loss": 0.2645,
+      "step": 486
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1279016844343919,
+      "learning_rate": 1.6508474576271187e-05,
+      "loss": 0.2576,
+      "step": 487
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5591821887853188,
+      "learning_rate": 1.654237288135593e-05,
+      "loss": 0.2851,
+      "step": 488
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2092586739280375,
+      "learning_rate": 1.657627118644068e-05,
+      "loss": 0.2801,
+      "step": 489
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1463301267011572,
+      "learning_rate": 1.6610169491525424e-05,
+      "loss": 0.2563,
+      "step": 490
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2444534543779873,
+      "learning_rate": 1.6644067796610172e-05,
+      "loss": 0.2642,
+      "step": 491
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3314286719081205,
+      "learning_rate": 1.6677966101694917e-05,
+      "loss": 0.2806,
+      "step": 492
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4953543966574252,
+      "learning_rate": 1.671186440677966e-05,
+      "loss": 0.2812,
+      "step": 493
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.229405126380989,
+      "learning_rate": 1.674576271186441e-05,
+      "loss": 0.2702,
+      "step": 494
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.226387305787883,
+      "learning_rate": 1.6779661016949154e-05,
+      "loss": 0.2672,
+      "step": 495
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2279350727222604,
+      "learning_rate": 1.68135593220339e-05,
+      "loss": 0.2746,
+      "step": 496
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4752201164031455,
+      "learning_rate": 1.6847457627118647e-05,
+      "loss": 0.2615,
+      "step": 497
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5171752111549475,
+      "learning_rate": 1.6881355932203392e-05,
+      "loss": 0.2511,
+      "step": 498
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4085635220195363,
+      "learning_rate": 1.6915254237288136e-05,
+      "loss": 0.2503,
+      "step": 499
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5619747875254615,
+      "learning_rate": 1.694915254237288e-05,
+      "loss": 0.279,
+      "step": 500
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6055369708361142,
+      "learning_rate": 1.698305084745763e-05,
+      "loss": 0.2639,
+      "step": 501
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6065530819560707,
+      "learning_rate": 1.7016949152542374e-05,
+      "loss": 0.2877,
+      "step": 502
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3686957461363003,
+      "learning_rate": 1.705084745762712e-05,
+      "loss": 0.2679,
+      "step": 503
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3246185897098746,
+      "learning_rate": 1.7084745762711867e-05,
+      "loss": 0.2803,
+      "step": 504
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.601972430115074,
+      "learning_rate": 1.711864406779661e-05,
+      "loss": 0.2557,
+      "step": 505
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4960766298227883,
+      "learning_rate": 1.715254237288136e-05,
+      "loss": 0.2722,
+      "step": 506
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7165264426107494,
+      "learning_rate": 1.7186440677966104e-05,
+      "loss": 0.2747,
+      "step": 507
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.300412910903492,
+      "learning_rate": 1.722033898305085e-05,
+      "loss": 0.2704,
+      "step": 508
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4490955968513795,
+      "learning_rate": 1.7254237288135597e-05,
+      "loss": 0.2591,
+      "step": 509
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3759244398932398,
+      "learning_rate": 1.728813559322034e-05,
+      "loss": 0.2969,
+      "step": 510
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6898284753441082,
+      "learning_rate": 1.7322033898305086e-05,
+      "loss": 0.2443,
+      "step": 511
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4318962229777292,
+      "learning_rate": 1.735593220338983e-05,
+      "loss": 0.2513,
+      "step": 512
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.743666701146938,
+      "learning_rate": 1.7389830508474576e-05,
+      "loss": 0.2814,
+      "step": 513
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7193073886473862,
+      "learning_rate": 1.7423728813559324e-05,
+      "loss": 0.2888,
+      "step": 514
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.557692647375557,
+      "learning_rate": 1.745762711864407e-05,
+      "loss": 0.2702,
+      "step": 515
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3850957268156736,
+      "learning_rate": 1.7491525423728813e-05,
+      "loss": 0.2806,
+      "step": 516
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.687921295616476,
+      "learning_rate": 1.752542372881356e-05,
+      "loss": 0.246,
+      "step": 517
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1706695625942483,
+      "learning_rate": 1.7559322033898306e-05,
+      "loss": 0.251,
+      "step": 518
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.037585934141602,
+      "learning_rate": 1.7593220338983054e-05,
+      "loss": 0.27,
+      "step": 519
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3341641611216237,
+      "learning_rate": 1.76271186440678e-05,
+      "loss": 0.2909,
+      "step": 520
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1083063405567553,
+      "learning_rate": 1.7661016949152543e-05,
+      "loss": 0.273,
+      "step": 521
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1166108132403247,
+      "learning_rate": 1.769491525423729e-05,
+      "loss": 0.2647,
+      "step": 522
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4730148417394753,
+      "learning_rate": 1.7728813559322036e-05,
+      "loss": 0.2574,
+      "step": 523
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3796044596292965,
+      "learning_rate": 1.776271186440678e-05,
+      "loss": 0.2809,
+      "step": 524
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.668912240239683,
+      "learning_rate": 1.7796610169491526e-05,
+      "loss": 0.2896,
+      "step": 525
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6221075546474033,
+      "learning_rate": 1.7830508474576274e-05,
+      "loss": 0.2429,
+      "step": 526
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.314326489119452,
+      "learning_rate": 1.7864406779661018e-05,
+      "loss": 0.2974,
+      "step": 527
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4011654180586948,
+      "learning_rate": 1.7898305084745763e-05,
+      "loss": 0.2805,
+      "step": 528
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.456593465147238,
+      "learning_rate": 1.7932203389830508e-05,
+      "loss": 0.271,
+      "step": 529
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4153772547557295,
+      "learning_rate": 1.7966101694915256e-05,
+      "loss": 0.2993,
+      "step": 530
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.460962611843914,
+      "learning_rate": 1.8e-05,
+      "loss": 0.288,
+      "step": 531
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3671451429644492,
+      "learning_rate": 1.803389830508475e-05,
+      "loss": 0.2591,
+      "step": 532
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.558548773828584,
+      "learning_rate": 1.8067796610169493e-05,
+      "loss": 0.2689,
+      "step": 533
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5127491344620718,
+      "learning_rate": 1.810169491525424e-05,
+      "loss": 0.3259,
+      "step": 534
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5486731950661254,
+      "learning_rate": 1.8135593220338986e-05,
+      "loss": 0.2712,
+      "step": 535
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3193510633104664,
+      "learning_rate": 1.816949152542373e-05,
+      "loss": 0.2838,
+      "step": 536
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0997689513888096,
+      "learning_rate": 1.8203389830508475e-05,
+      "loss": 0.2525,
+      "step": 537
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2756030235865132,
+      "learning_rate": 1.823728813559322e-05,
+      "loss": 0.307,
+      "step": 538
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4687241798166086,
+      "learning_rate": 1.8271186440677968e-05,
+      "loss": 0.2428,
+      "step": 539
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3399067424295357,
+      "learning_rate": 1.8305084745762713e-05,
+      "loss": 0.2732,
+      "step": 540
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6084487910737866,
+      "learning_rate": 1.8338983050847458e-05,
+      "loss": 0.2552,
+      "step": 541
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2944041185222261,
+      "learning_rate": 1.8372881355932202e-05,
+      "loss": 0.2804,
+      "step": 542
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2566959055725666,
+      "learning_rate": 1.840677966101695e-05,
+      "loss": 0.259,
+      "step": 543
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2933124442569186,
+      "learning_rate": 1.8440677966101695e-05,
+      "loss": 0.2676,
+      "step": 544
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3273078884878304,
+      "learning_rate": 1.8474576271186443e-05,
+      "loss": 0.2719,
+      "step": 545
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.263916893620895,
+      "learning_rate": 1.8508474576271188e-05,
+      "loss": 0.286,
+      "step": 546
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1352925308401638,
+      "learning_rate": 1.8542372881355936e-05,
+      "loss": 0.2449,
+      "step": 547
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4287019035022674,
+      "learning_rate": 1.857627118644068e-05,
+      "loss": 0.2854,
+      "step": 548
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.536114169133046,
+      "learning_rate": 1.8610169491525425e-05,
+      "loss": 0.2749,
+      "step": 549
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3636384713719114,
+      "learning_rate": 1.864406779661017e-05,
+      "loss": 0.2798,
+      "step": 550
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1282443226800742,
+      "learning_rate": 1.8677966101694918e-05,
+      "loss": 0.2545,
+      "step": 551
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4372325985492267,
+      "learning_rate": 1.8711864406779663e-05,
+      "loss": 0.2764,
+      "step": 552
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.431962567950389,
+      "learning_rate": 1.8745762711864407e-05,
+      "loss": 0.2535,
+      "step": 553
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2145997199986456,
+      "learning_rate": 1.8779661016949152e-05,
+      "loss": 0.2566,
+      "step": 554
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1352608835127773,
+      "learning_rate": 1.88135593220339e-05,
+      "loss": 0.2535,
+      "step": 555
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2210496354415843,
+      "learning_rate": 1.8847457627118645e-05,
+      "loss": 0.2643,
+      "step": 556
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1232924683009815,
+      "learning_rate": 1.8881355932203393e-05,
+      "loss": 0.2627,
+      "step": 557
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4024455416720127,
+      "learning_rate": 1.8915254237288138e-05,
+      "loss": 0.2525,
+      "step": 558
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3887487502150935,
+      "learning_rate": 1.8949152542372882e-05,
+      "loss": 0.2758,
+      "step": 559
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.232504365515857,
+      "learning_rate": 1.898305084745763e-05,
+      "loss": 0.2406,
+      "step": 560
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3940857376689135,
+      "learning_rate": 1.9016949152542375e-05,
+      "loss": 0.2558,
+      "step": 561
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.345932352677997,
+      "learning_rate": 1.905084745762712e-05,
+      "loss": 0.2626,
+      "step": 562
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1157309455332585,
+      "learning_rate": 1.9084745762711868e-05,
+      "loss": 0.2632,
+      "step": 563
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.156636964733872,
+      "learning_rate": 1.9118644067796613e-05,
+      "loss": 0.2781,
+      "step": 564
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0729683625503257,
+      "learning_rate": 1.9152542372881357e-05,
+      "loss": 0.2457,
+      "step": 565
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1647734275933048,
+      "learning_rate": 1.9186440677966102e-05,
+      "loss": 0.2393,
+      "step": 566
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5138294469581877,
+      "learning_rate": 1.9220338983050847e-05,
+      "loss": 0.2876,
+      "step": 567
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.323351272486395,
+      "learning_rate": 1.9254237288135595e-05,
+      "loss": 0.2396,
+      "step": 568
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9451745243471656,
+      "learning_rate": 1.928813559322034e-05,
+      "loss": 0.2514,
+      "step": 569
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2208860898726068,
+      "learning_rate": 1.9322033898305087e-05,
+      "loss": 0.2516,
+      "step": 570
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3855491705683636,
+      "learning_rate": 1.9355932203389832e-05,
+      "loss": 0.2758,
+      "step": 571
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.429772998796651,
+      "learning_rate": 1.938983050847458e-05,
+      "loss": 0.2356,
+      "step": 572
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7057138660762692,
+      "learning_rate": 1.9423728813559325e-05,
+      "loss": 0.2283,
+      "step": 573
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.51201577923191,
+      "learning_rate": 1.945762711864407e-05,
+      "loss": 0.2671,
+      "step": 574
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2651595136689437,
+      "learning_rate": 1.9491525423728814e-05,
+      "loss": 0.2801,
+      "step": 575
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3249553121989968,
+      "learning_rate": 1.9525423728813562e-05,
+      "loss": 0.2787,
+      "step": 576
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.74385495359413,
+      "learning_rate": 1.9559322033898307e-05,
+      "loss": 0.2608,
+      "step": 577
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6741267385582472,
+      "learning_rate": 1.9593220338983052e-05,
+      "loss": 0.258,
+      "step": 578
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4888420560673257,
+      "learning_rate": 1.9627118644067796e-05,
+      "loss": 0.281,
+      "step": 579
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4190781041419613,
+      "learning_rate": 1.9661016949152545e-05,
+      "loss": 0.2438,
+      "step": 580
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.12210293134904,
+      "learning_rate": 1.969491525423729e-05,
+      "loss": 0.2877,
+      "step": 581
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.441475622095831,
+      "learning_rate": 1.9728813559322034e-05,
+      "loss": 0.253,
+      "step": 582
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.745675132840984,
+      "learning_rate": 1.9762711864406782e-05,
+      "loss": 0.2806,
+      "step": 583
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3836639201458314,
+      "learning_rate": 1.9796610169491527e-05,
+      "loss": 0.3177,
+      "step": 584
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.30532147466716,
+      "learning_rate": 1.9830508474576275e-05,
+      "loss": 0.2846,
+      "step": 585
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5409660828860403,
+      "learning_rate": 1.986440677966102e-05,
+      "loss": 0.2681,
+      "step": 586
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.731352615733341,
+      "learning_rate": 1.9898305084745764e-05,
+      "loss": 0.2567,
+      "step": 587
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.475042781560892,
+      "learning_rate": 1.9932203389830512e-05,
+      "loss": 0.2628,
+      "step": 588
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.0834959806057185,
+      "learning_rate": 1.9966101694915257e-05,
+      "loss": 0.24,
+      "step": 589
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3873201495539649,
+      "learning_rate": 2e-05,
+      "loss": 0.26,
+      "step": 590
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3568449606917727,
+      "learning_rate": 1.999999986437472e-05,
+      "loss": 0.279,
+      "step": 591
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4736919604577718,
+      "learning_rate": 1.9999999457498875e-05,
+      "loss": 0.2684,
+      "step": 592
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2388398729176897,
+      "learning_rate": 1.9999998779372483e-05,
+      "loss": 0.296,
+      "step": 593
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.266582495935962,
+      "learning_rate": 1.9999997829995557e-05,
+      "loss": 0.2539,
+      "step": 594
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.563096295904235,
+      "learning_rate": 1.9999996609368124e-05,
+      "loss": 0.2957,
+      "step": 595
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5591054305818515,
+      "learning_rate": 1.9999995117490224e-05,
+      "loss": 0.2662,
+      "step": 596
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.65621574378536,
+      "learning_rate": 1.9999993354361887e-05,
+      "loss": 0.2403,
+      "step": 597
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2548699939118981,
+      "learning_rate": 1.9999991319983162e-05,
+      "loss": 0.2813,
+      "step": 598
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8410217890854863,
+      "learning_rate": 1.9999989014354117e-05,
+      "loss": 0.2371,
+      "step": 599
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4153342484952254,
+      "learning_rate": 1.9999986437474797e-05,
+      "loss": 0.2565,
+      "step": 600
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.10289019805283,
+      "learning_rate": 1.9999983589345282e-05,
+      "loss": 0.2698,
+      "step": 601
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.453681957989118,
+      "learning_rate": 1.9999980469965646e-05,
+      "loss": 0.2506,
+      "step": 602
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.948473664806546,
+      "learning_rate": 1.999997707933598e-05,
+      "loss": 0.2494,
+      "step": 603
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3618436212411482,
+      "learning_rate": 1.9999973417456367e-05,
+      "loss": 0.2644,
+      "step": 604
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3724384206589046,
+      "learning_rate": 1.999996948432691e-05,
+      "loss": 0.295,
+      "step": 605
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4885167672678044,
+      "learning_rate": 1.999996527994772e-05,
+      "loss": 0.2799,
+      "step": 606
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.177622041217393,
+      "learning_rate": 1.9999960804318904e-05,
+      "loss": 0.2878,
+      "step": 607
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0149772969197515,
+      "learning_rate": 1.999995605744059e-05,
+      "loss": 0.257,
+      "step": 608
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.691709339113414,
+      "learning_rate": 1.9999951039312897e-05,
+      "loss": 0.2524,
+      "step": 609
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0999445528030725,
+      "learning_rate": 1.999994574993597e-05,
+      "loss": 0.2612,
+      "step": 610
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5337654360480533,
+      "learning_rate": 1.999994018930995e-05,
+      "loss": 0.2948,
+      "step": 611
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1245129713817184,
+      "learning_rate": 1.9999934357434986e-05,
+      "loss": 0.2783,
+      "step": 612
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.051000583585932,
+      "learning_rate": 1.9999928254311242e-05,
+      "loss": 0.2663,
+      "step": 613
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6065272631126895,
+      "learning_rate": 1.9999921879938875e-05,
+      "loss": 0.2648,
+      "step": 614
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.040871656632654,
+      "learning_rate": 1.9999915234318064e-05,
+      "loss": 0.258,
+      "step": 615
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0489036653596298,
+      "learning_rate": 1.9999908317448985e-05,
+      "loss": 0.2619,
+      "step": 616
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2406900738256312,
+      "learning_rate": 1.9999901129331832e-05,
+      "loss": 0.2688,
+      "step": 617
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3093334914999668,
+      "learning_rate": 1.9999893669966794e-05,
+      "loss": 0.2856,
+      "step": 618
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0150981610834626,
+      "learning_rate": 1.9999885939354077e-05,
+      "loss": 0.267,
+      "step": 619
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1050649685427374,
+      "learning_rate": 1.9999877937493886e-05,
+      "loss": 0.2558,
+      "step": 620
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0775846893782826,
+      "learning_rate": 1.9999869664386443e-05,
+      "loss": 0.2675,
+      "step": 621
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0419452826452065,
+      "learning_rate": 1.999986112003197e-05,
+      "loss": 0.2645,
+      "step": 622
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3183338349176241,
+      "learning_rate": 1.99998523044307e-05,
+      "loss": 0.3004,
+      "step": 623
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9943967830586373,
+      "learning_rate": 1.999984321758287e-05,
+      "loss": 0.3,
+      "step": 624
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.498860798184339,
+      "learning_rate": 1.999983385948873e-05,
+      "loss": 0.2547,
+      "step": 625
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.537978887587886,
+      "learning_rate": 1.9999824230148532e-05,
+      "loss": 0.2938,
+      "step": 626
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9265795045788969,
+      "learning_rate": 1.999981432956254e-05,
+      "loss": 0.2698,
+      "step": 627
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.92643007386415,
+      "learning_rate": 1.999980415773101e-05,
+      "loss": 0.3034,
+      "step": 628
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0860148365443336,
+      "learning_rate": 1.9999793714654236e-05,
+      "loss": 0.2617,
+      "step": 629
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.232951628312408,
+      "learning_rate": 1.9999783000332486e-05,
+      "loss": 0.2436,
+      "step": 630
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6804316633229621,
+      "learning_rate": 1.9999772014766062e-05,
+      "loss": 0.305,
+      "step": 631
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0772468232195704,
+      "learning_rate": 1.9999760757955258e-05,
+      "loss": 0.284,
+      "step": 632
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4342896278895008,
+      "learning_rate": 1.9999749229900376e-05,
+      "loss": 0.2677,
+      "step": 633
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8471599254915816,
+      "learning_rate": 1.9999737430601734e-05,
+      "loss": 0.2557,
+      "step": 634
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.281316376095995,
+      "learning_rate": 1.9999725360059648e-05,
+      "loss": 0.2817,
+      "step": 635
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3402740898159577,
+      "learning_rate": 1.9999713018274444e-05,
+      "loss": 0.265,
+      "step": 636
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3958741995969437,
+      "learning_rate": 1.9999700405246462e-05,
+      "loss": 0.3047,
+      "step": 637
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4108555514675845,
+      "learning_rate": 1.9999687520976043e-05,
+      "loss": 0.2721,
+      "step": 638
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2241788885846367,
+      "learning_rate": 1.9999674365463532e-05,
+      "loss": 0.2645,
+      "step": 639
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.211714854940445,
+      "learning_rate": 1.999966093870929e-05,
+      "loss": 0.2947,
+      "step": 640
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2759128182319426,
+      "learning_rate": 1.999964724071368e-05,
+      "loss": 0.2615,
+      "step": 641
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.437640733114652,
+      "learning_rate": 1.999963327147708e-05,
+      "loss": 0.2614,
+      "step": 642
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4274344790032485,
+      "learning_rate": 1.9999619030999853e-05,
+      "loss": 0.2999,
+      "step": 643
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1171499634893127,
+      "learning_rate": 1.9999604519282403e-05,
+      "loss": 0.2506,
+      "step": 644
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3503945628596226,
+      "learning_rate": 1.999958973632511e-05,
+      "loss": 0.2618,
+      "step": 645
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.259518347435971,
+      "learning_rate": 1.9999574682128385e-05,
+      "loss": 0.2696,
+      "step": 646
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2025649744782203,
+      "learning_rate": 1.999955935669263e-05,
+      "loss": 0.2773,
+      "step": 647
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3748457996542727,
+      "learning_rate": 1.9999543760018264e-05,
+      "loss": 0.2731,
+      "step": 648
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.286447108152475,
+      "learning_rate": 1.999952789210571e-05,
+      "loss": 0.2513,
+      "step": 649
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1625739396436945,
+      "learning_rate": 1.999951175295539e-05,
+      "loss": 0.2805,
+      "step": 650
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8934326598683626,
+      "learning_rate": 1.9999495342567754e-05,
+      "loss": 0.2703,
+      "step": 651
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2766979998578325,
+      "learning_rate": 1.999947866094324e-05,
+      "loss": 0.2709,
+      "step": 652
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1893920363930803,
+      "learning_rate": 1.999946170808231e-05,
+      "loss": 0.27,
+      "step": 653
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4471066279469218,
+      "learning_rate": 1.9999444483985408e-05,
+      "loss": 0.2706,
+      "step": 654
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0597340037872298,
+      "learning_rate": 1.9999426988653012e-05,
+      "loss": 0.2862,
+      "step": 655
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1600664080975713,
+      "learning_rate": 1.9999409222085596e-05,
+      "loss": 0.2725,
+      "step": 656
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2152493757257978,
+      "learning_rate": 1.9999391184283638e-05,
+      "loss": 0.2869,
+      "step": 657
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2957871045516876,
+      "learning_rate": 1.999937287524763e-05,
+      "loss": 0.2851,
+      "step": 658
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1059643575300484,
+      "learning_rate": 1.9999354294978066e-05,
+      "loss": 0.2534,
+      "step": 659
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0210468864347002,
+      "learning_rate": 1.9999335443475452e-05,
+      "loss": 0.2608,
+      "step": 660
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2949778276813804,
+      "learning_rate": 1.9999316320740302e-05,
+      "loss": 0.2689,
+      "step": 661
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1647843952730665,
+      "learning_rate": 1.9999296926773133e-05,
+      "loss": 0.2436,
+      "step": 662
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3682879718976364,
+      "learning_rate": 1.9999277261574468e-05,
+      "loss": 0.2564,
+      "step": 663
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2468871282706768,
+      "learning_rate": 1.999925732514484e-05,
+      "loss": 0.2714,
+      "step": 664
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7519741264204172,
+      "learning_rate": 1.99992371174848e-05,
+      "loss": 0.2612,
+      "step": 665
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.161839164666825,
+      "learning_rate": 1.999921663859488e-05,
+      "loss": 0.2479,
+      "step": 666
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2751282318849002,
+      "learning_rate": 1.999919588847565e-05,
+      "loss": 0.2568,
+      "step": 667
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0965692887004719,
+      "learning_rate": 1.999917486712766e-05,
+      "loss": 0.269,
+      "step": 668
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0701031723437957,
+      "learning_rate": 1.9999153574551492e-05,
+      "loss": 0.2679,
+      "step": 669
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2142348695150131,
+      "learning_rate": 1.999913201074772e-05,
+      "loss": 0.265,
+      "step": 670
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.06668955891612,
+      "learning_rate": 1.9999110175716924e-05,
+      "loss": 0.2521,
+      "step": 671
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0533089187255256,
+      "learning_rate": 1.99990880694597e-05,
+      "loss": 0.256,
+      "step": 672
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0623846929343919,
+      "learning_rate": 1.9999065691976648e-05,
+      "loss": 0.2956,
+      "step": 673
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9917831191421542,
+      "learning_rate": 1.9999043043268375e-05,
+      "loss": 0.2521,
+      "step": 674
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0909337781250295,
+      "learning_rate": 1.9999020123335496e-05,
+      "loss": 0.2541,
+      "step": 675
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2088864473371073,
+      "learning_rate": 1.9998996932178625e-05,
+      "loss": 0.2353,
+      "step": 676
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.267002024540383,
+      "learning_rate": 1.9998973469798404e-05,
+      "loss": 0.2871,
+      "step": 677
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.26664688395502,
+      "learning_rate": 1.9998949736195464e-05,
+      "loss": 0.2697,
+      "step": 678
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1258218909222184,
+      "learning_rate": 1.9998925731370443e-05,
+      "loss": 0.2767,
+      "step": 679
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4333128018170362,
+      "learning_rate": 1.9998901455324e-05,
+      "loss": 0.272,
+      "step": 680
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1539615363566418,
+      "learning_rate": 1.999887690805679e-05,
+      "loss": 0.2703,
+      "step": 681
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3055896441511934,
+      "learning_rate": 1.9998852089569473e-05,
+      "loss": 0.2805,
+      "step": 682
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9049653729869355,
+      "learning_rate": 1.9998826999862736e-05,
+      "loss": 0.2588,
+      "step": 683
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.638495294480143,
+      "learning_rate": 1.9998801638937245e-05,
+      "loss": 0.2485,
+      "step": 684
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2575064722605738,
+      "learning_rate": 1.9998776006793702e-05,
+      "loss": 0.2661,
+      "step": 685
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1924520993769014,
+      "learning_rate": 1.999875010343279e-05,
+      "loss": 0.2907,
+      "step": 686
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0528956283908704,
+      "learning_rate": 1.999872392885522e-05,
+      "loss": 0.2597,
+      "step": 687
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0761853664702166,
+      "learning_rate": 1.9998697483061698e-05,
+      "loss": 0.2267,
+      "step": 688
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7045786645036725,
+      "learning_rate": 1.9998670766052942e-05,
+      "loss": 0.2515,
+      "step": 689
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0632182594277078,
+      "learning_rate": 1.9998643777829674e-05,
+      "loss": 0.2673,
+      "step": 690
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3976256225702135,
+      "learning_rate": 1.9998616518392633e-05,
+      "loss": 0.2639,
+      "step": 691
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0113588487491112,
+      "learning_rate": 1.999858898774255e-05,
+      "loss": 0.2794,
+      "step": 692
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0553084109359292,
+      "learning_rate": 1.999856118588018e-05,
+      "loss": 0.2411,
+      "step": 693
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0623825667352056,
+      "learning_rate": 1.999853311280627e-05,
+      "loss": 0.2567,
+      "step": 694
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6341578895660698,
+      "learning_rate": 1.9998504768521588e-05,
+      "loss": 0.2903,
+      "step": 695
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5184615805888484,
+      "learning_rate": 1.99984761530269e-05,
+      "loss": 0.2781,
+      "step": 696
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3640734092403373,
+      "learning_rate": 1.9998447266322974e-05,
+      "loss": 0.2571,
+      "step": 697
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0545124494588674,
+      "learning_rate": 1.9998418108410606e-05,
+      "loss": 0.2442,
+      "step": 698
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2493399526273568,
+      "learning_rate": 1.9998388679290583e-05,
+      "loss": 0.2984,
+      "step": 699
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1295762074044364,
+      "learning_rate": 1.9998358978963702e-05,
+      "loss": 0.2512,
+      "step": 700
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1603953126986037,
+      "learning_rate": 1.9998329007430767e-05,
+      "loss": 0.2696,
+      "step": 701
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2509835069608402,
+      "learning_rate": 1.9998298764692596e-05,
+      "loss": 0.2651,
+      "step": 702
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3348261112019737,
+      "learning_rate": 1.9998268250750006e-05,
+      "loss": 0.2583,
+      "step": 703
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3375545121799972,
+      "learning_rate": 1.9998237465603822e-05,
+      "loss": 0.28,
+      "step": 704
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.363579686038191,
+      "learning_rate": 1.9998206409254886e-05,
+      "loss": 0.2937,
+      "step": 705
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6025082506089567,
+      "learning_rate": 1.9998175081704035e-05,
+      "loss": 0.2634,
+      "step": 706
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2722761240387512,
+      "learning_rate": 1.9998143482952117e-05,
+      "loss": 0.2599,
+      "step": 707
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.599998265070209,
+      "learning_rate": 1.9998111612999995e-05,
+      "loss": 0.2505,
+      "step": 708
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.317103804306265,
+      "learning_rate": 1.999807947184853e-05,
+      "loss": 0.2807,
+      "step": 709
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3710100936702176,
+      "learning_rate": 1.99980470594986e-05,
+      "loss": 0.2641,
+      "step": 710
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.108376334270745,
+      "learning_rate": 1.9998014375951073e-05,
+      "loss": 0.2403,
+      "step": 711
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.252535265060163,
+      "learning_rate": 1.999798142120684e-05,
+      "loss": 0.2769,
+      "step": 712
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3742660328462404,
+      "learning_rate": 1.99979481952668e-05,
+      "loss": 0.2906,
+      "step": 713
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.644541022537266,
+      "learning_rate": 1.999791469813185e-05,
+      "loss": 0.257,
+      "step": 714
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0749151196722126,
+      "learning_rate": 1.9997880929802895e-05,
+      "loss": 0.262,
+      "step": 715
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1730231560972142,
+      "learning_rate": 1.999784689028086e-05,
+      "loss": 0.2644,
+      "step": 716
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6294158398432395,
+      "learning_rate": 1.999781257956666e-05,
+      "loss": 0.2706,
+      "step": 717
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0705353040911285,
+      "learning_rate": 1.999777799766123e-05,
+      "loss": 0.265,
+      "step": 718
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1518326861280468,
+      "learning_rate": 1.9997743144565513e-05,
+      "loss": 0.2823,
+      "step": 719
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.17251836953955,
+      "learning_rate": 1.999770802028044e-05,
+      "loss": 0.2623,
+      "step": 720
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1856563147370942,
+      "learning_rate": 1.9997672624806976e-05,
+      "loss": 0.2777,
+      "step": 721
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4003250521694823,
+      "learning_rate": 1.999763695814608e-05,
+      "loss": 0.2464,
+      "step": 722
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5113897733883497,
+      "learning_rate": 1.9997601020298713e-05,
+      "loss": 0.2613,
+      "step": 723
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2219147153367094,
+      "learning_rate": 1.9997564811265854e-05,
+      "loss": 0.2616,
+      "step": 724
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4067252099576408,
+      "learning_rate": 1.999752833104849e-05,
+      "loss": 0.2842,
+      "step": 725
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0208594719205228,
+      "learning_rate": 1.9997491579647595e-05,
+      "loss": 0.2505,
+      "step": 726
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1036370771367006,
+      "learning_rate": 1.9997454557064185e-05,
+      "loss": 0.2498,
+      "step": 727
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7207498654087374,
+      "learning_rate": 1.9997417263299256e-05,
+      "loss": 0.2713,
+      "step": 728
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3992568424381269,
+      "learning_rate": 1.999737969835381e-05,
+      "loss": 0.2955,
+      "step": 729
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9982495451376947,
+      "learning_rate": 1.9997341862228886e-05,
+      "loss": 0.2583,
+      "step": 730
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3960847560767105,
+      "learning_rate": 1.999730375492549e-05,
+      "loss": 0.2939,
+      "step": 731
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.214687864500847,
+      "learning_rate": 1.999726537644467e-05,
+      "loss": 0.254,
+      "step": 732
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4135176505419116,
+      "learning_rate": 1.9997226726787462e-05,
+      "loss": 0.3036,
+      "step": 733
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3404019047194708,
+      "learning_rate": 1.9997187805954916e-05,
+      "loss": 0.2803,
+      "step": 734
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0201205942546145,
+      "learning_rate": 1.999714861394808e-05,
+      "loss": 0.273,
+      "step": 735
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0396421943283112,
+      "learning_rate": 1.999710915076803e-05,
+      "loss": 0.2595,
+      "step": 736
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0398657233769981,
+      "learning_rate": 1.9997069416415824e-05,
+      "loss": 0.2853,
+      "step": 737
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0920534147148957,
+      "learning_rate": 1.9997029410892546e-05,
+      "loss": 0.241,
+      "step": 738
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0928042063839294,
+      "learning_rate": 1.9996989134199287e-05,
+      "loss": 0.2144,
+      "step": 739
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0198803572927333,
+      "learning_rate": 1.9996948586337127e-05,
+      "loss": 0.2572,
+      "step": 740
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4319185641684506,
+      "learning_rate": 1.9996907767307175e-05,
+      "loss": 0.2694,
+      "step": 741
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0858520979144273,
+      "learning_rate": 1.9996866677110534e-05,
+      "loss": 0.2594,
+      "step": 742
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3255623969997914,
+      "learning_rate": 1.999682531574832e-05,
+      "loss": 0.2538,
+      "step": 743
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0108943423155112,
+      "learning_rate": 1.9996783683221652e-05,
+      "loss": 0.2516,
+      "step": 744
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4094537154588183,
+      "learning_rate": 1.9996741779531665e-05,
+      "loss": 0.2707,
+      "step": 745
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1793015053224885,
+      "learning_rate": 1.9996699604679493e-05,
+      "loss": 0.2646,
+      "step": 746
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0792637757557646,
+      "learning_rate": 1.9996657158666276e-05,
+      "loss": 0.2478,
+      "step": 747
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.039527097370698,
+      "learning_rate": 1.999661444149317e-05,
+      "loss": 0.2534,
+      "step": 748
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2559400175483508,
+      "learning_rate": 1.9996571453161338e-05,
+      "loss": 0.2882,
+      "step": 749
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1366971909735748,
+      "learning_rate": 1.9996528193671936e-05,
+      "loss": 0.2505,
+      "step": 750
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3078857802825534,
+      "learning_rate": 1.9996484663026143e-05,
+      "loss": 0.2835,
+      "step": 751
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0829375306862181,
+      "learning_rate": 1.999644086122514e-05,
+      "loss": 0.2479,
+      "step": 752
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4273810049424516,
+      "learning_rate": 1.999639678827011e-05,
+      "loss": 0.2669,
+      "step": 753
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3117200257674768,
+      "learning_rate": 1.9996352444162257e-05,
+      "loss": 0.2834,
+      "step": 754
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3033840703007538,
+      "learning_rate": 1.999630782890278e-05,
+      "loss": 0.24,
+      "step": 755
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4419353437645621,
+      "learning_rate": 1.9996262942492882e-05,
+      "loss": 0.2608,
+      "step": 756
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0632655417967007,
+      "learning_rate": 1.9996217784933794e-05,
+      "loss": 0.293,
+      "step": 757
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.138628289340763,
+      "learning_rate": 1.999617235622673e-05,
+      "loss": 0.2835,
+      "step": 758
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8702915385259948,
+      "learning_rate": 1.999612665637293e-05,
+      "loss": 0.2692,
+      "step": 759
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.130927538844785,
+      "learning_rate": 1.9996080685373628e-05,
+      "loss": 0.2563,
+      "step": 760
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1131329974735835,
+      "learning_rate": 1.999603444323007e-05,
+      "loss": 0.2744,
+      "step": 761
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5508343576824948,
+      "learning_rate": 1.9995987929943518e-05,
+      "loss": 0.2657,
+      "step": 762
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3218831953876735,
+      "learning_rate": 1.9995941145515224e-05,
+      "loss": 0.2584,
+      "step": 763
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1460227432051737,
+      "learning_rate": 1.9995894089946466e-05,
+      "loss": 0.2883,
+      "step": 764
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0196084487669637,
+      "learning_rate": 1.9995846763238514e-05,
+      "loss": 0.2501,
+      "step": 765
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0954833712086312,
+      "learning_rate": 1.9995799165392653e-05,
+      "loss": 0.2882,
+      "step": 766
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2774372509323548,
+      "learning_rate": 1.9995751296410176e-05,
+      "loss": 0.2535,
+      "step": 767
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.103024655690764,
+      "learning_rate": 1.9995703156292382e-05,
+      "loss": 0.2699,
+      "step": 768
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.126010661023456,
+      "learning_rate": 1.9995654745040575e-05,
+      "loss": 0.2746,
+      "step": 769
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2036940841199122,
+      "learning_rate": 1.999560606265607e-05,
+      "loss": 0.2846,
+      "step": 770
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.053928772845716,
+      "learning_rate": 1.999555710914018e-05,
+      "loss": 0.2536,
+      "step": 771
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1459541009748426,
+      "learning_rate": 1.9995507884494244e-05,
+      "loss": 0.2503,
+      "step": 772
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2796107407868114,
+      "learning_rate": 1.999545838871959e-05,
+      "loss": 0.2575,
+      "step": 773
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.313315822502975,
+      "learning_rate": 1.9995408621817566e-05,
+      "loss": 0.2616,
+      "step": 774
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6089597975605014,
+      "learning_rate": 1.9995358583789514e-05,
+      "loss": 0.245,
+      "step": 775
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1189952245805992,
+      "learning_rate": 1.99953082746368e-05,
+      "loss": 0.2719,
+      "step": 776
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1676561428296084,
+      "learning_rate": 1.9995257694360778e-05,
+      "loss": 0.2515,
+      "step": 777
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2390945203221548,
+      "learning_rate": 1.9995206842962833e-05,
+      "loss": 0.2815,
+      "step": 778
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.284354717320916,
+      "learning_rate": 1.9995155720444336e-05,
+      "loss": 0.2474,
+      "step": 779
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2223633934856006,
+      "learning_rate": 1.9995104326806675e-05,
+      "loss": 0.2695,
+      "step": 780
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1957041076695898,
+      "learning_rate": 1.9995052662051244e-05,
+      "loss": 0.2401,
+      "step": 781
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6931627034823373,
+      "learning_rate": 1.999500072617945e-05,
+      "loss": 0.2469,
+      "step": 782
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1597265079302603,
+      "learning_rate": 1.999494851919269e-05,
+      "loss": 0.2868,
+      "step": 783
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8762144776870082,
+      "learning_rate": 1.999489604109239e-05,
+      "loss": 0.2919,
+      "step": 784
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1508642549126946,
+      "learning_rate": 1.999484329187997e-05,
+      "loss": 0.2577,
+      "step": 785
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1939915989038365,
+      "learning_rate": 1.9994790271556862e-05,
+      "loss": 0.2597,
+      "step": 786
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1376437711036782,
+      "learning_rate": 1.9994736980124502e-05,
+      "loss": 0.2614,
+      "step": 787
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1683273366283589,
+      "learning_rate": 1.9994683417584336e-05,
+      "loss": 0.2736,
+      "step": 788
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6975838851394127,
+      "learning_rate": 1.999462958393782e-05,
+      "loss": 0.2642,
+      "step": 789
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3563670815198583,
+      "learning_rate": 1.999457547918641e-05,
+      "loss": 0.2493,
+      "step": 790
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1854453853936364,
+      "learning_rate": 1.999452110333158e-05,
+      "loss": 0.2397,
+      "step": 791
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5034728936630222,
+      "learning_rate": 1.9994466456374796e-05,
+      "loss": 0.2834,
+      "step": 792
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9981359156760405,
+      "learning_rate": 1.9994411538317546e-05,
+      "loss": 0.2322,
+      "step": 793
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2749037646530144,
+      "learning_rate": 1.999435634916132e-05,
+      "loss": 0.2485,
+      "step": 794
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1573639241050016,
+      "learning_rate": 1.9994300888907613e-05,
+      "loss": 0.2847,
+      "step": 795
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9231922613268816,
+      "learning_rate": 1.999424515755793e-05,
+      "loss": 0.2389,
+      "step": 796
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.092813811557751,
+      "learning_rate": 1.9994189155113778e-05,
+      "loss": 0.2567,
+      "step": 797
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.097683225891256,
+      "learning_rate": 1.9994132881576685e-05,
+      "loss": 0.2702,
+      "step": 798
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1363590369891523,
+      "learning_rate": 1.9994076336948175e-05,
+      "loss": 0.2784,
+      "step": 799
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8879204954807518,
+      "learning_rate": 1.999401952122978e-05,
+      "loss": 0.2435,
+      "step": 800
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.033548685282584,
+      "learning_rate": 1.9993962434423037e-05,
+      "loss": 0.2656,
+      "step": 801
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1875030010672916,
+      "learning_rate": 1.99939050765295e-05,
+      "loss": 0.2726,
+      "step": 802
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.079155108315839,
+      "learning_rate": 1.9993847447550722e-05,
+      "loss": 0.2373,
+      "step": 803
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1246899751217057,
+      "learning_rate": 1.9993789547488268e-05,
+      "loss": 0.2584,
+      "step": 804
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0258829824673186,
+      "learning_rate": 1.999373137634371e-05,
+      "loss": 0.2619,
+      "step": 805
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.105454260644435,
+      "learning_rate": 1.9993672934118625e-05,
+      "loss": 0.3153,
+      "step": 806
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.971938599700853,
+      "learning_rate": 1.9993614220814594e-05,
+      "loss": 0.2531,
+      "step": 807
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1308239034584113,
+      "learning_rate": 1.9993555236433216e-05,
+      "loss": 0.2378,
+      "step": 808
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1821286909927475,
+      "learning_rate": 1.9993495980976084e-05,
+      "loss": 0.2461,
+      "step": 809
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0491796907842892,
+      "learning_rate": 1.9993436454444814e-05,
+      "loss": 0.2863,
+      "step": 810
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1794335903141098,
+      "learning_rate": 1.999337665684101e-05,
+      "loss": 0.2662,
+      "step": 811
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1534916181430674,
+      "learning_rate": 1.9993316588166307e-05,
+      "loss": 0.2768,
+      "step": 812
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.127511520512681,
+      "learning_rate": 1.999325624842232e-05,
+      "loss": 0.2873,
+      "step": 813
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0337648173317513,
+      "learning_rate": 1.9993195637610695e-05,
+      "loss": 0.2319,
+      "step": 814
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1132394984821739,
+      "learning_rate": 1.9993134755733075e-05,
+      "loss": 0.2471,
+      "step": 815
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1475184237917153,
+      "learning_rate": 1.9993073602791108e-05,
+      "loss": 0.2765,
+      "step": 816
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.333803344892296,
+      "learning_rate": 1.999301217878646e-05,
+      "loss": 0.2513,
+      "step": 817
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0499954721840008,
+      "learning_rate": 1.9992950483720787e-05,
+      "loss": 0.2605,
+      "step": 818
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3012802506530357,
+      "learning_rate": 1.999288851759577e-05,
+      "loss": 0.2676,
+      "step": 819
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4136693582266662,
+      "learning_rate": 1.9992826280413087e-05,
+      "loss": 0.2397,
+      "step": 820
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3517183530845074,
+      "learning_rate": 1.9992763772174427e-05,
+      "loss": 0.2772,
+      "step": 821
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9780065045060403,
+      "learning_rate": 1.9992700992881486e-05,
+      "loss": 0.2579,
+      "step": 822
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.107034582008244,
+      "learning_rate": 1.9992637942535963e-05,
+      "loss": 0.3015,
+      "step": 823
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.208106797361871,
+      "learning_rate": 1.9992574621139575e-05,
+      "loss": 0.2684,
+      "step": 824
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0879601895663045,
+      "learning_rate": 1.9992511028694036e-05,
+      "loss": 0.288,
+      "step": 825
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.116435309323238,
+      "learning_rate": 1.999244716520107e-05,
+      "loss": 0.2718,
+      "step": 826
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3059275763713378,
+      "learning_rate": 1.9992383030662412e-05,
+      "loss": 0.307,
+      "step": 827
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6808967976744835,
+      "learning_rate": 1.9992318625079796e-05,
+      "loss": 0.2571,
+      "step": 828
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2453853874792142,
+      "learning_rate": 1.9992253948454975e-05,
+      "loss": 0.265,
+      "step": 829
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2037147547935059,
+      "learning_rate": 1.99921890007897e-05,
+      "loss": 0.2994,
+      "step": 830
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.420063191800624,
+      "learning_rate": 1.9992123782085738e-05,
+      "loss": 0.2597,
+      "step": 831
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2273296556423356,
+      "learning_rate": 1.9992058292344853e-05,
+      "loss": 0.2513,
+      "step": 832
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1062136182815394,
+      "learning_rate": 1.9991992531568817e-05,
+      "loss": 0.2663,
+      "step": 833
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.08435612624478,
+      "learning_rate": 1.9991926499759426e-05,
+      "loss": 0.2818,
+      "step": 834
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.159958172965142,
+      "learning_rate": 1.999186019691846e-05,
+      "loss": 0.2732,
+      "step": 835
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0939438567851123,
+      "learning_rate": 1.9991793623047724e-05,
+      "loss": 0.2321,
+      "step": 836
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8205766434289725,
+      "learning_rate": 1.999172677814902e-05,
+      "loss": 0.2246,
+      "step": 837
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1672477240865653,
+      "learning_rate": 1.9991659662224166e-05,
+      "loss": 0.2701,
+      "step": 838
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.129230834652707,
+      "learning_rate": 1.9991592275274976e-05,
+      "loss": 0.2614,
+      "step": 839
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7214179515732553,
+      "learning_rate": 1.9991524617303282e-05,
+      "loss": 0.2762,
+      "step": 840
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2950757995564135,
+      "learning_rate": 1.999145668831092e-05,
+      "loss": 0.2626,
+      "step": 841
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5470842349719043,
+      "learning_rate": 1.999138848829973e-05,
+      "loss": 0.2667,
+      "step": 842
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5064490380389803,
+      "learning_rate": 1.9991320017271562e-05,
+      "loss": 0.2616,
+      "step": 843
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1662089984074246,
+      "learning_rate": 1.9991251275228274e-05,
+      "loss": 0.2637,
+      "step": 844
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2511580698565579,
+      "learning_rate": 1.9991182262171734e-05,
+      "loss": 0.2811,
+      "step": 845
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1962141316679449,
+      "learning_rate": 1.9991112978103807e-05,
+      "loss": 0.2535,
+      "step": 846
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2435377890556183,
+      "learning_rate": 1.9991043423026377e-05,
+      "loss": 0.2707,
+      "step": 847
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.125868593032254,
+      "learning_rate": 1.999097359694133e-05,
+      "loss": 0.2737,
+      "step": 848
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.171550604371576,
+      "learning_rate": 1.999090349985056e-05,
+      "loss": 0.2519,
+      "step": 849
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3676752292310235,
+      "learning_rate": 1.999083313175597e-05,
+      "loss": 0.2485,
+      "step": 850
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2978859424907145,
+      "learning_rate": 1.9990762492659466e-05,
+      "loss": 0.2642,
+      "step": 851
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1216709783383152,
+      "learning_rate": 1.9990691582562963e-05,
+      "loss": 0.2571,
+      "step": 852
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4066273485739214,
+      "learning_rate": 1.9990620401468392e-05,
+      "loss": 0.2656,
+      "step": 853
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0970791764561405,
+      "learning_rate": 1.9990548949377674e-05,
+      "loss": 0.2225,
+      "step": 854
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1687899119933483,
+      "learning_rate": 1.999047722629275e-05,
+      "loss": 0.2735,
+      "step": 855
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5732209820343088,
+      "learning_rate": 1.999040523221557e-05,
+      "loss": 0.2543,
+      "step": 856
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.725733160854452,
+      "learning_rate": 1.9990332967148082e-05,
+      "loss": 0.2354,
+      "step": 857
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.28006853250872,
+      "learning_rate": 1.999026043109225e-05,
+      "loss": 0.2678,
+      "step": 858
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.172545431709117,
+      "learning_rate": 1.9990187624050038e-05,
+      "loss": 0.2394,
+      "step": 859
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2752673779385229,
+      "learning_rate": 1.9990114546023423e-05,
+      "loss": 0.2663,
+      "step": 860
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0882818240971561,
+      "learning_rate": 1.999004119701439e-05,
+      "loss": 0.2499,
+      "step": 861
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9920608283590259,
+      "learning_rate": 1.9989967577024922e-05,
+      "loss": 0.2615,
+      "step": 862
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0741793303028535,
+      "learning_rate": 1.9989893686057016e-05,
+      "loss": 0.2781,
+      "step": 863
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1532640931043636,
+      "learning_rate": 1.9989819524112683e-05,
+      "loss": 0.2701,
+      "step": 864
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0592006955548727,
+      "learning_rate": 1.998974509119393e-05,
+      "loss": 0.2384,
+      "step": 865
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0794821739805955,
+      "learning_rate": 1.9989670387302783e-05,
+      "loss": 0.2575,
+      "step": 866
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9108946007980193,
+      "learning_rate": 1.9989595412441252e-05,
+      "loss": 0.2738,
+      "step": 867
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2301240916180438,
+      "learning_rate": 1.9989520166611388e-05,
+      "loss": 0.285,
+      "step": 868
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1232165575207511,
+      "learning_rate": 1.9989444649815226e-05,
+      "loss": 0.2408,
+      "step": 869
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3153763910607554,
+      "learning_rate": 1.9989368862054814e-05,
+      "loss": 0.2648,
+      "step": 870
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.067960049988388,
+      "learning_rate": 1.9989292803332203e-05,
+      "loss": 0.2372,
+      "step": 871
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1159094593203644,
+      "learning_rate": 1.9989216473649466e-05,
+      "loss": 0.2354,
+      "step": 872
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0141369128982671,
+      "learning_rate": 1.998913987300866e-05,
+      "loss": 0.2533,
+      "step": 873
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0006771174963607,
+      "learning_rate": 1.998906300141188e-05,
+      "loss": 0.2808,
+      "step": 874
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.116211648499011,
+      "learning_rate": 1.9988985858861193e-05,
+      "loss": 0.2637,
+      "step": 875
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2578617963336698,
+      "learning_rate": 1.9988908445358705e-05,
+      "loss": 0.2743,
+      "step": 876
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5131417020303695,
+      "learning_rate": 1.998883076090651e-05,
+      "loss": 0.2753,
+      "step": 877
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2599095826842825,
+      "learning_rate": 1.9988752805506723e-05,
+      "loss": 0.2678,
+      "step": 878
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9360444508399847,
+      "learning_rate": 1.9988674579161444e-05,
+      "loss": 0.2284,
+      "step": 879
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4269838584838916,
+      "learning_rate": 1.9988596081872805e-05,
+      "loss": 0.2552,
+      "step": 880
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2384542201113082,
+      "learning_rate": 1.9988517313642934e-05,
+      "loss": 0.2617,
+      "step": 881
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4157479977396126,
+      "learning_rate": 1.9988438274473966e-05,
+      "loss": 0.2937,
+      "step": 882
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.481445241615998,
+      "learning_rate": 1.9988358964368046e-05,
+      "loss": 0.2472,
+      "step": 883
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4095283492098256,
+      "learning_rate": 1.9988279383327324e-05,
+      "loss": 0.279,
+      "step": 884
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1291656446775273,
+      "learning_rate": 1.9988199531353963e-05,
+      "loss": 0.2332,
+      "step": 885
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.556219150900177,
+      "learning_rate": 1.998811940845012e-05,
+      "loss": 0.2166,
+      "step": 886
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1722666149602972,
+      "learning_rate": 1.998803901461798e-05,
+      "loss": 0.2168,
+      "step": 887
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.362960636009129,
+      "learning_rate": 1.998795834985971e-05,
+      "loss": 0.2438,
+      "step": 888
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2382505267852677,
+      "learning_rate": 1.998787741417751e-05,
+      "loss": 0.2446,
+      "step": 889
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2519924974220127,
+      "learning_rate": 1.9987796207573573e-05,
+      "loss": 0.2247,
+      "step": 890
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.138520059051484,
+      "learning_rate": 1.9987714730050098e-05,
+      "loss": 0.2394,
+      "step": 891
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.184081525001222,
+      "learning_rate": 1.9987632981609297e-05,
+      "loss": 0.2342,
+      "step": 892
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4228525464925417,
+      "learning_rate": 1.9987550962253387e-05,
+      "loss": 0.2505,
+      "step": 893
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1732583727474242,
+      "learning_rate": 1.998746867198459e-05,
+      "loss": 0.2408,
+      "step": 894
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2929452831438806,
+      "learning_rate": 1.9987386110805146e-05,
+      "loss": 0.2636,
+      "step": 895
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1957505225992544,
+      "learning_rate": 1.9987303278717288e-05,
+      "loss": 0.2597,
+      "step": 896
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3406390837672997,
+      "learning_rate": 1.9987220175723265e-05,
+      "loss": 0.2643,
+      "step": 897
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1369635625540935,
+      "learning_rate": 1.998713680182533e-05,
+      "loss": 0.2523,
+      "step": 898
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5937134878395935,
+      "learning_rate": 1.9987053157025748e-05,
+      "loss": 0.2677,
+      "step": 899
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.305439396439621,
+      "learning_rate": 1.998696924132678e-05,
+      "loss": 0.2856,
+      "step": 900
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3427530047357976,
+      "learning_rate": 1.9986885054730708e-05,
+      "loss": 0.2232,
+      "step": 901
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3349578827424304,
+      "learning_rate": 1.9986800597239817e-05,
+      "loss": 0.2687,
+      "step": 902
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2393032091995557,
+      "learning_rate": 1.9986715868856396e-05,
+      "loss": 0.2342,
+      "step": 903
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.349656127097357,
+      "learning_rate": 1.998663086958274e-05,
+      "loss": 0.2683,
+      "step": 904
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4183409271315817,
+      "learning_rate": 1.998654559942116e-05,
+      "loss": 0.2843,
+      "step": 905
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8087602257457172,
+      "learning_rate": 1.998646005837397e-05,
+      "loss": 0.2616,
+      "step": 906
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5093979621304114,
+      "learning_rate": 1.998637424644348e-05,
+      "loss": 0.2367,
+      "step": 907
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2972565928670108,
+      "learning_rate": 1.998628816363203e-05,
+      "loss": 0.2765,
+      "step": 908
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1342135617558307,
+      "learning_rate": 1.9986201809941945e-05,
+      "loss": 0.2605,
+      "step": 909
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.622288471832347,
+      "learning_rate": 1.9986115185375576e-05,
+      "loss": 0.2756,
+      "step": 910
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.712971571944573,
+      "learning_rate": 1.9986028289935263e-05,
+      "loss": 0.3367,
+      "step": 911
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.831004903324416,
+      "learning_rate": 1.9985941123623374e-05,
+      "loss": 0.2635,
+      "step": 912
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 37.253091376437865,
+      "learning_rate": 1.9985853686442266e-05,
+      "loss": 0.425,
+      "step": 913
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6139419406539022,
+      "learning_rate": 1.9985765978394315e-05,
+      "loss": 0.2935,
+      "step": 914
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6099763905589324,
+      "learning_rate": 1.9985677999481898e-05,
+      "loss": 0.2702,
+      "step": 915
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6413119822026072,
+      "learning_rate": 1.9985589749707395e-05,
+      "loss": 0.2629,
+      "step": 916
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3524571716304779,
+      "learning_rate": 1.9985501229073213e-05,
+      "loss": 0.2703,
+      "step": 917
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0938158950381753,
+      "learning_rate": 1.998541243758174e-05,
+      "loss": 0.2704,
+      "step": 918
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1489715020729094,
+      "learning_rate": 1.9985323375235395e-05,
+      "loss": 0.2543,
+      "step": 919
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1731682133705816,
+      "learning_rate": 1.9985234042036588e-05,
+      "loss": 0.2608,
+      "step": 920
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9738308855528613,
+      "learning_rate": 1.9985144437987743e-05,
+      "loss": 0.2424,
+      "step": 921
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0626982152800977,
+      "learning_rate": 1.9985054563091295e-05,
+      "loss": 0.2758,
+      "step": 922
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.426479209067473,
+      "learning_rate": 1.9984964417349675e-05,
+      "loss": 0.2733,
+      "step": 923
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1284519447109926,
+      "learning_rate": 1.998487400076533e-05,
+      "loss": 0.2457,
+      "step": 924
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0912469894763157,
+      "learning_rate": 1.9984783313340715e-05,
+      "loss": 0.2408,
+      "step": 925
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1812135886717172,
+      "learning_rate": 1.998469235507829e-05,
+      "loss": 0.2503,
+      "step": 926
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0073430784559474,
+      "learning_rate": 1.998460112598052e-05,
+      "loss": 0.244,
+      "step": 927
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3404204748861077,
+      "learning_rate": 1.998450962604988e-05,
+      "loss": 0.2885,
+      "step": 928
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7811568491496874,
+      "learning_rate": 1.9984417855288853e-05,
+      "loss": 0.2516,
+      "step": 929
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5443231025515465,
+      "learning_rate": 1.998432581369993e-05,
+      "loss": 0.2572,
+      "step": 930
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4293337544325062,
+      "learning_rate": 1.9984233501285602e-05,
+      "loss": 0.2992,
+      "step": 931
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.284982874168832,
+      "learning_rate": 1.9984140918048376e-05,
+      "loss": 0.2399,
+      "step": 932
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3961467293148901,
+      "learning_rate": 1.9984048063990766e-05,
+      "loss": 0.2564,
+      "step": 933
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1794505830153215,
+      "learning_rate": 1.9983954939115286e-05,
+      "loss": 0.2417,
+      "step": 934
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2760274907553164,
+      "learning_rate": 1.9983861543424467e-05,
+      "loss": 0.2662,
+      "step": 935
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1984319455046042,
+      "learning_rate": 1.998376787692084e-05,
+      "loss": 0.2402,
+      "step": 936
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5134177656969858,
+      "learning_rate": 1.9983673939606946e-05,
+      "loss": 0.2641,
+      "step": 937
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.08231174372219,
+      "learning_rate": 1.9983579731485326e-05,
+      "loss": 0.2398,
+      "step": 938
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3307559472577821,
+      "learning_rate": 1.998348525255855e-05,
+      "loss": 0.2773,
+      "step": 939
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.256364695732902,
+      "learning_rate": 1.9983390502829168e-05,
+      "loss": 0.2815,
+      "step": 940
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2766424443497033,
+      "learning_rate": 1.9983295482299752e-05,
+      "loss": 0.2555,
+      "step": 941
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3076871853427046,
+      "learning_rate": 1.998320019097289e-05,
+      "loss": 0.234,
+      "step": 942
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7653987555897812,
+      "learning_rate": 1.9983104628851154e-05,
+      "loss": 0.2476,
+      "step": 943
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.147047712588286,
+      "learning_rate": 1.9983008795937142e-05,
+      "loss": 0.291,
+      "step": 944
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9005551250112118,
+      "learning_rate": 1.9982912692233455e-05,
+      "loss": 0.2655,
+      "step": 945
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2830320142702902,
+      "learning_rate": 1.9982816317742694e-05,
+      "loss": 0.2607,
+      "step": 946
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6188483702984462,
+      "learning_rate": 1.9982719672467476e-05,
+      "loss": 0.2535,
+      "step": 947
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.494189922904832,
+      "learning_rate": 1.9982622756410425e-05,
+      "loss": 0.263,
+      "step": 948
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6891101259528638,
+      "learning_rate": 1.998252556957417e-05,
+      "loss": 0.2674,
+      "step": 949
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4203377893373739,
+      "learning_rate": 1.998242811196134e-05,
+      "loss": 0.2332,
+      "step": 950
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5932351561865459,
+      "learning_rate": 1.9982330383574586e-05,
+      "loss": 0.2421,
+      "step": 951
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.03124317920117,
+      "learning_rate": 1.998223238441656e-05,
+      "loss": 0.2498,
+      "step": 952
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8931422127100161,
+      "learning_rate": 1.9982134114489912e-05,
+      "loss": 0.2673,
+      "step": 953
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5731660080894065,
+      "learning_rate": 1.9982035573797315e-05,
+      "loss": 0.2766,
+      "step": 954
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2214546421076942,
+      "learning_rate": 1.9981936762341438e-05,
+      "loss": 0.2317,
+      "step": 955
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3338466940653078,
+      "learning_rate": 1.9981837680124963e-05,
+      "loss": 0.2238,
+      "step": 956
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4219834080515363,
+      "learning_rate": 1.9981738327150575e-05,
+      "loss": 0.2452,
+      "step": 957
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.281365091349734,
+      "learning_rate": 1.9981638703420977e-05,
+      "loss": 0.2621,
+      "step": 958
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2303691416352147,
+      "learning_rate": 1.998153880893886e-05,
+      "loss": 0.2442,
+      "step": 959
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7164286811059375,
+      "learning_rate": 1.998143864370694e-05,
+      "loss": 0.25,
+      "step": 960
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1619798707542857,
+      "learning_rate": 1.998133820772793e-05,
+      "loss": 0.2448,
+      "step": 961
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4654543872298886,
+      "learning_rate": 1.998123750100456e-05,
+      "loss": 0.2592,
+      "step": 962
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0862977637195583,
+      "learning_rate": 1.9981136523539565e-05,
+      "loss": 0.2426,
+      "step": 963
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2390086218034442,
+      "learning_rate": 1.9981035275335672e-05,
+      "loss": 0.2557,
+      "step": 964
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.504056753106432,
+      "learning_rate": 1.9980933756395635e-05,
+      "loss": 0.2697,
+      "step": 965
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.398764234923714,
+      "learning_rate": 1.9980831966722204e-05,
+      "loss": 0.2398,
+      "step": 966
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2339532264497461,
+      "learning_rate": 1.9980729906318145e-05,
+      "loss": 0.264,
+      "step": 967
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4824979512173875,
+      "learning_rate": 1.998062757518622e-05,
+      "loss": 0.2453,
+      "step": 968
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7206314812947987,
+      "learning_rate": 1.998052497332921e-05,
+      "loss": 0.2569,
+      "step": 969
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6681319774857464,
+      "learning_rate": 1.99804221007499e-05,
+      "loss": 0.2418,
+      "step": 970
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2922792928709421,
+      "learning_rate": 1.9980318957451073e-05,
+      "loss": 0.233,
+      "step": 971
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3283707199306902,
+      "learning_rate": 1.9980215543435532e-05,
+      "loss": 0.2715,
+      "step": 972
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7199472670960634,
+      "learning_rate": 1.998011185870608e-05,
+      "loss": 0.2525,
+      "step": 973
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2668493722493168,
+      "learning_rate": 1.998000790326553e-05,
+      "loss": 0.25,
+      "step": 974
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4536780933019773,
+      "learning_rate": 1.9979903677116705e-05,
+      "loss": 0.2754,
+      "step": 975
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.395271992192874,
+      "learning_rate": 1.9979799180262423e-05,
+      "loss": 0.2559,
+      "step": 976
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.691859147708012,
+      "learning_rate": 1.997969441270553e-05,
+      "loss": 0.2739,
+      "step": 977
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3249583444644388,
+      "learning_rate": 1.997958937444886e-05,
+      "loss": 0.2406,
+      "step": 978
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1293700203760684,
+      "learning_rate": 1.9979484065495264e-05,
+      "loss": 0.2522,
+      "step": 979
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6352020696036085,
+      "learning_rate": 1.99793784858476e-05,
+      "loss": 0.2365,
+      "step": 980
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4766571166239415,
+      "learning_rate": 1.997927263550873e-05,
+      "loss": 0.2377,
+      "step": 981
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5825322786834481,
+      "learning_rate": 1.997916651448153e-05,
+      "loss": 0.283,
+      "step": 982
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5960016847073644,
+      "learning_rate": 1.997906012276887e-05,
+      "loss": 0.2336,
+      "step": 983
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7464179676633318,
+      "learning_rate": 1.9978953460373643e-05,
+      "loss": 0.2491,
+      "step": 984
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.332022414231344,
+      "learning_rate": 1.997884652729874e-05,
+      "loss": 0.2708,
+      "step": 985
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4284193259813256,
+      "learning_rate": 1.997873932354706e-05,
+      "loss": 0.2558,
+      "step": 986
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2548510619479858,
+      "learning_rate": 1.9978631849121514e-05,
+      "loss": 0.23,
+      "step": 987
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5253198374407417,
+      "learning_rate": 1.997852410402501e-05,
+      "loss": 0.2598,
+      "step": 988
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6856588529543604,
+      "learning_rate": 1.9978416088260483e-05,
+      "loss": 0.2466,
+      "step": 989
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.204933102528033,
+      "learning_rate": 1.9978307801830855e-05,
+      "loss": 0.2608,
+      "step": 990
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.403160012398721,
+      "learning_rate": 1.997819924473906e-05,
+      "loss": 0.2457,
+      "step": 991
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3417670872825422,
+      "learning_rate": 1.997809041698805e-05,
+      "loss": 0.2481,
+      "step": 992
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9679789941414283,
+      "learning_rate": 1.9977981318580773e-05,
+      "loss": 0.2481,
+      "step": 993
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.802566030293649,
+      "learning_rate": 1.9977871949520188e-05,
+      "loss": 0.2454,
+      "step": 994
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1290121014796652,
+      "learning_rate": 1.9977762309809266e-05,
+      "loss": 0.2532,
+      "step": 995
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4914383201086971,
+      "learning_rate": 1.9977652399450976e-05,
+      "loss": 0.273,
+      "step": 996
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4738780508956397,
+      "learning_rate": 1.99775422184483e-05,
+      "loss": 0.2562,
+      "step": 997
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.431594373320625,
+      "learning_rate": 1.997743176680423e-05,
+      "loss": 0.2681,
+      "step": 998
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2916825659721045,
+      "learning_rate": 1.997732104452176e-05,
+      "loss": 0.2434,
+      "step": 999
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.887959469677641,
+      "learning_rate": 1.997721005160389e-05,
+      "loss": 0.2451,
+      "step": 1000
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4206349978995914,
+      "learning_rate": 1.9977098788053637e-05,
+      "loss": 0.262,
+      "step": 1001
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.146152194745477,
+      "learning_rate": 1.9976987253874016e-05,
+      "loss": 0.2638,
+      "step": 1002
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.318586277766884,
+      "learning_rate": 1.997687544906805e-05,
+      "loss": 0.2635,
+      "step": 1003
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3449964495777744,
+      "learning_rate": 1.9976763373638773e-05,
+      "loss": 0.2805,
+      "step": 1004
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3295176549041057,
+      "learning_rate": 1.997665102758923e-05,
+      "loss": 0.2925,
+      "step": 1005
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.356327669023187,
+      "learning_rate": 1.997653841092246e-05,
+      "loss": 0.2487,
+      "step": 1006
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4765934626224548,
+      "learning_rate": 1.9976425523641527e-05,
+      "loss": 0.2929,
+      "step": 1007
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.497001144648849,
+      "learning_rate": 1.9976312365749484e-05,
+      "loss": 0.2477,
+      "step": 1008
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5166775118877334,
+      "learning_rate": 1.9976198937249408e-05,
+      "loss": 0.2361,
+      "step": 1009
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4083767570633934,
+      "learning_rate": 1.997608523814437e-05,
+      "loss": 0.254,
+      "step": 1010
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.540794321329871,
+      "learning_rate": 1.9975971268437457e-05,
+      "loss": 0.2738,
+      "step": 1011
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.668700196997005,
+      "learning_rate": 1.997585702813176e-05,
+      "loss": 0.2793,
+      "step": 1012
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5112483099308371,
+      "learning_rate": 1.9975742517230377e-05,
+      "loss": 0.2443,
+      "step": 1013
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5205434194057652,
+      "learning_rate": 1.9975627735736416e-05,
+      "loss": 0.2532,
+      "step": 1014
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.40062000800819,
+      "learning_rate": 1.9975512683652985e-05,
+      "loss": 0.2509,
+      "step": 1015
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4963993851155675,
+      "learning_rate": 1.9975397360983216e-05,
+      "loss": 0.24,
+      "step": 1016
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1069124215856547,
+      "learning_rate": 1.9975281767730226e-05,
+      "loss": 0.2256,
+      "step": 1017
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4490239252320818,
+      "learning_rate": 1.9975165903897155e-05,
+      "loss": 0.2318,
+      "step": 1018
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.594997764001681,
+      "learning_rate": 1.9975049769487147e-05,
+      "loss": 0.2416,
+      "step": 1019
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7645257393696694,
+      "learning_rate": 1.9974933364503347e-05,
+      "loss": 0.2589,
+      "step": 1020
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.288053085705425,
+      "learning_rate": 1.9974816688948923e-05,
+      "loss": 0.2536,
+      "step": 1021
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2860181373293385,
+      "learning_rate": 1.9974699742827028e-05,
+      "loss": 0.2414,
+      "step": 1022
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4077289076246466,
+      "learning_rate": 1.997458252614084e-05,
+      "loss": 0.2416,
+      "step": 1023
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.378778097881167,
+      "learning_rate": 1.9974465038893535e-05,
+      "loss": 0.2535,
+      "step": 1024
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3503965136445129,
+      "learning_rate": 1.9974347281088305e-05,
+      "loss": 0.2676,
+      "step": 1025
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.236924180487533,
+      "learning_rate": 1.9974229252728345e-05,
+      "loss": 0.2545,
+      "step": 1026
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0346377489547494,
+      "learning_rate": 1.9974110953816846e-05,
+      "loss": 0.239,
+      "step": 1027
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3754934009638555,
+      "learning_rate": 1.997399238435703e-05,
+      "loss": 0.288,
+      "step": 1028
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2183415405104483,
+      "learning_rate": 1.99738735443521e-05,
+      "loss": 0.2624,
+      "step": 1029
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9613397656490309,
+      "learning_rate": 1.9973754433805294e-05,
+      "loss": 0.2424,
+      "step": 1030
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6289514132511271,
+      "learning_rate": 1.9973635052719836e-05,
+      "loss": 0.2523,
+      "step": 1031
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0487668403001444,
+      "learning_rate": 1.997351540109896e-05,
+      "loss": 0.2854,
+      "step": 1032
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3730747343031353,
+      "learning_rate": 1.9973395478945917e-05,
+      "loss": 0.251,
+      "step": 1033
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.247569885594585,
+      "learning_rate": 1.9973275286263955e-05,
+      "loss": 0.2465,
+      "step": 1034
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.200220973084389,
+      "learning_rate": 1.9973154823056343e-05,
+      "loss": 0.2733,
+      "step": 1035
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1170295827926,
+      "learning_rate": 1.997303408932634e-05,
+      "loss": 0.2553,
+      "step": 1036
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.234375683527027,
+      "learning_rate": 1.9972913085077225e-05,
+      "loss": 0.2682,
+      "step": 1037
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2679487419761017,
+      "learning_rate": 1.997279181031228e-05,
+      "loss": 0.272,
+      "step": 1038
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6447639613922878,
+      "learning_rate": 1.997267026503479e-05,
+      "loss": 0.3011,
+      "step": 1039
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3634475769344572,
+      "learning_rate": 1.997254844924806e-05,
+      "loss": 0.2631,
+      "step": 1040
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2718450470082385,
+      "learning_rate": 1.997242636295539e-05,
+      "loss": 0.2412,
+      "step": 1041
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0108447929482707,
+      "learning_rate": 1.997230400616009e-05,
+      "loss": 0.237,
+      "step": 1042
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1645213079118888,
+      "learning_rate": 1.997218137886548e-05,
+      "loss": 0.2645,
+      "step": 1043
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0614791909352574,
+      "learning_rate": 1.997205848107489e-05,
+      "loss": 0.2693,
+      "step": 1044
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4634827744750394,
+      "learning_rate": 1.9971935312791646e-05,
+      "loss": 0.2357,
+      "step": 1045
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9872275898199967,
+      "learning_rate": 1.9971811874019096e-05,
+      "loss": 0.2554,
+      "step": 1046
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2016246852199257,
+      "learning_rate": 1.9971688164760588e-05,
+      "loss": 0.2769,
+      "step": 1047
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9982730088115166,
+      "learning_rate": 1.997156418501947e-05,
+      "loss": 0.2465,
+      "step": 1048
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6992277962147386,
+      "learning_rate": 1.9971439934799113e-05,
+      "loss": 0.2723,
+      "step": 1049
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0004795906439163,
+      "learning_rate": 1.9971315414102886e-05,
+      "loss": 0.2699,
+      "step": 1050
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2129500121770678,
+      "learning_rate": 1.9971190622934164e-05,
+      "loss": 0.2474,
+      "step": 1051
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.131678439889529,
+      "learning_rate": 1.9971065561296334e-05,
+      "loss": 0.2638,
+      "step": 1052
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1363756749163911,
+      "learning_rate": 1.9970940229192785e-05,
+      "loss": 0.2406,
+      "step": 1053
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2498200840780742,
+      "learning_rate": 1.997081462662692e-05,
+      "loss": 0.2639,
+      "step": 1054
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1762334890050876,
+      "learning_rate": 1.997068875360215e-05,
+      "loss": 0.2543,
+      "step": 1055
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0024363148955997,
+      "learning_rate": 1.9970562610121878e-05,
+      "loss": 0.2289,
+      "step": 1056
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1811960528971635,
+      "learning_rate": 1.9970436196189534e-05,
+      "loss": 0.2478,
+      "step": 1057
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1637800873859996,
+      "learning_rate": 1.9970309511808544e-05,
+      "loss": 0.2465,
+      "step": 1058
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.979550169262176,
+      "learning_rate": 1.997018255698235e-05,
+      "loss": 0.2214,
+      "step": 1059
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.16589763971035,
+      "learning_rate": 1.9970055331714383e-05,
+      "loss": 0.2711,
+      "step": 1060
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1246384724575211,
+      "learning_rate": 1.9969927836008106e-05,
+      "loss": 0.2775,
+      "step": 1061
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7036162064431464,
+      "learning_rate": 1.9969800069866977e-05,
+      "loss": 0.2495,
+      "step": 1062
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1887124846196262,
+      "learning_rate": 1.9969672033294456e-05,
+      "loss": 0.2403,
+      "step": 1063
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1111538706770012,
+      "learning_rate": 1.9969543726294015e-05,
+      "loss": 0.2787,
+      "step": 1064
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.344197692512503,
+      "learning_rate": 1.996941514886914e-05,
+      "loss": 0.2406,
+      "step": 1065
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9370858572857556,
+      "learning_rate": 1.9969286301023313e-05,
+      "loss": 0.2453,
+      "step": 1066
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.317987916437125,
+      "learning_rate": 1.9969157182760038e-05,
+      "loss": 0.2469,
+      "step": 1067
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4308374640331052,
+      "learning_rate": 1.9969027794082805e-05,
+      "loss": 0.2618,
+      "step": 1068
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0128360475469447,
+      "learning_rate": 1.9968898134995133e-05,
+      "loss": 0.2589,
+      "step": 1069
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2206987095575548,
+      "learning_rate": 1.9968768205500537e-05,
+      "loss": 0.2539,
+      "step": 1070
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2616967144566469,
+      "learning_rate": 1.996863800560254e-05,
+      "loss": 0.2561,
+      "step": 1071
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2420509612739192,
+      "learning_rate": 1.9968507535304673e-05,
+      "loss": 0.2488,
+      "step": 1072
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6531805252472223,
+      "learning_rate": 1.9968376794610476e-05,
+      "loss": 0.2473,
+      "step": 1073
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.512339883342327,
+      "learning_rate": 1.9968245783523494e-05,
+      "loss": 0.2761,
+      "step": 1074
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1497848143449239,
+      "learning_rate": 1.9968114502047285e-05,
+      "loss": 0.2543,
+      "step": 1075
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3298769541275925,
+      "learning_rate": 1.9967982950185406e-05,
+      "loss": 0.2466,
+      "step": 1076
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1633256939356942,
+      "learning_rate": 1.9967851127941428e-05,
+      "loss": 0.2452,
+      "step": 1077
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4220440751364178,
+      "learning_rate": 1.9967719035318923e-05,
+      "loss": 0.2764,
+      "step": 1078
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1132936198571883,
+      "learning_rate": 1.996758667232148e-05,
+      "loss": 0.26,
+      "step": 1079
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6312747105898855,
+      "learning_rate": 1.996745403895268e-05,
+      "loss": 0.2669,
+      "step": 1080
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6402142909338209,
+      "learning_rate": 1.996732113521613e-05,
+      "loss": 0.2457,
+      "step": 1081
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0667893373200155,
+      "learning_rate": 1.996718796111543e-05,
+      "loss": 0.2608,
+      "step": 1082
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.852776127345349,
+      "learning_rate": 1.9967054516654192e-05,
+      "loss": 0.2296,
+      "step": 1083
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.409690469886677,
+      "learning_rate": 1.996692080183604e-05,
+      "loss": 0.2808,
+      "step": 1084
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3043579795586937,
+      "learning_rate": 1.9966786816664595e-05,
+      "loss": 0.2389,
+      "step": 1085
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5235254463247614,
+      "learning_rate": 1.9966652561143497e-05,
+      "loss": 0.2704,
+      "step": 1086
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1451019290362825,
+      "learning_rate": 1.9966518035276386e-05,
+      "loss": 0.2436,
+      "step": 1087
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1315370371336715,
+      "learning_rate": 1.996638323906691e-05,
+      "loss": 0.2551,
+      "step": 1088
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2459189550689043,
+      "learning_rate": 1.9966248172518724e-05,
+      "loss": 0.2388,
+      "step": 1089
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0925475069915203,
+      "learning_rate": 1.9966112835635493e-05,
+      "loss": 0.2755,
+      "step": 1090
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5310245971490808,
+      "learning_rate": 1.996597722842089e-05,
+      "loss": 0.2449,
+      "step": 1091
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1293064532018613,
+      "learning_rate": 1.9965841350878594e-05,
+      "loss": 0.2457,
+      "step": 1092
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.270511290540653,
+      "learning_rate": 1.9965705203012288e-05,
+      "loss": 0.2584,
+      "step": 1093
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3994165582848725,
+      "learning_rate": 1.9965568784825665e-05,
+      "loss": 0.2679,
+      "step": 1094
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2426772281808531,
+      "learning_rate": 1.9965432096322423e-05,
+      "loss": 0.2421,
+      "step": 1095
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3840821487760482,
+      "learning_rate": 1.9965295137506275e-05,
+      "loss": 0.2817,
+      "step": 1096
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5904477416527278,
+      "learning_rate": 1.9965157908380934e-05,
+      "loss": 0.2625,
+      "step": 1097
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0619560169492481,
+      "learning_rate": 1.996502040895012e-05,
+      "loss": 0.2489,
+      "step": 1098
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1477414335171585,
+      "learning_rate": 1.9964882639217564e-05,
+      "loss": 0.2561,
+      "step": 1099
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2290065958182672,
+      "learning_rate": 1.9964744599187006e-05,
+      "loss": 0.2423,
+      "step": 1100
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0360763227358927,
+      "learning_rate": 1.9964606288862187e-05,
+      "loss": 0.2513,
+      "step": 1101
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0421375048836479,
+      "learning_rate": 1.9964467708246858e-05,
+      "loss": 0.2321,
+      "step": 1102
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1176614561117166,
+      "learning_rate": 1.9964328857344782e-05,
+      "loss": 0.2346,
+      "step": 1103
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1864842864031386,
+      "learning_rate": 1.9964189736159724e-05,
+      "loss": 0.2379,
+      "step": 1104
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9928267143128622,
+      "learning_rate": 1.9964050344695454e-05,
+      "loss": 0.2471,
+      "step": 1105
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2547197785017616,
+      "learning_rate": 1.9963910682955755e-05,
+      "loss": 0.2796,
+      "step": 1106
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.108318882422375,
+      "learning_rate": 1.9963770750944416e-05,
+      "loss": 0.2731,
+      "step": 1107
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3881418626610218,
+      "learning_rate": 1.9963630548665234e-05,
+      "loss": 0.2565,
+      "step": 1108
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0514992110398136,
+      "learning_rate": 1.9963490076122013e-05,
+      "loss": 0.2517,
+      "step": 1109
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0593218515735172,
+      "learning_rate": 1.9963349333318557e-05,
+      "loss": 0.2534,
+      "step": 1110
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.118821356451065,
+      "learning_rate": 1.996320832025869e-05,
+      "loss": 0.2423,
+      "step": 1111
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.303669999694777,
+      "learning_rate": 1.9963067036946234e-05,
+      "loss": 0.2487,
+      "step": 1112
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9298685030587246,
+      "learning_rate": 1.996292548338502e-05,
+      "loss": 0.255,
+      "step": 1113
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1029318356890085,
+      "learning_rate": 1.9962783659578893e-05,
+      "loss": 0.2728,
+      "step": 1114
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0197604969095162,
+      "learning_rate": 1.9962641565531694e-05,
+      "loss": 0.2539,
+      "step": 1115
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4415415643758744,
+      "learning_rate": 1.9962499201247278e-05,
+      "loss": 0.2628,
+      "step": 1116
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.083338942951384,
+      "learning_rate": 1.996235656672951e-05,
+      "loss": 0.2233,
+      "step": 1117
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1160686080415427,
+      "learning_rate": 1.9962213661982258e-05,
+      "loss": 0.2383,
+      "step": 1118
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3011374875117159,
+      "learning_rate": 1.99620704870094e-05,
+      "loss": 0.2546,
+      "step": 1119
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2663679967062615,
+      "learning_rate": 1.9961927041814818e-05,
+      "loss": 0.2764,
+      "step": 1120
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.08364336655249,
+      "learning_rate": 1.99617833264024e-05,
+      "loss": 0.2728,
+      "step": 1121
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1913191519966877,
+      "learning_rate": 1.9961639340776044e-05,
+      "loss": 0.2662,
+      "step": 1122
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.337006850164463,
+      "learning_rate": 1.9961495084939663e-05,
+      "loss": 0.2617,
+      "step": 1123
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9984066446142431,
+      "learning_rate": 1.9961350558897165e-05,
+      "loss": 0.2304,
+      "step": 1124
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1472444721091424,
+      "learning_rate": 1.996120576265247e-05,
+      "loss": 0.2582,
+      "step": 1125
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0510611344781502,
+      "learning_rate": 1.99610606962095e-05,
+      "loss": 0.2411,
+      "step": 1126
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1413980782504782,
+      "learning_rate": 1.99609153595722e-05,
+      "loss": 0.2501,
+      "step": 1127
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9585331468733518,
+      "learning_rate": 1.9960769752744508e-05,
+      "loss": 0.2266,
+      "step": 1128
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0384063965422534,
+      "learning_rate": 1.9960623875730376e-05,
+      "loss": 0.2433,
+      "step": 1129
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9887923427738782,
+      "learning_rate": 1.9960477728533756e-05,
+      "loss": 0.242,
+      "step": 1130
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2281455484801602,
+      "learning_rate": 1.9960331311158618e-05,
+      "loss": 0.2478,
+      "step": 1131
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1420820808451997,
+      "learning_rate": 1.9960184623608927e-05,
+      "loss": 0.2313,
+      "step": 1132
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1291272243771713,
+      "learning_rate": 1.996003766588867e-05,
+      "loss": 0.2435,
+      "step": 1133
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.291842843117987,
+      "learning_rate": 1.9959890438001826e-05,
+      "loss": 0.2505,
+      "step": 1134
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0968369271525031,
+      "learning_rate": 1.9959742939952393e-05,
+      "loss": 0.2585,
+      "step": 1135
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1421824643955416,
+      "learning_rate": 1.9959595171744367e-05,
+      "loss": 0.2725,
+      "step": 1136
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9766039383545265,
+      "learning_rate": 1.9959447133381762e-05,
+      "loss": 0.2533,
+      "step": 1137
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9345271213518466,
+      "learning_rate": 1.995929882486859e-05,
+      "loss": 0.241,
+      "step": 1138
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.187322374200805,
+      "learning_rate": 1.9959150246208876e-05,
+      "loss": 0.2347,
+      "step": 1139
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0532689372785176,
+      "learning_rate": 1.995900139740665e-05,
+      "loss": 0.2603,
+      "step": 1140
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5080362513166135,
+      "learning_rate": 1.9958852278465946e-05,
+      "loss": 0.2028,
+      "step": 1141
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1438427953969899,
+      "learning_rate": 1.9958702889390813e-05,
+      "loss": 0.2538,
+      "step": 1142
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3178650171897648,
+      "learning_rate": 1.99585532301853e-05,
+      "loss": 0.2687,
+      "step": 1143
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0837016465004479,
+      "learning_rate": 1.9958403300853472e-05,
+      "loss": 0.2732,
+      "step": 1144
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.151385755054102,
+      "learning_rate": 1.9958253101399388e-05,
+      "loss": 0.2318,
+      "step": 1145
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0980774803415834,
+      "learning_rate": 1.9958102631827127e-05,
+      "loss": 0.2398,
+      "step": 1146
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0518834716128322,
+      "learning_rate": 1.995795189214077e-05,
+      "loss": 0.2567,
+      "step": 1147
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9774327625199531,
+      "learning_rate": 1.9957800882344406e-05,
+      "loss": 0.2912,
+      "step": 1148
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9752705364199556,
+      "learning_rate": 1.9957649602442132e-05,
+      "loss": 0.2637,
+      "step": 1149
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.22461622064381,
+      "learning_rate": 1.9957498052438046e-05,
+      "loss": 0.2682,
+      "step": 1150
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9363047948353536,
+      "learning_rate": 1.9957346232336264e-05,
+      "loss": 0.2608,
+      "step": 1151
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0908039664636027,
+      "learning_rate": 1.9957194142140907e-05,
+      "loss": 0.2642,
+      "step": 1152
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.298250529492443,
+      "learning_rate": 1.9957041781856094e-05,
+      "loss": 0.2428,
+      "step": 1153
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8691045152210589,
+      "learning_rate": 1.995688915148596e-05,
+      "loss": 0.252,
+      "step": 1154
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5017543963557427,
+      "learning_rate": 1.9956736251034643e-05,
+      "loss": 0.2807,
+      "step": 1155
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.425121321973548,
+      "learning_rate": 1.99565830805063e-05,
+      "loss": 0.2524,
+      "step": 1156
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3588406022460622,
+      "learning_rate": 1.995642963990507e-05,
+      "loss": 0.2631,
+      "step": 1157
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.012820438420422,
+      "learning_rate": 1.995627592923513e-05,
+      "loss": 0.2583,
+      "step": 1158
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0423757816916552,
+      "learning_rate": 1.995612194850064e-05,
+      "loss": 0.231,
+      "step": 1159
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1506603528204253,
+      "learning_rate": 1.9955967697705782e-05,
+      "loss": 0.2758,
+      "step": 1160
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.120430771480744,
+      "learning_rate": 1.9955813176854735e-05,
+      "loss": 0.2574,
+      "step": 1161
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0850296741610588,
+      "learning_rate": 1.9955658385951695e-05,
+      "loss": 0.2612,
+      "step": 1162
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1501268535631535,
+      "learning_rate": 1.9955503325000857e-05,
+      "loss": 0.26,
+      "step": 1163
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2212111778915191,
+      "learning_rate": 1.9955347994006432e-05,
+      "loss": 0.2737,
+      "step": 1164
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4028969597594856,
+      "learning_rate": 1.9955192392972628e-05,
+      "loss": 0.2324,
+      "step": 1165
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3102571784603596,
+      "learning_rate": 1.995503652190367e-05,
+      "loss": 0.2674,
+      "step": 1166
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9897987458802096,
+      "learning_rate": 1.9954880380803787e-05,
+      "loss": 0.2423,
+      "step": 1167
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9207734644130189,
+      "learning_rate": 1.9954723969677206e-05,
+      "loss": 0.2306,
+      "step": 1168
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1295488992326463,
+      "learning_rate": 1.9954567288528174e-05,
+      "loss": 0.2418,
+      "step": 1169
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2685761887350202,
+      "learning_rate": 1.9954410337360945e-05,
+      "loss": 0.2357,
+      "step": 1170
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9861608199592001,
+      "learning_rate": 1.9954253116179772e-05,
+      "loss": 0.2426,
+      "step": 1171
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0280451362848153,
+      "learning_rate": 1.9954095624988924e-05,
+      "loss": 0.2339,
+      "step": 1172
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0937539428083092,
+      "learning_rate": 1.9953937863792666e-05,
+      "loss": 0.2407,
+      "step": 1173
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5164023930909465,
+      "learning_rate": 1.9953779832595285e-05,
+      "loss": 0.271,
+      "step": 1174
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0085274941712419,
+      "learning_rate": 1.995362153140106e-05,
+      "loss": 0.2501,
+      "step": 1175
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0757387120719115,
+      "learning_rate": 1.9953462960214293e-05,
+      "loss": 0.2299,
+      "step": 1176
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.562846048479879,
+      "learning_rate": 1.995330411903928e-05,
+      "loss": 0.2913,
+      "step": 1177
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.670780598619956,
+      "learning_rate": 1.995314500788033e-05,
+      "loss": 0.2559,
+      "step": 1178
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0999296700893124,
+      "learning_rate": 1.9952985626741757e-05,
+      "loss": 0.2413,
+      "step": 1179
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2266840303048263,
+      "learning_rate": 1.995282597562789e-05,
+      "loss": 0.2299,
+      "step": 1180
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.12382611130755,
+      "learning_rate": 1.9952666054543053e-05,
+      "loss": 0.2557,
+      "step": 1181
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8852569009332808,
+      "learning_rate": 1.995250586349159e-05,
+      "loss": 0.2543,
+      "step": 1182
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9698648745098383,
+      "learning_rate": 1.9952345402477844e-05,
+      "loss": 0.2397,
+      "step": 1183
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0757537964497887,
+      "learning_rate": 1.9952184671506167e-05,
+      "loss": 0.2867,
+      "step": 1184
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.047570354258211,
+      "learning_rate": 1.9952023670580915e-05,
+      "loss": 0.2373,
+      "step": 1185
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3164713777096313,
+      "learning_rate": 1.9951862399706463e-05,
+      "loss": 0.2745,
+      "step": 1186
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.254144229233726,
+      "learning_rate": 1.995170085888718e-05,
+      "loss": 0.2355,
+      "step": 1187
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0187453637220607,
+      "learning_rate": 1.9951539048127447e-05,
+      "loss": 0.2368,
+      "step": 1188
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.985069297112168,
+      "learning_rate": 1.9951376967431658e-05,
+      "loss": 0.2376,
+      "step": 1189
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1871515603562464,
+      "learning_rate": 1.9951214616804203e-05,
+      "loss": 0.2481,
+      "step": 1190
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1725021607449277,
+      "learning_rate": 1.9951051996249492e-05,
+      "loss": 0.2618,
+      "step": 1191
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0558607145471346,
+      "learning_rate": 1.9950889105771937e-05,
+      "loss": 0.2689,
+      "step": 1192
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8759610787740514,
+      "learning_rate": 1.995072594537595e-05,
+      "loss": 0.258,
+      "step": 1193
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0177805167346698,
+      "learning_rate": 1.9950562515065957e-05,
+      "loss": 0.2662,
+      "step": 1194
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.21473439251119,
+      "learning_rate": 1.9950398814846396e-05,
+      "loss": 0.2414,
+      "step": 1195
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0701077252548996,
+      "learning_rate": 1.9950234844721707e-05,
+      "loss": 0.2294,
+      "step": 1196
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.304305485860522,
+      "learning_rate": 1.9950070604696332e-05,
+      "loss": 0.2575,
+      "step": 1197
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4979470689692984,
+      "learning_rate": 1.994990609477473e-05,
+      "loss": 0.2103,
+      "step": 1198
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.881741820714266,
+      "learning_rate": 1.994974131496137e-05,
+      "loss": 0.2499,
+      "step": 1199
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4797426522385466,
+      "learning_rate": 1.9949576265260708e-05,
+      "loss": 0.2396,
+      "step": 1200
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3031654108455368,
+      "learning_rate": 1.994941094567723e-05,
+      "loss": 0.2342,
+      "step": 1201
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5757734679789772,
+      "learning_rate": 1.9949245356215415e-05,
+      "loss": 0.2518,
+      "step": 1202
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2205859526102094,
+      "learning_rate": 1.9949079496879763e-05,
+      "loss": 0.2384,
+      "step": 1203
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3153888292338682,
+      "learning_rate": 1.9948913367674766e-05,
+      "loss": 0.2674,
+      "step": 1204
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.118783886908271,
+      "learning_rate": 1.994874696860493e-05,
+      "loss": 0.2526,
+      "step": 1205
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.51742792693216,
+      "learning_rate": 1.9948580299674774e-05,
+      "loss": 0.2485,
+      "step": 1206
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0380364345723683,
+      "learning_rate": 1.994841336088881e-05,
+      "loss": 0.2567,
+      "step": 1207
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1678784841168433,
+      "learning_rate": 1.9948246152251576e-05,
+      "loss": 0.2431,
+      "step": 1208
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3366080159215235,
+      "learning_rate": 1.9948078673767604e-05,
+      "loss": 0.2875,
+      "step": 1209
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.63814443418371,
+      "learning_rate": 1.9947910925441435e-05,
+      "loss": 0.2752,
+      "step": 1210
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2934593570974755,
+      "learning_rate": 1.9947742907277617e-05,
+      "loss": 0.2762,
+      "step": 1211
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.348419593558008,
+      "learning_rate": 1.9947574619280713e-05,
+      "loss": 0.2296,
+      "step": 1212
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9139456450488086,
+      "learning_rate": 1.9947406061455287e-05,
+      "loss": 0.2519,
+      "step": 1213
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9703076073163888,
+      "learning_rate": 1.994723723380591e-05,
+      "loss": 0.2621,
+      "step": 1214
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3465532732447363,
+      "learning_rate": 1.994706813633716e-05,
+      "loss": 0.2558,
+      "step": 1215
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.284682342016494,
+      "learning_rate": 1.9946898769053625e-05,
+      "loss": 0.2363,
+      "step": 1216
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1236941900852206,
+      "learning_rate": 1.9946729131959902e-05,
+      "loss": 0.2692,
+      "step": 1217
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3346742349950482,
+      "learning_rate": 1.9946559225060585e-05,
+      "loss": 0.2413,
+      "step": 1218
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2810878909267052,
+      "learning_rate": 1.9946389048360288e-05,
+      "loss": 0.2243,
+      "step": 1219
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2926565224995998,
+      "learning_rate": 1.9946218601863626e-05,
+      "loss": 0.2635,
+      "step": 1220
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3114698308075319,
+      "learning_rate": 1.9946047885575224e-05,
+      "loss": 0.2714,
+      "step": 1221
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1178514386861154,
+      "learning_rate": 1.9945876899499712e-05,
+      "loss": 0.2486,
+      "step": 1222
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.114371429943034,
+      "learning_rate": 1.9945705643641727e-05,
+      "loss": 0.2325,
+      "step": 1223
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4891443815170118,
+      "learning_rate": 1.9945534118005913e-05,
+      "loss": 0.2446,
+      "step": 1224
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2381114684264194,
+      "learning_rate": 1.9945362322596926e-05,
+      "loss": 0.2606,
+      "step": 1225
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1075696539896358,
+      "learning_rate": 1.9945190257419424e-05,
+      "loss": 0.2496,
+      "step": 1226
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2353450742088958,
+      "learning_rate": 1.9945017922478076e-05,
+      "loss": 0.2721,
+      "step": 1227
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.225499079059848,
+      "learning_rate": 1.994484531777755e-05,
+      "loss": 0.2375,
+      "step": 1228
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3675113150352771,
+      "learning_rate": 1.994467244332254e-05,
+      "loss": 0.2543,
+      "step": 1229
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.483126597664391,
+      "learning_rate": 1.9944499299117724e-05,
+      "loss": 0.2659,
+      "step": 1230
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4329247239553178,
+      "learning_rate": 1.9944325885167807e-05,
+      "loss": 0.2541,
+      "step": 1231
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1956395859727627,
+      "learning_rate": 1.9944152201477483e-05,
+      "loss": 0.2331,
+      "step": 1232
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0789307326922348,
+      "learning_rate": 1.994397824805147e-05,
+      "loss": 0.2247,
+      "step": 1233
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2295779281449082,
+      "learning_rate": 1.9943804024894486e-05,
+      "loss": 0.2609,
+      "step": 1234
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1590034791235422,
+      "learning_rate": 1.994362953201126e-05,
+      "loss": 0.2666,
+      "step": 1235
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0564212568535971,
+      "learning_rate": 1.9943454769406515e-05,
+      "loss": 0.2412,
+      "step": 1236
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5078051441569378,
+      "learning_rate": 1.9943279737085003e-05,
+      "loss": 0.2494,
+      "step": 1237
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.216524313899798,
+      "learning_rate": 1.9943104435051466e-05,
+      "loss": 0.2622,
+      "step": 1238
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5140052363285328,
+      "learning_rate": 1.994292886331066e-05,
+      "loss": 0.2611,
+      "step": 1239
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2042848747214754,
+      "learning_rate": 1.994275302186734e-05,
+      "loss": 0.2446,
+      "step": 1240
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2721633003954307,
+      "learning_rate": 1.994257691072629e-05,
+      "loss": 0.252,
+      "step": 1241
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1749578657632214,
+      "learning_rate": 1.994240052989228e-05,
+      "loss": 0.2317,
+      "step": 1242
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2140607390698037,
+      "learning_rate": 1.994222387937009e-05,
+      "loss": 0.2465,
+      "step": 1243
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2318297370878173,
+      "learning_rate": 1.9942046959164516e-05,
+      "loss": 0.2477,
+      "step": 1244
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0822095702256438,
+      "learning_rate": 1.994186976928036e-05,
+      "loss": 0.2563,
+      "step": 1245
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2863539137070017,
+      "learning_rate": 1.9941692309722422e-05,
+      "loss": 0.2482,
+      "step": 1246
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9198636083932812,
+      "learning_rate": 1.994151458049552e-05,
+      "loss": 0.2487,
+      "step": 1247
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0390473686422546,
+      "learning_rate": 1.9941336581604474e-05,
+      "loss": 0.2725,
+      "step": 1248
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9622996627966769,
+      "learning_rate": 1.994115831305411e-05,
+      "loss": 0.2508,
+      "step": 1249
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1572359650318538,
+      "learning_rate": 1.9940979774849264e-05,
+      "loss": 0.2259,
+      "step": 1250
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0814302644289238,
+      "learning_rate": 1.9940800966994785e-05,
+      "loss": 0.2188,
+      "step": 1251
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1509504124606063,
+      "learning_rate": 1.9940621889495516e-05,
+      "loss": 0.2391,
+      "step": 1252
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1232777139387475,
+      "learning_rate": 1.9940442542356315e-05,
+      "loss": 0.2296,
+      "step": 1253
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1470671685669127,
+      "learning_rate": 1.9940262925582052e-05,
+      "loss": 0.2782,
+      "step": 1254
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4878598163784487,
+      "learning_rate": 1.9940083039177594e-05,
+      "loss": 0.2556,
+      "step": 1255
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.419684202169054,
+      "learning_rate": 1.993990288314782e-05,
+      "loss": 0.2509,
+      "step": 1256
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0335157524606504,
+      "learning_rate": 1.9939722457497625e-05,
+      "loss": 0.2177,
+      "step": 1257
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1965718148733537,
+      "learning_rate": 1.993954176223189e-05,
+      "loss": 0.268,
+      "step": 1258
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0653255609824037,
+      "learning_rate": 1.9939360797355527e-05,
+      "loss": 0.2258,
+      "step": 1259
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0147997894620475,
+      "learning_rate": 1.9939179562873437e-05,
+      "loss": 0.2402,
+      "step": 1260
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6590099622321888,
+      "learning_rate": 1.9938998058790546e-05,
+      "loss": 0.241,
+      "step": 1261
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9272334105915023,
+      "learning_rate": 1.9938816285111768e-05,
+      "loss": 0.2633,
+      "step": 1262
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9696781916820303,
+      "learning_rate": 1.9938634241842037e-05,
+      "loss": 0.2746,
+      "step": 1263
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.075879149837418,
+      "learning_rate": 1.993845192898629e-05,
+      "loss": 0.2481,
+      "step": 1264
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.050216008594986,
+      "learning_rate": 1.9938269346549473e-05,
+      "loss": 0.2562,
+      "step": 1265
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0603398741984067,
+      "learning_rate": 1.993808649453654e-05,
+      "loss": 0.2201,
+      "step": 1266
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1759766212394545,
+      "learning_rate": 1.993790337295245e-05,
+      "loss": 0.2338,
+      "step": 1267
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3895706761656446,
+      "learning_rate": 1.993771998180217e-05,
+      "loss": 0.2334,
+      "step": 1268
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.007624866161374,
+      "learning_rate": 1.9937536321090673e-05,
+      "loss": 0.2513,
+      "step": 1269
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.468659081059867,
+      "learning_rate": 1.9937352390822945e-05,
+      "loss": 0.2439,
+      "step": 1270
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.126146886783081,
+      "learning_rate": 1.993716819100397e-05,
+      "loss": 0.2822,
+      "step": 1271
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9947140946371513,
+      "learning_rate": 1.9936983721638745e-05,
+      "loss": 0.2452,
+      "step": 1272
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.022459747505446,
+      "learning_rate": 1.9936798982732274e-05,
+      "loss": 0.2747,
+      "step": 1273
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0124242493363043,
+      "learning_rate": 1.9936613974289575e-05,
+      "loss": 0.2578,
+      "step": 1274
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7048341660404465,
+      "learning_rate": 1.9936428696315656e-05,
+      "loss": 0.2352,
+      "step": 1275
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0907511105393048,
+      "learning_rate": 1.993624314881555e-05,
+      "loss": 0.265,
+      "step": 1276
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.08612995999494,
+      "learning_rate": 1.9936057331794284e-05,
+      "loss": 0.2474,
+      "step": 1277
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0279733992904583,
+      "learning_rate": 1.9935871245256907e-05,
+      "loss": 0.2478,
+      "step": 1278
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3072664522397148,
+      "learning_rate": 1.9935684889208455e-05,
+      "loss": 0.244,
+      "step": 1279
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9953628542258297,
+      "learning_rate": 1.9935498263653994e-05,
+      "loss": 0.2457,
+      "step": 1280
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8782714215203581,
+      "learning_rate": 1.993531136859858e-05,
+      "loss": 0.2319,
+      "step": 1281
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0964019610858486,
+      "learning_rate": 1.9935124204047283e-05,
+      "loss": 0.2362,
+      "step": 1282
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1545713070308408,
+      "learning_rate": 1.9934936770005184e-05,
+      "loss": 0.2196,
+      "step": 1283
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7314670261178784,
+      "learning_rate": 1.993474906647736e-05,
+      "loss": 0.2384,
+      "step": 1284
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0590502228754426,
+      "learning_rate": 1.993456109346891e-05,
+      "loss": 0.2324,
+      "step": 1285
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0498547537114502,
+      "learning_rate": 1.9934372850984925e-05,
+      "loss": 0.2585,
+      "step": 1286
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1848581217189293,
+      "learning_rate": 1.9934184339030517e-05,
+      "loss": 0.2664,
+      "step": 1287
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.072168224985916,
+      "learning_rate": 1.99339955576108e-05,
+      "loss": 0.2768,
+      "step": 1288
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3175290108699542,
+      "learning_rate": 1.993380650673089e-05,
+      "loss": 0.2889,
+      "step": 1289
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9720547354553507,
+      "learning_rate": 1.9933617186395917e-05,
+      "loss": 0.2231,
+      "step": 1290
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.373607960688602,
+      "learning_rate": 1.993342759661102e-05,
+      "loss": 0.2704,
+      "step": 1291
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3201599152256596,
+      "learning_rate": 1.9933237737381336e-05,
+      "loss": 0.2411,
+      "step": 1292
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.090537077664811,
+      "learning_rate": 1.993304760871202e-05,
+      "loss": 0.2439,
+      "step": 1293
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2988085961178586,
+      "learning_rate": 1.993285721060822e-05,
+      "loss": 0.2552,
+      "step": 1294
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.12247096396022,
+      "learning_rate": 1.9932666543075113e-05,
+      "loss": 0.2555,
+      "step": 1295
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.474840116009373,
+      "learning_rate": 1.9932475606117865e-05,
+      "loss": 0.2577,
+      "step": 1296
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5661222125417564,
+      "learning_rate": 1.9932284399741653e-05,
+      "loss": 0.2386,
+      "step": 1297
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3560953139033376,
+      "learning_rate": 1.9932092923951667e-05,
+      "loss": 0.2418,
+      "step": 1298
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1428106762278663,
+      "learning_rate": 1.99319011787531e-05,
+      "loss": 0.2286,
+      "step": 1299
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6241541244210773,
+      "learning_rate": 1.993170916415115e-05,
+      "loss": 0.2679,
+      "step": 1300
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1611966906057736,
+      "learning_rate": 1.993151688015103e-05,
+      "loss": 0.2799,
+      "step": 1301
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.06454836864526,
+      "learning_rate": 1.993132432675795e-05,
+      "loss": 0.224,
+      "step": 1302
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4358338142793268,
+      "learning_rate": 1.993113150397714e-05,
+      "loss": 0.2535,
+      "step": 1303
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1185775618889204,
+      "learning_rate": 1.993093841181383e-05,
+      "loss": 0.2704,
+      "step": 1304
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8895921291367405,
+      "learning_rate": 1.993074505027325e-05,
+      "loss": 0.2508,
+      "step": 1305
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9817015060406692,
+      "learning_rate": 1.9930551419360653e-05,
+      "loss": 0.2346,
+      "step": 1306
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9722902810607787,
+      "learning_rate": 1.9930357519081286e-05,
+      "loss": 0.2384,
+      "step": 1307
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.250573129564375,
+      "learning_rate": 1.993016334944041e-05,
+      "loss": 0.2486,
+      "step": 1308
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1498427550944075,
+      "learning_rate": 1.9929968910443294e-05,
+      "loss": 0.2396,
+      "step": 1309
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0067228344309367,
+      "learning_rate": 1.992977420209521e-05,
+      "loss": 0.2576,
+      "step": 1310
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2293727519204154,
+      "learning_rate": 1.9929579224401436e-05,
+      "loss": 0.2297,
+      "step": 1311
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5442224539939846,
+      "learning_rate": 1.992938397736727e-05,
+      "loss": 0.2366,
+      "step": 1312
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3856656195152042,
+      "learning_rate": 1.9929188460998e-05,
+      "loss": 0.2534,
+      "step": 1313
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0568400725170732,
+      "learning_rate": 1.992899267529893e-05,
+      "loss": 0.2555,
+      "step": 1314
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3167276569636372,
+      "learning_rate": 1.9928796620275377e-05,
+      "loss": 0.2651,
+      "step": 1315
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5992657211036883,
+      "learning_rate": 1.9928600295932655e-05,
+      "loss": 0.2644,
+      "step": 1316
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2137062647139218,
+      "learning_rate": 1.992840370227609e-05,
+      "loss": 0.2527,
+      "step": 1317
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1268154425617247,
+      "learning_rate": 1.992820683931101e-05,
+      "loss": 0.2532,
+      "step": 1318
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.83134570568081,
+      "learning_rate": 1.992800970704276e-05,
+      "loss": 0.2679,
+      "step": 1319
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2780231232806325,
+      "learning_rate": 1.9927812305476685e-05,
+      "loss": 0.2515,
+      "step": 1320
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9825579835071978,
+      "learning_rate": 1.9927614634618142e-05,
+      "loss": 0.2491,
+      "step": 1321
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0270718091675821,
+      "learning_rate": 1.9927416694472493e-05,
+      "loss": 0.2598,
+      "step": 1322
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8900160649600508,
+      "learning_rate": 1.9927218485045103e-05,
+      "loss": 0.2355,
+      "step": 1323
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4166876201090004,
+      "learning_rate": 1.992702000634135e-05,
+      "loss": 0.2445,
+      "step": 1324
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9133842813076238,
+      "learning_rate": 1.9926821258366622e-05,
+      "loss": 0.2533,
+      "step": 1325
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1537899021882343,
+      "learning_rate": 1.9926622241126306e-05,
+      "loss": 0.2521,
+      "step": 1326
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.237239241656995,
+      "learning_rate": 1.99264229546258e-05,
+      "loss": 0.2769,
+      "step": 1327
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9828066104197971,
+      "learning_rate": 1.992622339887051e-05,
+      "loss": 0.2688,
+      "step": 1328
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9653344834028247,
+      "learning_rate": 1.992602357386585e-05,
+      "loss": 0.2613,
+      "step": 1329
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1792847900807026,
+      "learning_rate": 1.9925823479617242e-05,
+      "loss": 0.2516,
+      "step": 1330
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9584914944637997,
+      "learning_rate": 1.9925623116130105e-05,
+      "loss": 0.2474,
+      "step": 1331
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1377392233359997,
+      "learning_rate": 1.9925422483409886e-05,
+      "loss": 0.2352,
+      "step": 1332
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2468284883092173,
+      "learning_rate": 1.992522158146202e-05,
+      "loss": 0.241,
+      "step": 1333
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3281788584533107,
+      "learning_rate": 1.9925020410291963e-05,
+      "loss": 0.2457,
+      "step": 1334
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3139206487167114,
+      "learning_rate": 1.992481896990516e-05,
+      "loss": 0.2725,
+      "step": 1335
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1181599807995584,
+      "learning_rate": 1.9924617260307088e-05,
+      "loss": 0.2496,
+      "step": 1336
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.095318282833729,
+      "learning_rate": 1.9924415281503204e-05,
+      "loss": 0.2635,
+      "step": 1337
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3388267621206682,
+      "learning_rate": 1.9924213033499e-05,
+      "loss": 0.254,
+      "step": 1338
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2570545418236239,
+      "learning_rate": 1.9924010516299956e-05,
+      "loss": 0.2343,
+      "step": 1339
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8375633870522983,
+      "learning_rate": 1.9923807729911567e-05,
+      "loss": 0.2485,
+      "step": 1340
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2080640469217854,
+      "learning_rate": 1.9923604674339336e-05,
+      "loss": 0.2477,
+      "step": 1341
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3859634411593529,
+      "learning_rate": 1.9923401349588762e-05,
+      "loss": 0.2701,
+      "step": 1342
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2651163070331262,
+      "learning_rate": 1.9923197755665368e-05,
+      "loss": 0.237,
+      "step": 1343
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.212850146390197,
+      "learning_rate": 1.9922993892574676e-05,
+      "loss": 0.2636,
+      "step": 1344
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5856155338509108,
+      "learning_rate": 1.9922789760322213e-05,
+      "loss": 0.2572,
+      "step": 1345
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3322458710609553,
+      "learning_rate": 1.9922585358913515e-05,
+      "loss": 0.2538,
+      "step": 1346
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1453763616701749,
+      "learning_rate": 1.992238068835413e-05,
+      "loss": 0.2127,
+      "step": 1347
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4671152149814002,
+      "learning_rate": 1.9922175748649612e-05,
+      "loss": 0.2467,
+      "step": 1348
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.96344423338482,
+      "learning_rate": 1.9921970539805513e-05,
+      "loss": 0.2449,
+      "step": 1349
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3164717303997437,
+      "learning_rate": 1.9921765061827405e-05,
+      "loss": 0.2527,
+      "step": 1350
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1273036629419297,
+      "learning_rate": 1.992155931472086e-05,
+      "loss": 0.2335,
+      "step": 1351
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3271998637630977,
+      "learning_rate": 1.9921353298491453e-05,
+      "loss": 0.233,
+      "step": 1352
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.083621975560061,
+      "learning_rate": 1.9921147013144782e-05,
+      "loss": 0.26,
+      "step": 1353
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1429372463643215,
+      "learning_rate": 1.9920940458686434e-05,
+      "loss": 0.2267,
+      "step": 1354
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3916262889864177,
+      "learning_rate": 1.992073363512202e-05,
+      "loss": 0.2518,
+      "step": 1355
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5964270594798335,
+      "learning_rate": 1.9920526542457143e-05,
+      "loss": 0.2459,
+      "step": 1356
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1391799157648104,
+      "learning_rate": 1.9920319180697422e-05,
+      "loss": 0.2825,
+      "step": 1357
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1659412582972448,
+      "learning_rate": 1.9920111549848486e-05,
+      "loss": 0.2372,
+      "step": 1358
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4113772014832484,
+      "learning_rate": 1.991990364991596e-05,
+      "loss": 0.2844,
+      "step": 1359
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2647682628121506,
+      "learning_rate": 1.991969548090549e-05,
+      "loss": 0.2307,
+      "step": 1360
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2706451265074896,
+      "learning_rate": 1.9919487042822722e-05,
+      "loss": 0.2504,
+      "step": 1361
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3043546137160325,
+      "learning_rate": 1.9919278335673306e-05,
+      "loss": 0.2653,
+      "step": 1362
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.526648857603099,
+      "learning_rate": 1.9919069359462906e-05,
+      "loss": 0.2352,
+      "step": 1363
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9163989199581128,
+      "learning_rate": 1.9918860114197186e-05,
+      "loss": 0.2623,
+      "step": 1364
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2210780161581907,
+      "learning_rate": 1.9918650599881828e-05,
+      "loss": 0.2463,
+      "step": 1365
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2037239438393557,
+      "learning_rate": 1.9918440816522514e-05,
+      "loss": 0.2681,
+      "step": 1366
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2841089377890307,
+      "learning_rate": 1.991823076412493e-05,
+      "loss": 0.25,
+      "step": 1367
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2099082139270674,
+      "learning_rate": 1.9918020442694773e-05,
+      "loss": 0.2374,
+      "step": 1368
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5448627661147787,
+      "learning_rate": 1.9917809852237754e-05,
+      "loss": 0.2389,
+      "step": 1369
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3484794901039256,
+      "learning_rate": 1.9917598992759587e-05,
+      "loss": 0.2431,
+      "step": 1370
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3696444473216656,
+      "learning_rate": 1.9917387864265983e-05,
+      "loss": 0.2469,
+      "step": 1371
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1573883129027844,
+      "learning_rate": 1.9917176466762673e-05,
+      "loss": 0.2794,
+      "step": 1372
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9521797753529773,
+      "learning_rate": 1.991696480025539e-05,
+      "loss": 0.2609,
+      "step": 1373
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9701217511012938,
+      "learning_rate": 1.991675286474988e-05,
+      "loss": 0.2452,
+      "step": 1374
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2255199211334986,
+      "learning_rate": 1.9916540660251887e-05,
+      "loss": 0.2657,
+      "step": 1375
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2802572417504723,
+      "learning_rate": 1.9916328186767168e-05,
+      "loss": 0.2507,
+      "step": 1376
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2780752266677757,
+      "learning_rate": 1.9916115444301488e-05,
+      "loss": 0.2475,
+      "step": 1377
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1917746479811409,
+      "learning_rate": 1.9915902432860615e-05,
+      "loss": 0.2778,
+      "step": 1378
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1122713843909182,
+      "learning_rate": 1.9915689152450328e-05,
+      "loss": 0.257,
+      "step": 1379
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.372238893365632,
+      "learning_rate": 1.9915475603076414e-05,
+      "loss": 0.2623,
+      "step": 1380
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3604773532215377,
+      "learning_rate": 1.9915261784744664e-05,
+      "loss": 0.2317,
+      "step": 1381
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.200145793587868,
+      "learning_rate": 1.9915047697460878e-05,
+      "loss": 0.2662,
+      "step": 1382
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0781961411781062,
+      "learning_rate": 1.9914833341230863e-05,
+      "loss": 0.2551,
+      "step": 1383
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9837813270598623,
+      "learning_rate": 1.9914618716060437e-05,
+      "loss": 0.259,
+      "step": 1384
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1325359219128366,
+      "learning_rate": 1.9914403821955414e-05,
+      "loss": 0.2386,
+      "step": 1385
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0389864770057453,
+      "learning_rate": 1.9914188658921628e-05,
+      "loss": 0.236,
+      "step": 1386
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3836282909574422,
+      "learning_rate": 1.9913973226964917e-05,
+      "loss": 0.2576,
+      "step": 1387
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0113155814154398,
+      "learning_rate": 1.991375752609112e-05,
+      "loss": 0.2626,
+      "step": 1388
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0919641731577683,
+      "learning_rate": 1.991354155630609e-05,
+      "loss": 0.2562,
+      "step": 1389
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4916606982981637,
+      "learning_rate": 1.9913325317615684e-05,
+      "loss": 0.2225,
+      "step": 1390
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.269965773751471,
+      "learning_rate": 1.9913108810025776e-05,
+      "loss": 0.2719,
+      "step": 1391
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1927198996440365,
+      "learning_rate": 1.9912892033542225e-05,
+      "loss": 0.2448,
+      "step": 1392
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0247277433743283,
+      "learning_rate": 1.991267498817092e-05,
+      "loss": 0.2456,
+      "step": 1393
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.999133874275022,
+      "learning_rate": 1.9912457673917745e-05,
+      "loss": 0.2409,
+      "step": 1394
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.160387438965137,
+      "learning_rate": 1.9912240090788595e-05,
+      "loss": 0.2619,
+      "step": 1395
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.208793083153623,
+      "learning_rate": 1.9912022238789374e-05,
+      "loss": 0.242,
+      "step": 1396
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8843530643442636,
+      "learning_rate": 1.991180411792599e-05,
+      "loss": 0.2707,
+      "step": 1397
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1037363784292973,
+      "learning_rate": 1.9911585728204362e-05,
+      "loss": 0.2335,
+      "step": 1398
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.954476589084465,
+      "learning_rate": 1.9911367069630408e-05,
+      "loss": 0.2311,
+      "step": 1399
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0015353333121517,
+      "learning_rate": 1.991114814221006e-05,
+      "loss": 0.2325,
+      "step": 1400
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9980972684885128,
+      "learning_rate": 1.9910928945949264e-05,
+      "loss": 0.2394,
+      "step": 1401
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7186452306443267,
+      "learning_rate": 1.9910709480853957e-05,
+      "loss": 0.2424,
+      "step": 1402
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9696052813958719,
+      "learning_rate": 1.9910489746930097e-05,
+      "loss": 0.26,
+      "step": 1403
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7690335657622046,
+      "learning_rate": 1.9910269744183645e-05,
+      "loss": 0.2129,
+      "step": 1404
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1975760811845586,
+      "learning_rate": 1.9910049472620564e-05,
+      "loss": 0.2611,
+      "step": 1405
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8503953892746059,
+      "learning_rate": 1.990982893224683e-05,
+      "loss": 0.2246,
+      "step": 1406
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.847893913853286,
+      "learning_rate": 1.990960812306843e-05,
+      "loss": 0.2194,
+      "step": 1407
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9348019252601966,
+      "learning_rate": 1.990938704509135e-05,
+      "loss": 0.221,
+      "step": 1408
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9761686610221613,
+      "learning_rate": 1.9909165698321585e-05,
+      "loss": 0.243,
+      "step": 1409
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1722248378675564,
+      "learning_rate": 1.990894408276514e-05,
+      "loss": 0.2547,
+      "step": 1410
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8053092328818088,
+      "learning_rate": 1.9908722198428027e-05,
+      "loss": 0.227,
+      "step": 1411
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9665903219728691,
+      "learning_rate": 1.9908500045316264e-05,
+      "loss": 0.2451,
+      "step": 1412
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.953084935420593,
+      "learning_rate": 1.9908277623435878e-05,
+      "loss": 0.2318,
+      "step": 1413
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5767742424384938,
+      "learning_rate": 1.9908054932792903e-05,
+      "loss": 0.2865,
+      "step": 1414
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0425077282991027,
+      "learning_rate": 1.9907831973393377e-05,
+      "loss": 0.2654,
+      "step": 1415
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9923093106371809,
+      "learning_rate": 1.9907608745243356e-05,
+      "loss": 0.2429,
+      "step": 1416
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.104082202213175,
+      "learning_rate": 1.9907385248348882e-05,
+      "loss": 0.2542,
+      "step": 1417
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1450188139943347,
+      "learning_rate": 1.990716148271602e-05,
+      "loss": 0.2285,
+      "step": 1418
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0882818233930411,
+      "learning_rate": 1.990693744835085e-05,
+      "loss": 0.2331,
+      "step": 1419
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0449521038261802,
+      "learning_rate": 1.990671314525944e-05,
+      "loss": 0.2442,
+      "step": 1420
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1106530955831952,
+      "learning_rate": 1.9906488573447875e-05,
+      "loss": 0.2584,
+      "step": 1421
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0370821761153255,
+      "learning_rate": 1.990626373292225e-05,
+      "loss": 0.2335,
+      "step": 1422
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.13841965133708,
+      "learning_rate": 1.9906038623688658e-05,
+      "loss": 0.2388,
+      "step": 1423
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9337248189403204,
+      "learning_rate": 1.9905813245753214e-05,
+      "loss": 0.253,
+      "step": 1424
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8239707311358239,
+      "learning_rate": 1.9905587599122022e-05,
+      "loss": 0.2417,
+      "step": 1425
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8948997842459009,
+      "learning_rate": 1.990536168380121e-05,
+      "loss": 0.2574,
+      "step": 1426
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.344858167440341,
+      "learning_rate": 1.9905135499796903e-05,
+      "loss": 0.2557,
+      "step": 1427
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8315357711462754,
+      "learning_rate": 1.9904909047115233e-05,
+      "loss": 0.267,
+      "step": 1428
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7994238898963966,
+      "learning_rate": 1.990468232576235e-05,
+      "loss": 0.2377,
+      "step": 1429
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8653614776519574,
+      "learning_rate": 1.9904455335744395e-05,
+      "loss": 0.2432,
+      "step": 1430
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8806812047415478,
+      "learning_rate": 1.990422807706753e-05,
+      "loss": 0.2571,
+      "step": 1431
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9214838257317811,
+      "learning_rate": 1.990400054973792e-05,
+      "loss": 0.2043,
+      "step": 1432
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8867702484440099,
+      "learning_rate": 1.9903772753761736e-05,
+      "loss": 0.2562,
+      "step": 1433
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8546974694213872,
+      "learning_rate": 1.990354468914516e-05,
+      "loss": 0.2408,
+      "step": 1434
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1024048985123924,
+      "learning_rate": 1.990331635589437e-05,
+      "loss": 0.2714,
+      "step": 1435
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1446317404952169,
+      "learning_rate": 1.9903087754015567e-05,
+      "loss": 0.2347,
+      "step": 1436
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.919739139874813,
+      "learning_rate": 1.9902858883514948e-05,
+      "loss": 0.2527,
+      "step": 1437
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.435959400311835,
+      "learning_rate": 1.990262974439872e-05,
+      "loss": 0.2464,
+      "step": 1438
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8152490838875965,
+      "learning_rate": 1.9902400336673107e-05,
+      "loss": 0.2828,
+      "step": 1439
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9967524541458044,
+      "learning_rate": 1.9902170660344323e-05,
+      "loss": 0.2598,
+      "step": 1440
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.066795566885315,
+      "learning_rate": 1.99019407154186e-05,
+      "loss": 0.2262,
+      "step": 1441
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0092050775090406,
+      "learning_rate": 1.9901710501902177e-05,
+      "loss": 0.2339,
+      "step": 1442
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1680892173535489,
+      "learning_rate": 1.9901480019801297e-05,
+      "loss": 0.217,
+      "step": 1443
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9901721765525188,
+      "learning_rate": 1.990124926912221e-05,
+      "loss": 0.2208,
+      "step": 1444
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3413707759726372,
+      "learning_rate": 1.990101824987118e-05,
+      "loss": 0.2542,
+      "step": 1445
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0970427933144882,
+      "learning_rate": 1.9900786962054468e-05,
+      "loss": 0.2431,
+      "step": 1446
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1698370073558522,
+      "learning_rate": 1.9900555405678354e-05,
+      "loss": 0.264,
+      "step": 1447
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.996959606051725,
+      "learning_rate": 1.990032358074911e-05,
+      "loss": 0.2658,
+      "step": 1448
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3875556188415172,
+      "learning_rate": 1.9900091487273035e-05,
+      "loss": 0.2672,
+      "step": 1449
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1253872775336855,
+      "learning_rate": 1.9899859125256417e-05,
+      "loss": 0.2545,
+      "step": 1450
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.065186848280434,
+      "learning_rate": 1.989962649470556e-05,
+      "loss": 0.2541,
+      "step": 1451
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1268591743661116,
+      "learning_rate": 1.989939359562678e-05,
+      "loss": 0.2451,
+      "step": 1452
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2351628782733877,
+      "learning_rate": 1.9899160428026383e-05,
+      "loss": 0.2669,
+      "step": 1453
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2936651171469973,
+      "learning_rate": 1.9898926991910704e-05,
+      "loss": 0.2772,
+      "step": 1454
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2085038183262553,
+      "learning_rate": 1.989869328728607e-05,
+      "loss": 0.2427,
+      "step": 1455
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4147175272036487,
+      "learning_rate": 1.9898459314158825e-05,
+      "loss": 0.2632,
+      "step": 1456
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9223842251997182,
+      "learning_rate": 1.989822507253531e-05,
+      "loss": 0.221,
+      "step": 1457
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2893374006720397,
+      "learning_rate": 1.9897990562421882e-05,
+      "loss": 0.2401,
+      "step": 1458
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9799151191327158,
+      "learning_rate": 1.9897755783824897e-05,
+      "loss": 0.2362,
+      "step": 1459
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1257687249716646,
+      "learning_rate": 1.989752073675073e-05,
+      "loss": 0.2519,
+      "step": 1460
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0071942580571653,
+      "learning_rate": 1.9897285421205753e-05,
+      "loss": 0.2347,
+      "step": 1461
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.127580590734185,
+      "learning_rate": 1.989704983719635e-05,
+      "loss": 0.2469,
+      "step": 1462
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.490512732609794,
+      "learning_rate": 1.9896813984728915e-05,
+      "loss": 0.2626,
+      "step": 1463
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5208393177639197,
+      "learning_rate": 1.9896577863809836e-05,
+      "loss": 0.2562,
+      "step": 1464
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3154094170613637,
+      "learning_rate": 1.9896341474445526e-05,
+      "loss": 0.2369,
+      "step": 1465
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3400389440702367,
+      "learning_rate": 1.9896104816642393e-05,
+      "loss": 0.2435,
+      "step": 1466
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.245380768826678,
+      "learning_rate": 1.989586789040686e-05,
+      "loss": 0.2342,
+      "step": 1467
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0622069292754814,
+      "learning_rate": 1.9895630695745353e-05,
+      "loss": 0.2514,
+      "step": 1468
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3739800782133331,
+      "learning_rate": 1.98953932326643e-05,
+      "loss": 0.2709,
+      "step": 1469
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2438019019651285,
+      "learning_rate": 1.9895155501170153e-05,
+      "loss": 0.224,
+      "step": 1470
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1099411604601386,
+      "learning_rate": 1.9894917501269346e-05,
+      "loss": 0.2386,
+      "step": 1471
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0999919296078493,
+      "learning_rate": 1.989467923296835e-05,
+      "loss": 0.2301,
+      "step": 1472
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8744198630330524,
+      "learning_rate": 1.9894440696273615e-05,
+      "loss": 0.2443,
+      "step": 1473
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.462670102238129,
+      "learning_rate": 1.9894201891191624e-05,
+      "loss": 0.2574,
+      "step": 1474
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.522143438867789,
+      "learning_rate": 1.9893962817728842e-05,
+      "loss": 0.2598,
+      "step": 1475
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.66521405994223,
+      "learning_rate": 1.9893723475891762e-05,
+      "loss": 0.227,
+      "step": 1476
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.030802909005997,
+      "learning_rate": 1.9893483865686875e-05,
+      "loss": 0.2425,
+      "step": 1477
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.619965380448205,
+      "learning_rate": 1.989324398712068e-05,
+      "loss": 0.2383,
+      "step": 1478
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5818095937642507,
+      "learning_rate": 1.9893003840199677e-05,
+      "loss": 0.2516,
+      "step": 1479
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4259409101129912,
+      "learning_rate": 1.989276342493039e-05,
+      "loss": 0.2405,
+      "step": 1480
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9980110737147162,
+      "learning_rate": 1.989252274131934e-05,
+      "loss": 0.2323,
+      "step": 1481
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0821845943196076,
+      "learning_rate": 1.9892281789373047e-05,
+      "loss": 0.2628,
+      "step": 1482
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0011387530258766,
+      "learning_rate": 1.9892040569098054e-05,
+      "loss": 0.2427,
+      "step": 1483
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.296208187020809,
+      "learning_rate": 1.98917990805009e-05,
+      "loss": 0.2274,
+      "step": 1484
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.17330571312088,
+      "learning_rate": 1.989155732358814e-05,
+      "loss": 0.2519,
+      "step": 1485
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0333952943167788,
+      "learning_rate": 1.9891315298366327e-05,
+      "loss": 0.2308,
+      "step": 1486
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1104592349324152,
+      "learning_rate": 1.9891073004842026e-05,
+      "loss": 0.2546,
+      "step": 1487
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2835504367778445,
+      "learning_rate": 1.9890830443021814e-05,
+      "loss": 0.2398,
+      "step": 1488
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1519730418068233,
+      "learning_rate": 1.9890587612912268e-05,
+      "loss": 0.2599,
+      "step": 1489
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1704596933418903,
+      "learning_rate": 1.9890344514519974e-05,
+      "loss": 0.2392,
+      "step": 1490
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.349986371882224,
+      "learning_rate": 1.9890101147851526e-05,
+      "loss": 0.2604,
+      "step": 1491
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.097941104349231,
+      "learning_rate": 1.9889857512913523e-05,
+      "loss": 0.2556,
+      "step": 1492
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7466568104138178,
+      "learning_rate": 1.988961360971258e-05,
+      "loss": 0.2476,
+      "step": 1493
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1822046880072066,
+      "learning_rate": 1.988936943825531e-05,
+      "loss": 0.2595,
+      "step": 1494
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9970149311719365,
+      "learning_rate": 1.9889124998548332e-05,
+      "loss": 0.2517,
+      "step": 1495
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0518883109686619,
+      "learning_rate": 1.9888880290598282e-05,
+      "loss": 0.2647,
+      "step": 1496
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2107601881178482,
+      "learning_rate": 1.9888635314411797e-05,
+      "loss": 0.2647,
+      "step": 1497
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0645260886991201,
+      "learning_rate": 1.9888390069995516e-05,
+      "loss": 0.2764,
+      "step": 1498
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1982898262283437,
+      "learning_rate": 1.98881445573561e-05,
+      "loss": 0.243,
+      "step": 1499
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2710638079525047,
+      "learning_rate": 1.9887898776500203e-05,
+      "loss": 0.2552,
+      "step": 1500
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2871447436383632,
+      "learning_rate": 1.9887652727434492e-05,
+      "loss": 0.2446,
+      "step": 1501
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1434019027245008,
+      "learning_rate": 1.9887406410165644e-05,
+      "loss": 0.2352,
+      "step": 1502
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.405194585504932,
+      "learning_rate": 1.988715982470034e-05,
+      "loss": 0.2275,
+      "step": 1503
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4582424939786653,
+      "learning_rate": 1.9886912971045263e-05,
+      "loss": 0.216,
+      "step": 1504
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7604856830727356,
+      "learning_rate": 1.9886665849207116e-05,
+      "loss": 0.268,
+      "step": 1505
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0954405919257795,
+      "learning_rate": 1.98864184591926e-05,
+      "loss": 0.2502,
+      "step": 1506
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.226051267596389,
+      "learning_rate": 1.9886170801008423e-05,
+      "loss": 0.2505,
+      "step": 1507
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2075409268454134,
+      "learning_rate": 1.9885922874661308e-05,
+      "loss": 0.2461,
+      "step": 1508
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4044291962244422,
+      "learning_rate": 1.9885674680157974e-05,
+      "loss": 0.2331,
+      "step": 1509
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.892375728058062,
+      "learning_rate": 1.9885426217505154e-05,
+      "loss": 0.2579,
+      "step": 1510
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0109769323742304,
+      "learning_rate": 1.9885177486709595e-05,
+      "loss": 0.2407,
+      "step": 1511
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2523333881157754,
+      "learning_rate": 1.988492848777803e-05,
+      "loss": 0.2368,
+      "step": 1512
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9684049169883828,
+      "learning_rate": 1.9884679220717232e-05,
+      "loss": 0.238,
+      "step": 1513
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.985704737514277,
+      "learning_rate": 1.9884429685533947e-05,
+      "loss": 0.2292,
+      "step": 1514
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5076127760722757,
+      "learning_rate": 1.9884179882234946e-05,
+      "loss": 0.2582,
+      "step": 1515
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1491759101287482,
+      "learning_rate": 1.988392981082701e-05,
+      "loss": 0.2619,
+      "step": 1516
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.016203850638015,
+      "learning_rate": 1.9883679471316918e-05,
+      "loss": 0.2411,
+      "step": 1517
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.237079163922459,
+      "learning_rate": 1.9883428863711463e-05,
+      "loss": 0.2125,
+      "step": 1518
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1480551559538474,
+      "learning_rate": 1.9883177988017444e-05,
+      "loss": 0.2565,
+      "step": 1519
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1236030361228728,
+      "learning_rate": 1.9882926844241662e-05,
+      "loss": 0.2568,
+      "step": 1520
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0466955366612953,
+      "learning_rate": 1.988267543239093e-05,
+      "loss": 0.2243,
+      "step": 1521
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1433635714952062,
+      "learning_rate": 1.988242375247207e-05,
+      "loss": 0.2695,
+      "step": 1522
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9521221469115966,
+      "learning_rate": 1.9882171804491905e-05,
+      "loss": 0.2548,
+      "step": 1523
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.032790584538378,
+      "learning_rate": 1.9881919588457274e-05,
+      "loss": 0.219,
+      "step": 1524
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1202926836453848,
+      "learning_rate": 1.9881667104375018e-05,
+      "loss": 0.2256,
+      "step": 1525
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0362677561309366,
+      "learning_rate": 1.988141435225198e-05,
+      "loss": 0.2424,
+      "step": 1526
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.108121046698207,
+      "learning_rate": 1.9881161332095024e-05,
+      "loss": 0.259,
+      "step": 1527
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.106705413680249,
+      "learning_rate": 1.9880908043911004e-05,
+      "loss": 0.2494,
+      "step": 1528
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8919959929553708,
+      "learning_rate": 1.9880654487706797e-05,
+      "loss": 0.2384,
+      "step": 1529
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1403536176285372,
+      "learning_rate": 1.988040066348928e-05,
+      "loss": 0.2442,
+      "step": 1530
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0397301835981851,
+      "learning_rate": 1.9880146571265337e-05,
+      "loss": 0.2215,
+      "step": 1531
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0570264901588597,
+      "learning_rate": 1.987989221104186e-05,
+      "loss": 0.2186,
+      "step": 1532
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5454973258691305,
+      "learning_rate": 1.987963758282575e-05,
+      "loss": 0.2435,
+      "step": 1533
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.00095769815178,
+      "learning_rate": 1.987938268662391e-05,
+      "loss": 0.2398,
+      "step": 1534
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0314401299992415,
+      "learning_rate": 1.9879127522443255e-05,
+      "loss": 0.2516,
+      "step": 1535
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0239176880004133,
+      "learning_rate": 1.9878872090290714e-05,
+      "loss": 0.2209,
+      "step": 1536
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1516319522863183,
+      "learning_rate": 1.98786163901732e-05,
+      "loss": 0.2179,
+      "step": 1537
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0899641412385346,
+      "learning_rate": 1.9878360422097666e-05,
+      "loss": 0.2629,
+      "step": 1538
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1161182738196964,
+      "learning_rate": 1.9878104186071047e-05,
+      "loss": 0.2669,
+      "step": 1539
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9331629099333119,
+      "learning_rate": 1.9877847682100294e-05,
+      "loss": 0.2226,
+      "step": 1540
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8936707439041887,
+      "learning_rate": 1.9877590910192362e-05,
+      "loss": 0.2725,
+      "step": 1541
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.127775496155315,
+      "learning_rate": 1.987733387035422e-05,
+      "loss": 0.2563,
+      "step": 1542
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1260913141490958,
+      "learning_rate": 1.9877076562592844e-05,
+      "loss": 0.2322,
+      "step": 1543
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8980360219871005,
+      "learning_rate": 1.9876818986915202e-05,
+      "loss": 0.2206,
+      "step": 1544
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9796120458036948,
+      "learning_rate": 1.9876561143328287e-05,
+      "loss": 0.2516,
+      "step": 1545
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2629491204181673,
+      "learning_rate": 1.9876303031839094e-05,
+      "loss": 0.2083,
+      "step": 1546
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0874082802575349,
+      "learning_rate": 1.9876044652454627e-05,
+      "loss": 0.2649,
+      "step": 1547
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1914776664656248,
+      "learning_rate": 1.987578600518189e-05,
+      "loss": 0.2606,
+      "step": 1548
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9057171643487041,
+      "learning_rate": 1.9875527090027897e-05,
+      "loss": 0.2399,
+      "step": 1549
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0640987292525308,
+      "learning_rate": 1.9875267906999674e-05,
+      "loss": 0.2369,
+      "step": 1550
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1909353635895097,
+      "learning_rate": 1.987500845610425e-05,
+      "loss": 0.2415,
+      "step": 1551
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1777676002953559,
+      "learning_rate": 1.987474873734867e-05,
+      "loss": 0.2381,
+      "step": 1552
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0400464992950833,
+      "learning_rate": 1.987448875073997e-05,
+      "loss": 0.2272,
+      "step": 1553
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9159186299023466,
+      "learning_rate": 1.9874228496285203e-05,
+      "loss": 0.2326,
+      "step": 1554
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0776077764374352,
+      "learning_rate": 1.9873967973991432e-05,
+      "loss": 0.251,
+      "step": 1555
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4205363639530653,
+      "learning_rate": 1.987370718386572e-05,
+      "loss": 0.2427,
+      "step": 1556
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1243700994075487,
+      "learning_rate": 1.9873446125915145e-05,
+      "loss": 0.2768,
+      "step": 1557
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1028876890940438,
+      "learning_rate": 1.9873184800146785e-05,
+      "loss": 0.2489,
+      "step": 1558
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0126681453424553,
+      "learning_rate": 1.987292320656773e-05,
+      "loss": 0.2471,
+      "step": 1559
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9566294464744491,
+      "learning_rate": 1.9872661345185076e-05,
+      "loss": 0.2393,
+      "step": 1560
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3368115973005088,
+      "learning_rate": 1.9872399216005928e-05,
+      "loss": 0.2478,
+      "step": 1561
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.592650857695837,
+      "learning_rate": 1.9872136819037388e-05,
+      "loss": 0.2437,
+      "step": 1562
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1068509765308028,
+      "learning_rate": 1.9871874154286585e-05,
+      "loss": 0.2383,
+      "step": 1563
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0576437487873738,
+      "learning_rate": 1.987161122176063e-05,
+      "loss": 0.2348,
+      "step": 1564
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0079938384705214,
+      "learning_rate": 1.9871348021466673e-05,
+      "loss": 0.2355,
+      "step": 1565
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4585402407253452,
+      "learning_rate": 1.9871084553411835e-05,
+      "loss": 0.2369,
+      "step": 1566
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0320403920202317,
+      "learning_rate": 1.9870820817603276e-05,
+      "loss": 0.2191,
+      "step": 1567
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0270957769308535,
+      "learning_rate": 1.9870556814048145e-05,
+      "loss": 0.2427,
+      "step": 1568
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2074542042203558,
+      "learning_rate": 1.9870292542753596e-05,
+      "loss": 0.2504,
+      "step": 1569
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.942970746262455,
+      "learning_rate": 1.987002800372681e-05,
+      "loss": 0.2212,
+      "step": 1570
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9104940249554769,
+      "learning_rate": 1.9869763196974957e-05,
+      "loss": 0.2596,
+      "step": 1571
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8644721366170993,
+      "learning_rate": 1.986949812250522e-05,
+      "loss": 0.2258,
+      "step": 1572
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1098327646726607,
+      "learning_rate": 1.986923278032479e-05,
+      "loss": 0.2533,
+      "step": 1573
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4782949178290852,
+      "learning_rate": 1.986896717044086e-05,
+      "loss": 0.3009,
+      "step": 1574
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0298685807340697,
+      "learning_rate": 1.986870129286064e-05,
+      "loss": 0.2592,
+      "step": 1575
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2470618274705578,
+      "learning_rate": 1.986843514759134e-05,
+      "loss": 0.2426,
+      "step": 1576
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.266134386190633,
+      "learning_rate": 1.986816873464018e-05,
+      "loss": 0.2532,
+      "step": 1577
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4067436040344086,
+      "learning_rate": 1.9867902054014382e-05,
+      "loss": 0.279,
+      "step": 1578
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8698043011661225,
+      "learning_rate": 1.986763510572119e-05,
+      "loss": 0.2399,
+      "step": 1579
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0193089125859338,
+      "learning_rate": 1.986736788976783e-05,
+      "loss": 0.2277,
+      "step": 1580
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.17900717751552,
+      "learning_rate": 1.9867100406161563e-05,
+      "loss": 0.2609,
+      "step": 1581
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.062165241963998,
+      "learning_rate": 1.986683265490964e-05,
+      "loss": 0.2747,
+      "step": 1582
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.137051073249643,
+      "learning_rate": 1.9866564636019326e-05,
+      "loss": 0.2288,
+      "step": 1583
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1032651435013483,
+      "learning_rate": 1.9866296349497885e-05,
+      "loss": 0.2267,
+      "step": 1584
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9910774254352182,
+      "learning_rate": 1.9866027795352604e-05,
+      "loss": 0.2767,
+      "step": 1585
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0635540904541145,
+      "learning_rate": 1.986575897359076e-05,
+      "loss": 0.2722,
+      "step": 1586
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9511029033089041,
+      "learning_rate": 1.9865489884219643e-05,
+      "loss": 0.2284,
+      "step": 1587
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.139944510002614,
+      "learning_rate": 1.9865220527246556e-05,
+      "loss": 0.2355,
+      "step": 1588
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3021053893170482,
+      "learning_rate": 1.986495090267881e-05,
+      "loss": 0.2657,
+      "step": 1589
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0402349598351108,
+      "learning_rate": 1.986468101052371e-05,
+      "loss": 0.2491,
+      "step": 1590
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9833347233482201,
+      "learning_rate": 1.986441085078858e-05,
+      "loss": 0.2408,
+      "step": 1591
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9702896415133163,
+      "learning_rate": 1.986414042348075e-05,
+      "loss": 0.2296,
+      "step": 1592
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8403703196326306,
+      "learning_rate": 1.9863869728607553e-05,
+      "loss": 0.2327,
+      "step": 1593
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9887059891575193,
+      "learning_rate": 1.986359876617633e-05,
+      "loss": 0.2416,
+      "step": 1594
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.15694928653404,
+      "learning_rate": 1.9863327536194438e-05,
+      "loss": 0.2196,
+      "step": 1595
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0068578045610728,
+      "learning_rate": 1.9863056038669225e-05,
+      "loss": 0.2533,
+      "step": 1596
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9916793991803713,
+      "learning_rate": 1.9862784273608066e-05,
+      "loss": 0.2549,
+      "step": 1597
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0031710272512935,
+      "learning_rate": 1.986251224101832e-05,
+      "loss": 0.231,
+      "step": 1598
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.021474671000375,
+      "learning_rate": 1.9862239940907377e-05,
+      "loss": 0.2491,
+      "step": 1599
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8310914145791256,
+      "learning_rate": 1.986196737328262e-05,
+      "loss": 0.2451,
+      "step": 1600
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8690234167987716,
+      "learning_rate": 1.9861694538151436e-05,
+      "loss": 0.2687,
+      "step": 1601
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1227562514141765,
+      "learning_rate": 1.9861421435521234e-05,
+      "loss": 0.2742,
+      "step": 1602
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9654518363415745,
+      "learning_rate": 1.9861148065399414e-05,
+      "loss": 0.2318,
+      "step": 1603
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1076305396603072,
+      "learning_rate": 1.98608744277934e-05,
+      "loss": 0.2421,
+      "step": 1604
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4547345595987675,
+      "learning_rate": 1.986060052271061e-05,
+      "loss": 0.2305,
+      "step": 1605
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8491699775291529,
+      "learning_rate": 1.9860326350158472e-05,
+      "loss": 0.225,
+      "step": 1606
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8319543451433087,
+      "learning_rate": 1.9860051910144426e-05,
+      "loss": 0.2435,
+      "step": 1607
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8100604809756528,
+      "learning_rate": 1.9859777202675915e-05,
+      "loss": 0.2111,
+      "step": 1608
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9527615886593352,
+      "learning_rate": 1.985950222776039e-05,
+      "loss": 0.2644,
+      "step": 1609
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0894915270026155,
+      "learning_rate": 1.9859226985405312e-05,
+      "loss": 0.2577,
+      "step": 1610
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.215983338220503,
+      "learning_rate": 1.985895147561814e-05,
+      "loss": 0.2239,
+      "step": 1611
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9302355264268377,
+      "learning_rate": 1.985867569840636e-05,
+      "loss": 0.2541,
+      "step": 1612
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.600810373609161,
+      "learning_rate": 1.985839965377744e-05,
+      "loss": 0.24,
+      "step": 1613
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2071590122653046,
+      "learning_rate": 1.9858123341738877e-05,
+      "loss": 0.255,
+      "step": 1614
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.078164560236254,
+      "learning_rate": 1.9857846762298157e-05,
+      "loss": 0.2397,
+      "step": 1615
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2358085343430798,
+      "learning_rate": 1.9857569915462787e-05,
+      "loss": 0.2642,
+      "step": 1616
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.136443380213396,
+      "learning_rate": 1.9857292801240278e-05,
+      "loss": 0.2348,
+      "step": 1617
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.256005542958859,
+      "learning_rate": 1.985701541963814e-05,
+      "loss": 0.2263,
+      "step": 1618
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9861752701380436,
+      "learning_rate": 1.9856737770663908e-05,
+      "loss": 0.2171,
+      "step": 1619
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0699500859875848,
+      "learning_rate": 1.9856459854325108e-05,
+      "loss": 0.2312,
+      "step": 1620
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0277239815636015,
+      "learning_rate": 1.9856181670629272e-05,
+      "loss": 0.2221,
+      "step": 1621
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1622788998264484,
+      "learning_rate": 1.985590321958396e-05,
+      "loss": 0.2658,
+      "step": 1622
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9654394119932761,
+      "learning_rate": 1.985562450119671e-05,
+      "loss": 0.2127,
+      "step": 1623
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8818972933724379,
+      "learning_rate": 1.985534551547509e-05,
+      "loss": 0.2402,
+      "step": 1624
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0822541435500859,
+      "learning_rate": 1.9855066262426663e-05,
+      "loss": 0.245,
+      "step": 1625
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0792182766553684,
+      "learning_rate": 1.9854786742059012e-05,
+      "loss": 0.2229,
+      "step": 1626
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9164683009523013,
+      "learning_rate": 1.9854506954379714e-05,
+      "loss": 0.2442,
+      "step": 1627
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1155262220942679,
+      "learning_rate": 1.9854226899396356e-05,
+      "loss": 0.2657,
+      "step": 1628
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3021145519257342,
+      "learning_rate": 1.9853946577116536e-05,
+      "loss": 0.23,
+      "step": 1629
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1384442147020182,
+      "learning_rate": 1.985366598754786e-05,
+      "loss": 0.2619,
+      "step": 1630
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5198425729758447,
+      "learning_rate": 1.985338513069794e-05,
+      "loss": 0.2424,
+      "step": 1631
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2746960811741264,
+      "learning_rate": 1.9853104006574387e-05,
+      "loss": 0.2398,
+      "step": 1632
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9081894042267109,
+      "learning_rate": 1.9852822615184835e-05,
+      "loss": 0.2712,
+      "step": 1633
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9654287887774892,
+      "learning_rate": 1.9852540956536912e-05,
+      "loss": 0.2482,
+      "step": 1634
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9874173268909807,
+      "learning_rate": 1.985225903063826e-05,
+      "loss": 0.2628,
+      "step": 1635
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8587542362193802,
+      "learning_rate": 1.9851976837496522e-05,
+      "loss": 0.2567,
+      "step": 1636
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8944630553655413,
+      "learning_rate": 1.9851694377119358e-05,
+      "loss": 0.2313,
+      "step": 1637
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9580136632504814,
+      "learning_rate": 1.985141164951443e-05,
+      "loss": 0.2637,
+      "step": 1638
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0240173091734972,
+      "learning_rate": 1.98511286546894e-05,
+      "loss": 0.2588,
+      "step": 1639
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1632183346348985,
+      "learning_rate": 1.985084539265195e-05,
+      "loss": 0.2492,
+      "step": 1640
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2418168208945377,
+      "learning_rate": 1.9850561863409763e-05,
+      "loss": 0.245,
+      "step": 1641
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2675161459462643,
+      "learning_rate": 1.985027806697053e-05,
+      "loss": 0.2235,
+      "step": 1642
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4699273219751203,
+      "learning_rate": 1.984999400334195e-05,
+      "loss": 0.2787,
+      "step": 1643
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3416851354072679,
+      "learning_rate": 1.9849709672531724e-05,
+      "loss": 0.2144,
+      "step": 1644
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.588177614535934,
+      "learning_rate": 1.9849425074547565e-05,
+      "loss": 0.2643,
+      "step": 1645
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4621532798948187,
+      "learning_rate": 1.98491402093972e-05,
+      "loss": 0.2434,
+      "step": 1646
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.016718930677667,
+      "learning_rate": 1.9848855077088346e-05,
+      "loss": 0.245,
+      "step": 1647
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.960429641737833,
+      "learning_rate": 1.984856967762874e-05,
+      "loss": 0.2353,
+      "step": 1648
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0038547952300165,
+      "learning_rate": 1.984828401102613e-05,
+      "loss": 0.2471,
+      "step": 1649
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5716038810095376,
+      "learning_rate": 1.9847998077288255e-05,
+      "loss": 0.2624,
+      "step": 1650
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0463153168335042,
+      "learning_rate": 1.984771187642288e-05,
+      "loss": 0.2196,
+      "step": 1651
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1341781324491753,
+      "learning_rate": 1.9847425408437763e-05,
+      "loss": 0.2548,
+      "step": 1652
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3841618420365462,
+      "learning_rate": 1.9847138673340675e-05,
+      "loss": 0.2438,
+      "step": 1653
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0169550684413193,
+      "learning_rate": 1.9846851671139394e-05,
+      "loss": 0.2607,
+      "step": 1654
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3519043264751935,
+      "learning_rate": 1.984656440184171e-05,
+      "loss": 0.2607,
+      "step": 1655
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2608149031054832,
+      "learning_rate": 1.9846276865455403e-05,
+      "loss": 0.2437,
+      "step": 1656
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0108753451264925,
+      "learning_rate": 1.9845989061988283e-05,
+      "loss": 0.2487,
+      "step": 1657
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9027978818842535,
+      "learning_rate": 1.9845700991448154e-05,
+      "loss": 0.2359,
+      "step": 1658
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.250101052431056,
+      "learning_rate": 1.984541265384283e-05,
+      "loss": 0.2632,
+      "step": 1659
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3325062156063474,
+      "learning_rate": 1.9845124049180132e-05,
+      "loss": 0.2225,
+      "step": 1660
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9988706987608464,
+      "learning_rate": 1.9844835177467886e-05,
+      "loss": 0.2321,
+      "step": 1661
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.120658383739874,
+      "learning_rate": 1.984454603871393e-05,
+      "loss": 0.2576,
+      "step": 1662
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9601587400055499,
+      "learning_rate": 1.984425663292611e-05,
+      "loss": 0.2143,
+      "step": 1663
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9261788091835301,
+      "learning_rate": 1.984396696011227e-05,
+      "loss": 0.2469,
+      "step": 1664
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5508358929347974,
+      "learning_rate": 1.984367702028027e-05,
+      "loss": 0.2642,
+      "step": 1665
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9767778816683551,
+      "learning_rate": 1.9843386813437976e-05,
+      "loss": 0.2367,
+      "step": 1666
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0196991564528861,
+      "learning_rate": 1.984309633959326e-05,
+      "loss": 0.2413,
+      "step": 1667
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.316436967597623,
+      "learning_rate": 1.9842805598753997e-05,
+      "loss": 0.2642,
+      "step": 1668
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2167182228798439,
+      "learning_rate": 1.984251459092808e-05,
+      "loss": 0.2275,
+      "step": 1669
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1146683861566014,
+      "learning_rate": 1.9842223316123393e-05,
+      "loss": 0.2346,
+      "step": 1670
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3334225033325622,
+      "learning_rate": 1.9841931774347846e-05,
+      "loss": 0.2545,
+      "step": 1671
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3786897506461797,
+      "learning_rate": 1.9841639965609344e-05,
+      "loss": 0.2298,
+      "step": 1672
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.612244430590351,
+      "learning_rate": 1.9841347889915804e-05,
+      "loss": 0.2449,
+      "step": 1673
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2949843857838457,
+      "learning_rate": 1.9841055547275142e-05,
+      "loss": 0.2405,
+      "step": 1674
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.229655683707706,
+      "learning_rate": 1.9840762937695296e-05,
+      "loss": 0.2706,
+      "step": 1675
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5452909783071291,
+      "learning_rate": 1.98404700611842e-05,
+      "loss": 0.2369,
+      "step": 1676
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9981909241193303,
+      "learning_rate": 1.9840176917749795e-05,
+      "loss": 0.2494,
+      "step": 1677
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0870733215780806,
+      "learning_rate": 1.9839883507400037e-05,
+      "loss": 0.2528,
+      "step": 1678
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0528706352105188,
+      "learning_rate": 1.9839589830142882e-05,
+      "loss": 0.257,
+      "step": 1679
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1201032075648276,
+      "learning_rate": 1.98392958859863e-05,
+      "loss": 0.2492,
+      "step": 1680
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0474478400549914,
+      "learning_rate": 1.983900167493826e-05,
+      "loss": 0.2926,
+      "step": 1681
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0154346670627037,
+      "learning_rate": 1.983870719700674e-05,
+      "loss": 0.2386,
+      "step": 1682
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8657370203828151,
+      "learning_rate": 1.9838412452199732e-05,
+      "loss": 0.242,
+      "step": 1683
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.918886234549711,
+      "learning_rate": 1.9838117440525236e-05,
+      "loss": 0.2439,
+      "step": 1684
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9403024089951526,
+      "learning_rate": 1.9837822161991248e-05,
+      "loss": 0.2429,
+      "step": 1685
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.084076095905887,
+      "learning_rate": 1.9837526616605774e-05,
+      "loss": 0.2335,
+      "step": 1686
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7367729442066835,
+      "learning_rate": 1.9837230804376838e-05,
+      "loss": 0.2293,
+      "step": 1687
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9329776639522541,
+      "learning_rate": 1.983693472531246e-05,
+      "loss": 0.2169,
+      "step": 1688
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2153526164638409,
+      "learning_rate": 1.983663837942067e-05,
+      "loss": 0.2542,
+      "step": 1689
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.956868994422886,
+      "learning_rate": 1.983634176670951e-05,
+      "loss": 0.2591,
+      "step": 1690
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.029881962928075,
+      "learning_rate": 1.9836044887187023e-05,
+      "loss": 0.2213,
+      "step": 1691
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1669934190529483,
+      "learning_rate": 1.9835747740861266e-05,
+      "loss": 0.2626,
+      "step": 1692
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9980708214268654,
+      "learning_rate": 1.9835450327740293e-05,
+      "loss": 0.2788,
+      "step": 1693
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.158179447044645,
+      "learning_rate": 1.9835152647832175e-05,
+      "loss": 0.2486,
+      "step": 1694
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9135665839164294,
+      "learning_rate": 1.9834854701144986e-05,
+      "loss": 0.2286,
+      "step": 1695
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0453395818232172,
+      "learning_rate": 1.983455648768681e-05,
+      "loss": 0.2274,
+      "step": 1696
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0728487666452273,
+      "learning_rate": 1.983425800746573e-05,
+      "loss": 0.2582,
+      "step": 1697
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9397892154114075,
+      "learning_rate": 1.983395926048985e-05,
+      "loss": 0.2259,
+      "step": 1698
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9379066761125769,
+      "learning_rate": 1.9833660246767267e-05,
+      "loss": 0.2386,
+      "step": 1699
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2392563662188694,
+      "learning_rate": 1.9833360966306095e-05,
+      "loss": 0.237,
+      "step": 1700
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2609688920322724,
+      "learning_rate": 1.9833061419114452e-05,
+      "loss": 0.2885,
+      "step": 1701
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1060909223384836,
+      "learning_rate": 1.9832761605200464e-05,
+      "loss": 0.2539,
+      "step": 1702
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0220662279210742,
+      "learning_rate": 1.9832461524572258e-05,
+      "loss": 0.2272,
+      "step": 1703
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.071997778470365,
+      "learning_rate": 1.983216117723798e-05,
+      "loss": 0.2574,
+      "step": 1704
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0151680433606107,
+      "learning_rate": 1.9831860563205776e-05,
+      "loss": 0.2401,
+      "step": 1705
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.987979664831528,
+      "learning_rate": 1.98315596824838e-05,
+      "loss": 0.2201,
+      "step": 1706
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8212769811622768,
+      "learning_rate": 1.9831258535080206e-05,
+      "loss": 0.2338,
+      "step": 1707
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1635767738327358,
+      "learning_rate": 1.9830957121003176e-05,
+      "loss": 0.2394,
+      "step": 1708
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0173061683887226,
+      "learning_rate": 1.983065544026087e-05,
+      "loss": 0.264,
+      "step": 1709
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.833164217466478,
+      "learning_rate": 1.9830353492861493e-05,
+      "loss": 0.2578,
+      "step": 1710
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.022547453107026,
+      "learning_rate": 1.983005127881321e-05,
+      "loss": 0.2651,
+      "step": 1711
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0363103004554934,
+      "learning_rate": 1.9829748798124237e-05,
+      "loss": 0.2612,
+      "step": 1712
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9714966214210872,
+      "learning_rate": 1.9829446050802776e-05,
+      "loss": 0.2519,
+      "step": 1713
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9434534750808347,
+      "learning_rate": 1.9829143036857027e-05,
+      "loss": 0.2293,
+      "step": 1714
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.195774260854126,
+      "learning_rate": 1.9828839756295223e-05,
+      "loss": 0.2517,
+      "step": 1715
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0262577604343457,
+      "learning_rate": 1.9828536209125584e-05,
+      "loss": 0.2463,
+      "step": 1716
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.842309395951756,
+      "learning_rate": 1.9828232395356347e-05,
+      "loss": 0.2409,
+      "step": 1717
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2461010415205944,
+      "learning_rate": 1.982792831499575e-05,
+      "loss": 0.2425,
+      "step": 1718
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9829452005423613,
+      "learning_rate": 1.982762396805204e-05,
+      "loss": 0.2484,
+      "step": 1719
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.984884649513512,
+      "learning_rate": 1.982731935453348e-05,
+      "loss": 0.2389,
+      "step": 1720
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1119583351575009,
+      "learning_rate": 1.9827014474448324e-05,
+      "loss": 0.2518,
+      "step": 1721
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.002796824835171,
+      "learning_rate": 1.9826709327804846e-05,
+      "loss": 0.2284,
+      "step": 1722
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0157183698966548,
+      "learning_rate": 1.982640391461132e-05,
+      "loss": 0.2415,
+      "step": 1723
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.260041652519267,
+      "learning_rate": 1.982609823487604e-05,
+      "loss": 0.2351,
+      "step": 1724
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3519455184908302,
+      "learning_rate": 1.9825792288607284e-05,
+      "loss": 0.2451,
+      "step": 1725
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0062626123603624,
+      "learning_rate": 1.982548607581336e-05,
+      "loss": 0.2467,
+      "step": 1726
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2439664837325628,
+      "learning_rate": 1.9825179596502567e-05,
+      "loss": 0.2509,
+      "step": 1727
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8243535374289274,
+      "learning_rate": 1.9824872850683226e-05,
+      "loss": 0.2075,
+      "step": 1728
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1505514359912987,
+      "learning_rate": 1.9824565838363657e-05,
+      "loss": 0.2419,
+      "step": 1729
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.269435382278997,
+      "learning_rate": 1.9824258559552182e-05,
+      "loss": 0.239,
+      "step": 1730
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.31208623122475,
+      "learning_rate": 1.9823951014257138e-05,
+      "loss": 0.2527,
+      "step": 1731
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.206820363011641,
+      "learning_rate": 1.9823643202486867e-05,
+      "loss": 0.2221,
+      "step": 1732
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2041683786551478,
+      "learning_rate": 1.982333512424972e-05,
+      "loss": 0.2213,
+      "step": 1733
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0255291639405049,
+      "learning_rate": 1.9823026779554055e-05,
+      "loss": 0.2405,
+      "step": 1734
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1134042500477659,
+      "learning_rate": 1.982271816840823e-05,
+      "loss": 0.2355,
+      "step": 1735
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.122621923354994,
+      "learning_rate": 1.982240929082062e-05,
+      "loss": 0.2591,
+      "step": 1736
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0109011146022282,
+      "learning_rate": 1.9822100146799607e-05,
+      "loss": 0.2268,
+      "step": 1737
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1939519658509656,
+      "learning_rate": 1.982179073635357e-05,
+      "loss": 0.2648,
+      "step": 1738
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5178683098536014,
+      "learning_rate": 1.9821481059490906e-05,
+      "loss": 0.2232,
+      "step": 1739
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9516149012483963,
+      "learning_rate": 1.982117111622001e-05,
+      "loss": 0.2476,
+      "step": 1740
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2321003493682299,
+      "learning_rate": 1.98208609065493e-05,
+      "loss": 0.2064,
+      "step": 1741
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1263545813097904,
+      "learning_rate": 1.982055043048718e-05,
+      "loss": 0.219,
+      "step": 1742
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9164576815422517,
+      "learning_rate": 1.982023968804207e-05,
+      "loss": 0.2325,
+      "step": 1743
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.481776837320407,
+      "learning_rate": 1.981992867922241e-05,
+      "loss": 0.2732,
+      "step": 1744
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0480431065887095,
+      "learning_rate": 1.981961740403663e-05,
+      "loss": 0.225,
+      "step": 1745
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1453713530810714,
+      "learning_rate": 1.981930586249317e-05,
+      "loss": 0.2555,
+      "step": 1746
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0440512071308312,
+      "learning_rate": 1.9818994054600484e-05,
+      "loss": 0.2443,
+      "step": 1747
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3140536888347236,
+      "learning_rate": 1.981868198036703e-05,
+      "loss": 0.2296,
+      "step": 1748
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9589913304538162,
+      "learning_rate": 1.9818369639801273e-05,
+      "loss": 0.2269,
+      "step": 1749
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9421444481840565,
+      "learning_rate": 1.9818057032911687e-05,
+      "loss": 0.236,
+      "step": 1750
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.011810083369825,
+      "learning_rate": 1.9817744159706746e-05,
+      "loss": 0.2542,
+      "step": 1751
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9867132876670602,
+      "learning_rate": 1.9817431020194942e-05,
+      "loss": 0.2248,
+      "step": 1752
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2165872518452028,
+      "learning_rate": 1.9817117614384767e-05,
+      "loss": 0.2315,
+      "step": 1753
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6573462337949036,
+      "learning_rate": 1.9816803942284724e-05,
+      "loss": 0.2482,
+      "step": 1754
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0905321049950607,
+      "learning_rate": 1.9816490003903316e-05,
+      "loss": 0.2451,
+      "step": 1755
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2974795740116438,
+      "learning_rate": 1.9816175799249064e-05,
+      "loss": 0.262,
+      "step": 1756
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9930878435396299,
+      "learning_rate": 1.981586132833049e-05,
+      "loss": 0.2834,
+      "step": 1757
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9804621023472282,
+      "learning_rate": 1.981554659115612e-05,
+      "loss": 0.249,
+      "step": 1758
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8228160230043197,
+      "learning_rate": 1.98152315877345e-05,
+      "loss": 0.2267,
+      "step": 1759
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8087224046227851,
+      "learning_rate": 1.9814916318074163e-05,
+      "loss": 0.2372,
+      "step": 1760
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9374525942013675,
+      "learning_rate": 1.981460078218367e-05,
+      "loss": 0.2595,
+      "step": 1761
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9126980800579981,
+      "learning_rate": 1.9814284980071578e-05,
+      "loss": 0.2429,
+      "step": 1762
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7815278328067415,
+      "learning_rate": 1.9813968911746447e-05,
+      "loss": 0.2128,
+      "step": 1763
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1630201006053036,
+      "learning_rate": 1.9813652577216858e-05,
+      "loss": 0.2348,
+      "step": 1764
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0067825312667709,
+      "learning_rate": 1.9813335976491387e-05,
+      "loss": 0.2426,
+      "step": 1765
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7904482493750873,
+      "learning_rate": 1.9813019109578623e-05,
+      "loss": 0.2409,
+      "step": 1766
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0847359740312001,
+      "learning_rate": 1.981270197648716e-05,
+      "loss": 0.2616,
+      "step": 1767
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.935404199004101,
+      "learning_rate": 1.9812384577225604e-05,
+      "loss": 0.2277,
+      "step": 1768
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.289585307813352,
+      "learning_rate": 1.9812066911802565e-05,
+      "loss": 0.2553,
+      "step": 1769
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4646201642946093,
+      "learning_rate": 1.981174898022665e-05,
+      "loss": 0.2478,
+      "step": 1770
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3498844676297264,
+      "learning_rate": 1.9811430782506496e-05,
+      "loss": 0.2371,
+      "step": 1771
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.195156883689694,
+      "learning_rate": 1.9811112318650725e-05,
+      "loss": 0.2526,
+      "step": 1772
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7032110116525865,
+      "learning_rate": 1.9810793588667977e-05,
+      "loss": 0.258,
+      "step": 1773
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.886767523375247,
+      "learning_rate": 1.98104745925669e-05,
+      "loss": 0.2432,
+      "step": 1774
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0485652010268987,
+      "learning_rate": 1.9810155330356147e-05,
+      "loss": 0.2278,
+      "step": 1775
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0753335063192004,
+      "learning_rate": 1.9809835802044378e-05,
+      "loss": 0.2317,
+      "step": 1776
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2770843961058662,
+      "learning_rate": 1.9809516007640255e-05,
+      "loss": 0.229,
+      "step": 1777
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8403924810021453,
+      "learning_rate": 1.9809195947152458e-05,
+      "loss": 0.2282,
+      "step": 1778
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0793162103481402,
+      "learning_rate": 1.9808875620589667e-05,
+      "loss": 0.2234,
+      "step": 1779
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1238385198046674,
+      "learning_rate": 1.980855502796057e-05,
+      "loss": 0.2174,
+      "step": 1780
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2699355390187514,
+      "learning_rate": 1.9808234169273864e-05,
+      "loss": 0.2288,
+      "step": 1781
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.011158926633347,
+      "learning_rate": 1.9807913044538252e-05,
+      "loss": 0.2485,
+      "step": 1782
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4074623352058435,
+      "learning_rate": 1.9807591653762447e-05,
+      "loss": 0.2307,
+      "step": 1783
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9928912356239473,
+      "learning_rate": 1.980726999695516e-05,
+      "loss": 0.2675,
+      "step": 1784
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0184293111623297,
+      "learning_rate": 1.9806948074125123e-05,
+      "loss": 0.2908,
+      "step": 1785
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1122490801754585,
+      "learning_rate": 1.9806625885281065e-05,
+      "loss": 0.2621,
+      "step": 1786
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1598479098892676,
+      "learning_rate": 1.9806303430431727e-05,
+      "loss": 0.2441,
+      "step": 1787
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.979354192254303,
+      "learning_rate": 1.9805980709585855e-05,
+      "loss": 0.2375,
+      "step": 1788
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.267730797707709,
+      "learning_rate": 1.9805657722752202e-05,
+      "loss": 0.2284,
+      "step": 1789
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9952643727050997,
+      "learning_rate": 1.980533446993953e-05,
+      "loss": 0.2453,
+      "step": 1790
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0233808075949027,
+      "learning_rate": 1.9805010951156605e-05,
+      "loss": 0.2396,
+      "step": 1791
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.024484392071274,
+      "learning_rate": 1.9804687166412204e-05,
+      "loss": 0.2423,
+      "step": 1792
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1779536450407442,
+      "learning_rate": 1.980436311571511e-05,
+      "loss": 0.2506,
+      "step": 1793
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.110809751968605,
+      "learning_rate": 1.9804038799074114e-05,
+      "loss": 0.2379,
+      "step": 1794
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.911356847281592,
+      "learning_rate": 1.9803714216498013e-05,
+      "loss": 0.2459,
+      "step": 1795
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2508497514282289,
+      "learning_rate": 1.9803389367995606e-05,
+      "loss": 0.2559,
+      "step": 1796
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2547143824197837,
+      "learning_rate": 1.9803064253575713e-05,
+      "loss": 0.2678,
+      "step": 1797
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1171374299241046,
+      "learning_rate": 1.9802738873247146e-05,
+      "loss": 0.2373,
+      "step": 1798
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3682294514056483,
+      "learning_rate": 1.9802413227018732e-05,
+      "loss": 0.2634,
+      "step": 1799
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6593452822995967,
+      "learning_rate": 1.980208731489931e-05,
+      "loss": 0.2565,
+      "step": 1800
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.33397124612498,
+      "learning_rate": 1.9801761136897713e-05,
+      "loss": 0.2465,
+      "step": 1801
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2179198485821112,
+      "learning_rate": 1.980143469302279e-05,
+      "loss": 0.251,
+      "step": 1802
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.248443571251299,
+      "learning_rate": 1.9801107983283403e-05,
+      "loss": 0.2264,
+      "step": 1803
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2624127344026586,
+      "learning_rate": 1.9800781007688403e-05,
+      "loss": 0.2419,
+      "step": 1804
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0773293999784925,
+      "learning_rate": 1.9800453766246668e-05,
+      "loss": 0.2478,
+      "step": 1805
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2539597583340516,
+      "learning_rate": 1.980012625896707e-05,
+      "loss": 0.242,
+      "step": 1806
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0523566435986238,
+      "learning_rate": 1.979979848585849e-05,
+      "loss": 0.2438,
+      "step": 1807
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2783462412732363,
+      "learning_rate": 1.9799470446929827e-05,
+      "loss": 0.2322,
+      "step": 1808
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.501195818234163,
+      "learning_rate": 1.9799142142189974e-05,
+      "loss": 0.2366,
+      "step": 1809
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0351472253286227,
+      "learning_rate": 1.9798813571647835e-05,
+      "loss": 0.2475,
+      "step": 1810
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.141466973878634,
+      "learning_rate": 1.9798484735312327e-05,
+      "loss": 0.224,
+      "step": 1811
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3377399022823369,
+      "learning_rate": 1.9798155633192368e-05,
+      "loss": 0.2341,
+      "step": 1812
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4789245607733694,
+      "learning_rate": 1.979782626529688e-05,
+      "loss": 0.2666,
+      "step": 1813
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7178934408624136,
+      "learning_rate": 1.9797496631634804e-05,
+      "loss": 0.2506,
+      "step": 1814
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9579183307072763,
+      "learning_rate": 1.9797166732215078e-05,
+      "loss": 0.2334,
+      "step": 1815
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1702378825834747,
+      "learning_rate": 1.979683656704665e-05,
+      "loss": 0.2367,
+      "step": 1816
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2623264964220406,
+      "learning_rate": 1.979650613613848e-05,
+      "loss": 0.2264,
+      "step": 1817
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1794467571917981,
+      "learning_rate": 1.979617543949952e-05,
+      "loss": 0.2339,
+      "step": 1818
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.91157451744252,
+      "learning_rate": 1.9795844477138756e-05,
+      "loss": 0.225,
+      "step": 1819
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2309863507665357,
+      "learning_rate": 1.9795513249065155e-05,
+      "loss": 0.253,
+      "step": 1820
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1783442633169707,
+      "learning_rate": 1.97951817552877e-05,
+      "loss": 0.2379,
+      "step": 1821
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.094279223019567,
+      "learning_rate": 1.9794849995815392e-05,
+      "loss": 0.2457,
+      "step": 1822
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3138667208821027,
+      "learning_rate": 1.979451797065722e-05,
+      "loss": 0.259,
+      "step": 1823
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2491422875129075,
+      "learning_rate": 1.97941856798222e-05,
+      "loss": 0.2364,
+      "step": 1824
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.186450281638106,
+      "learning_rate": 1.979385312331934e-05,
+      "loss": 0.2312,
+      "step": 1825
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1194218673747123,
+      "learning_rate": 1.9793520301157656e-05,
+      "loss": 0.2473,
+      "step": 1826
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5625455110596913,
+      "learning_rate": 1.9793187213346183e-05,
+      "loss": 0.2544,
+      "step": 1827
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1394717462108623,
+      "learning_rate": 1.9792853859893953e-05,
+      "loss": 0.2318,
+      "step": 1828
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.979714128320049,
+      "learning_rate": 1.9792520240810012e-05,
+      "loss": 0.2555,
+      "step": 1829
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0604405691060503,
+      "learning_rate": 1.9792186356103403e-05,
+      "loss": 0.2421,
+      "step": 1830
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9962422834651873,
+      "learning_rate": 1.9791852205783186e-05,
+      "loss": 0.2274,
+      "step": 1831
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0115858456203548,
+      "learning_rate": 1.9791517789858428e-05,
+      "loss": 0.218,
+      "step": 1832
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1584982837673712,
+      "learning_rate": 1.9791183108338195e-05,
+      "loss": 0.2291,
+      "step": 1833
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4205913603650355,
+      "learning_rate": 1.9790848161231568e-05,
+      "loss": 0.2258,
+      "step": 1834
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9593511399242401,
+      "learning_rate": 1.9790512948547633e-05,
+      "loss": 0.2421,
+      "step": 1835
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.205982035714139,
+      "learning_rate": 1.9790177470295474e-05,
+      "loss": 0.2452,
+      "step": 1836
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0021327684946961,
+      "learning_rate": 1.9789841726484208e-05,
+      "loss": 0.2571,
+      "step": 1837
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1537386848613858,
+      "learning_rate": 1.9789505717122926e-05,
+      "loss": 0.2639,
+      "step": 1838
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.251241669990851,
+      "learning_rate": 1.978916944222075e-05,
+      "loss": 0.2597,
+      "step": 1839
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3627173950389766,
+      "learning_rate": 1.97888329017868e-05,
+      "loss": 0.2549,
+      "step": 1840
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1280632425903117,
+      "learning_rate": 1.9788496095830205e-05,
+      "loss": 0.2427,
+      "step": 1841
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2424096456708325,
+      "learning_rate": 1.97881590243601e-05,
+      "loss": 0.2507,
+      "step": 1842
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1268147690233024,
+      "learning_rate": 1.978782168738563e-05,
+      "loss": 0.2313,
+      "step": 1843
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9731217683506375,
+      "learning_rate": 1.9787484084915943e-05,
+      "loss": 0.1961,
+      "step": 1844
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9932726226384986,
+      "learning_rate": 1.9787146216960196e-05,
+      "loss": 0.2704,
+      "step": 1845
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3258885680331116,
+      "learning_rate": 1.978680808352756e-05,
+      "loss": 0.2223,
+      "step": 1846
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.179125947707762,
+      "learning_rate": 1.9786469684627193e-05,
+      "loss": 0.2233,
+      "step": 1847
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.181157569743411,
+      "learning_rate": 1.978613102026829e-05,
+      "loss": 0.2614,
+      "step": 1848
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2465357791351852,
+      "learning_rate": 1.9785792090460026e-05,
+      "loss": 0.1947,
+      "step": 1849
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1762006333045996,
+      "learning_rate": 1.9785452895211606e-05,
+      "loss": 0.2488,
+      "step": 1850
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9732313500927374,
+      "learning_rate": 1.978511343453222e-05,
+      "loss": 0.2208,
+      "step": 1851
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.930795569689893,
+      "learning_rate": 1.9784773708431076e-05,
+      "loss": 0.2319,
+      "step": 1852
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1402431579192085,
+      "learning_rate": 1.9784433716917397e-05,
+      "loss": 0.2394,
+      "step": 1853
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0338244690058258,
+      "learning_rate": 1.97840934600004e-05,
+      "loss": 0.2477,
+      "step": 1854
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6535929415927046,
+      "learning_rate": 1.9783752937689312e-05,
+      "loss": 0.2408,
+      "step": 1855
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.370824985835338,
+      "learning_rate": 1.9783412149993374e-05,
+      "loss": 0.2349,
+      "step": 1856
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3956801985306262,
+      "learning_rate": 1.978307109692183e-05,
+      "loss": 0.2534,
+      "step": 1857
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3940946373038978,
+      "learning_rate": 1.978272977848393e-05,
+      "loss": 0.286,
+      "step": 1858
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0987035112902765,
+      "learning_rate": 1.9782388194688933e-05,
+      "loss": 0.2779,
+      "step": 1859
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1882153135309836,
+      "learning_rate": 1.9782046345546102e-05,
+      "loss": 0.2235,
+      "step": 1860
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2008243551349194,
+      "learning_rate": 1.9781704231064715e-05,
+      "loss": 0.2693,
+      "step": 1861
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1751556702307224,
+      "learning_rate": 1.9781361851254044e-05,
+      "loss": 0.2543,
+      "step": 1862
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1081469900399938,
+      "learning_rate": 1.9781019206123382e-05,
+      "loss": 0.2512,
+      "step": 1863
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9908211347171663,
+      "learning_rate": 1.978067629568202e-05,
+      "loss": 0.2035,
+      "step": 1864
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2227189853228269,
+      "learning_rate": 1.9780333119939264e-05,
+      "loss": 0.235,
+      "step": 1865
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2011150507746902,
+      "learning_rate": 1.9779989678904416e-05,
+      "loss": 0.2399,
+      "step": 1866
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3756169840284638,
+      "learning_rate": 1.97796459725868e-05,
+      "loss": 0.2533,
+      "step": 1867
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7413831052381767,
+      "learning_rate": 1.9779302000995732e-05,
+      "loss": 0.2174,
+      "step": 1868
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2214321925490699,
+      "learning_rate": 1.9778957764140545e-05,
+      "loss": 0.2509,
+      "step": 1869
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.048946823233343,
+      "learning_rate": 1.9778613262030577e-05,
+      "loss": 0.2396,
+      "step": 1870
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0703035941982966,
+      "learning_rate": 1.9778268494675172e-05,
+      "loss": 0.2352,
+      "step": 1871
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.010644102322746,
+      "learning_rate": 1.977792346208368e-05,
+      "loss": 0.2269,
+      "step": 1872
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2422707614304045,
+      "learning_rate": 1.9777578164265464e-05,
+      "loss": 0.2476,
+      "step": 1873
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2907222314859268,
+      "learning_rate": 1.9777232601229887e-05,
+      "loss": 0.2204,
+      "step": 1874
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3521410331749686,
+      "learning_rate": 1.9776886772986325e-05,
+      "loss": 0.2366,
+      "step": 1875
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0863288578749053,
+      "learning_rate": 1.9776540679544154e-05,
+      "loss": 0.2297,
+      "step": 1876
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0241804063042204,
+      "learning_rate": 1.977619432091277e-05,
+      "loss": 0.2277,
+      "step": 1877
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.359207853748398,
+      "learning_rate": 1.977584769710156e-05,
+      "loss": 0.2602,
+      "step": 1878
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.488727423100929,
+      "learning_rate": 1.9775500808119926e-05,
+      "loss": 0.2123,
+      "step": 1879
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9310924451888012,
+      "learning_rate": 1.9775153653977284e-05,
+      "loss": 0.2347,
+      "step": 1880
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0882134751609247,
+      "learning_rate": 1.9774806234683047e-05,
+      "loss": 0.2168,
+      "step": 1881
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1032916586088626,
+      "learning_rate": 1.9774458550246636e-05,
+      "loss": 0.2346,
+      "step": 1882
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8964220837477248,
+      "learning_rate": 1.977411060067749e-05,
+      "loss": 0.2192,
+      "step": 1883
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2271518408134103,
+      "learning_rate": 1.977376238598504e-05,
+      "loss": 0.2207,
+      "step": 1884
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2657120013958592,
+      "learning_rate": 1.977341390617873e-05,
+      "loss": 0.2216,
+      "step": 1885
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1240450447057986,
+      "learning_rate": 1.9773065161268015e-05,
+      "loss": 0.2152,
+      "step": 1886
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.140021817870832,
+      "learning_rate": 1.977271615126236e-05,
+      "loss": 0.2405,
+      "step": 1887
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9646328044306108,
+      "learning_rate": 1.9772366876171224e-05,
+      "loss": 0.2214,
+      "step": 1888
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0229657136088213,
+      "learning_rate": 1.9772017336004085e-05,
+      "loss": 0.2486,
+      "step": 1889
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0801181204271537,
+      "learning_rate": 1.9771667530770427e-05,
+      "loss": 0.2413,
+      "step": 1890
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0024092708899657,
+      "learning_rate": 1.9771317460479733e-05,
+      "loss": 0.2554,
+      "step": 1891
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9461916604616549,
+      "learning_rate": 1.9770967125141502e-05,
+      "loss": 0.2465,
+      "step": 1892
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4241403509851787,
+      "learning_rate": 1.9770616524765236e-05,
+      "loss": 0.2618,
+      "step": 1893
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2192273387499462,
+      "learning_rate": 1.9770265659360445e-05,
+      "loss": 0.2477,
+      "step": 1894
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9149517263311271,
+      "learning_rate": 1.9769914528936646e-05,
+      "loss": 0.2203,
+      "step": 1895
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9640595665851582,
+      "learning_rate": 1.976956313350336e-05,
+      "loss": 0.255,
+      "step": 1896
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2296809453079092,
+      "learning_rate": 1.9769211473070124e-05,
+      "loss": 0.2257,
+      "step": 1897
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0503214717220533,
+      "learning_rate": 1.9768859547646476e-05,
+      "loss": 0.245,
+      "step": 1898
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0261216658899286,
+      "learning_rate": 1.976850735724196e-05,
+      "loss": 0.2228,
+      "step": 1899
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8879896992019953,
+      "learning_rate": 1.9768154901866136e-05,
+      "loss": 0.2362,
+      "step": 1900
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.944882194134527,
+      "learning_rate": 1.9767802181528552e-05,
+      "loss": 0.2428,
+      "step": 1901
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4309989587560117,
+      "learning_rate": 1.9767449196238785e-05,
+      "loss": 0.2513,
+      "step": 1902
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1054306441049362,
+      "learning_rate": 1.9767095946006405e-05,
+      "loss": 0.2416,
+      "step": 1903
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9790156761267768,
+      "learning_rate": 1.9766742430840998e-05,
+      "loss": 0.2196,
+      "step": 1904
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.225542879755533,
+      "learning_rate": 1.9766388650752152e-05,
+      "loss": 0.2531,
+      "step": 1905
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1451682777421768,
+      "learning_rate": 1.976603460574946e-05,
+      "loss": 0.2252,
+      "step": 1906
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1697641021216751,
+      "learning_rate": 1.9765680295842525e-05,
+      "loss": 0.2374,
+      "step": 1907
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3119766391360652,
+      "learning_rate": 1.9765325721040964e-05,
+      "loss": 0.2328,
+      "step": 1908
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8776972094573054,
+      "learning_rate": 1.976497088135439e-05,
+      "loss": 0.2255,
+      "step": 1909
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.901221087412436,
+      "learning_rate": 1.976461577679243e-05,
+      "loss": 0.2538,
+      "step": 1910
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9364330186554849,
+      "learning_rate": 1.9764260407364714e-05,
+      "loss": 0.2266,
+      "step": 1911
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8343122185679239,
+      "learning_rate": 1.9763904773080886e-05,
+      "loss": 0.241,
+      "step": 1912
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5044879595592187,
+      "learning_rate": 1.9763548873950586e-05,
+      "loss": 0.2721,
+      "step": 1913
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0495765131484909,
+      "learning_rate": 1.9763192709983473e-05,
+      "loss": 0.2417,
+      "step": 1914
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.05508985163837,
+      "learning_rate": 1.9762836281189207e-05,
+      "loss": 0.2227,
+      "step": 1915
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0155022843798927,
+      "learning_rate": 1.9762479587577457e-05,
+      "loss": 0.2425,
+      "step": 1916
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0967121155196264,
+      "learning_rate": 1.976212262915789e-05,
+      "loss": 0.2335,
+      "step": 1917
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.37545605462818,
+      "learning_rate": 1.9761765405940203e-05,
+      "loss": 0.2407,
+      "step": 1918
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9818909497523616,
+      "learning_rate": 1.9761407917934073e-05,
+      "loss": 0.2493,
+      "step": 1919
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.137871337305637,
+      "learning_rate": 1.9761050165149208e-05,
+      "loss": 0.246,
+      "step": 1920
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9860186561324186,
+      "learning_rate": 1.9760692147595298e-05,
+      "loss": 0.243,
+      "step": 1921
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0181895969287278,
+      "learning_rate": 1.9760333865282067e-05,
+      "loss": 0.2473,
+      "step": 1922
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8406082864526969,
+      "learning_rate": 1.975997531821923e-05,
+      "loss": 0.2075,
+      "step": 1923
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9227285386867792,
+      "learning_rate": 1.9759616506416506e-05,
+      "loss": 0.2367,
+      "step": 1924
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5616621406325664,
+      "learning_rate": 1.975925742988364e-05,
+      "loss": 0.2383,
+      "step": 1925
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3935247996994191,
+      "learning_rate": 1.975889808863036e-05,
+      "loss": 0.2452,
+      "step": 1926
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2943208678791482,
+      "learning_rate": 1.975853848266642e-05,
+      "loss": 0.2407,
+      "step": 1927
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.450294411426727,
+      "learning_rate": 1.975817861200157e-05,
+      "loss": 0.2277,
+      "step": 1928
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9018073347670972,
+      "learning_rate": 1.9757818476645573e-05,
+      "loss": 0.2498,
+      "step": 1929
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0314892631011348,
+      "learning_rate": 1.9757458076608204e-05,
+      "loss": 0.2412,
+      "step": 1930
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1697931974945825,
+      "learning_rate": 1.975709741189923e-05,
+      "loss": 0.2312,
+      "step": 1931
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9365982325740557,
+      "learning_rate": 1.975673648252844e-05,
+      "loss": 0.2147,
+      "step": 1932
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8698438621254598,
+      "learning_rate": 1.975637528850562e-05,
+      "loss": 0.2396,
+      "step": 1933
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1560916597632185,
+      "learning_rate": 1.9756013829840568e-05,
+      "loss": 0.2024,
+      "step": 1934
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8347486142298968,
+      "learning_rate": 1.9755652106543094e-05,
+      "loss": 0.2421,
+      "step": 1935
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.920153099028764,
+      "learning_rate": 1.9755290118623e-05,
+      "loss": 0.2228,
+      "step": 1936
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9913299797122175,
+      "learning_rate": 1.9754927866090115e-05,
+      "loss": 0.251,
+      "step": 1937
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0404221817815027,
+      "learning_rate": 1.975456534895426e-05,
+      "loss": 0.2099,
+      "step": 1938
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.115116727740608,
+      "learning_rate": 1.975420256722527e-05,
+      "loss": 0.248,
+      "step": 1939
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1369950824968853,
+      "learning_rate": 1.9753839520912984e-05,
+      "loss": 0.2212,
+      "step": 1940
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4411550000757647,
+      "learning_rate": 1.9753476210027248e-05,
+      "loss": 0.204,
+      "step": 1941
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.877079500076538,
+      "learning_rate": 1.975311263457792e-05,
+      "loss": 0.236,
+      "step": 1942
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2958517471483217,
+      "learning_rate": 1.9752748794574858e-05,
+      "loss": 0.2547,
+      "step": 1943
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.102209497560389,
+      "learning_rate": 1.9752384690027937e-05,
+      "loss": 0.2082,
+      "step": 1944
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0121374445198599,
+      "learning_rate": 1.975202032094703e-05,
+      "loss": 0.2849,
+      "step": 1945
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8864239305753996,
+      "learning_rate": 1.9751655687342022e-05,
+      "loss": 0.2282,
+      "step": 1946
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9967365811704525,
+      "learning_rate": 1.9751290789222804e-05,
+      "loss": 0.2322,
+      "step": 1947
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1234867484412785,
+      "learning_rate": 1.975092562659927e-05,
+      "loss": 0.2789,
+      "step": 1948
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8313426797342862,
+      "learning_rate": 1.9750560199481325e-05,
+      "loss": 0.218,
+      "step": 1949
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0417507918529505,
+      "learning_rate": 1.975019450787889e-05,
+      "loss": 0.2595,
+      "step": 1950
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1414670268025355,
+      "learning_rate": 1.9749828551801875e-05,
+      "loss": 0.2483,
+      "step": 1951
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0376239315288212,
+      "learning_rate": 1.974946233126021e-05,
+      "loss": 0.231,
+      "step": 1952
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1091441540992661,
+      "learning_rate": 1.9749095846263828e-05,
+      "loss": 0.2234,
+      "step": 1953
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.087702774894137,
+      "learning_rate": 1.974872909682267e-05,
+      "loss": 0.2496,
+      "step": 1954
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0600141655170918,
+      "learning_rate": 1.974836208294669e-05,
+      "loss": 0.2432,
+      "step": 1955
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3603611052790359,
+      "learning_rate": 1.9747994804645835e-05,
+      "loss": 0.2283,
+      "step": 1956
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0742920942448784,
+      "learning_rate": 1.9747627261930066e-05,
+      "loss": 0.2387,
+      "step": 1957
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1454768361187495,
+      "learning_rate": 1.974725945480936e-05,
+      "loss": 0.2743,
+      "step": 1958
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.033670082913112,
+      "learning_rate": 1.9746891383293692e-05,
+      "loss": 0.2661,
+      "step": 1959
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2940527105563957,
+      "learning_rate": 1.9746523047393046e-05,
+      "loss": 0.2791,
+      "step": 1960
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8717314285910045,
+      "learning_rate": 1.974615444711741e-05,
+      "loss": 0.2261,
+      "step": 1961
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1675295248221975,
+      "learning_rate": 1.974578558247678e-05,
+      "loss": 0.2474,
+      "step": 1962
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1768908708954267,
+      "learning_rate": 1.9745416453481168e-05,
+      "loss": 0.2453,
+      "step": 1963
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0965930944631577,
+      "learning_rate": 1.974504706014059e-05,
+      "loss": 0.2422,
+      "step": 1964
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3209827499231572,
+      "learning_rate": 1.9744677402465053e-05,
+      "loss": 0.2617,
+      "step": 1965
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0704828420643713,
+      "learning_rate": 1.9744307480464595e-05,
+      "loss": 0.24,
+      "step": 1966
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.000209014735969,
+      "learning_rate": 1.9743937294149244e-05,
+      "loss": 0.2334,
+      "step": 1967
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2831002301749044,
+      "learning_rate": 1.9743566843529045e-05,
+      "loss": 0.2512,
+      "step": 1968
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1344015282589346,
+      "learning_rate": 1.9743196128614045e-05,
+      "loss": 0.2488,
+      "step": 1969
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9536869918552521,
+      "learning_rate": 1.97428251494143e-05,
+      "loss": 0.2471,
+      "step": 1970
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2110892765554457,
+      "learning_rate": 1.974245390593987e-05,
+      "loss": 0.2298,
+      "step": 1971
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1996062868964599,
+      "learning_rate": 1.974208239820083e-05,
+      "loss": 0.2441,
+      "step": 1972
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8974209218869812,
+      "learning_rate": 1.9741710626207255e-05,
+      "loss": 0.2358,
+      "step": 1973
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4554750921379593,
+      "learning_rate": 1.9741338589969226e-05,
+      "loss": 0.263,
+      "step": 1974
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0392813150984546,
+      "learning_rate": 1.9740966289496844e-05,
+      "loss": 0.2458,
+      "step": 1975
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9963122217447735,
+      "learning_rate": 1.9740593724800194e-05,
+      "loss": 0.26,
+      "step": 1976
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9737554516896297,
+      "learning_rate": 1.9740220895889393e-05,
+      "loss": 0.2259,
+      "step": 1977
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8923416935134135,
+      "learning_rate": 1.973984780277455e-05,
+      "loss": 0.2476,
+      "step": 1978
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.172847400701586,
+      "learning_rate": 1.9739474445465783e-05,
+      "loss": 0.2455,
+      "step": 1979
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9256808107004492,
+      "learning_rate": 1.9739100823973226e-05,
+      "loss": 0.2285,
+      "step": 1980
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5604533353401182,
+      "learning_rate": 1.9738726938307e-05,
+      "loss": 0.2567,
+      "step": 1981
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8144426438895492,
+      "learning_rate": 1.9738352788477268e-05,
+      "loss": 0.2263,
+      "step": 1982
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0438242662359962,
+      "learning_rate": 1.9737978374494157e-05,
+      "loss": 0.2192,
+      "step": 1983
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3678826201890977,
+      "learning_rate": 1.9737603696367836e-05,
+      "loss": 0.2379,
+      "step": 1984
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.909867140325504,
+      "learning_rate": 1.9737228754108467e-05,
+      "loss": 0.2626,
+      "step": 1985
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0062470083627526,
+      "learning_rate": 1.9736853547726214e-05,
+      "loss": 0.2115,
+      "step": 1986
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9377825018993832,
+      "learning_rate": 1.973647807723126e-05,
+      "loss": 0.255,
+      "step": 1987
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0494091703851158,
+      "learning_rate": 1.973610234263379e-05,
+      "loss": 0.2518,
+      "step": 1988
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1987444586881062,
+      "learning_rate": 1.9735726343943992e-05,
+      "loss": 0.2384,
+      "step": 1989
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.02642903480859,
+      "learning_rate": 1.973535008117207e-05,
+      "loss": 0.267,
+      "step": 1990
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9161340787331664,
+      "learning_rate": 1.9734973554328223e-05,
+      "loss": 0.2229,
+      "step": 1991
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8948289353381529,
+      "learning_rate": 1.9734596763422672e-05,
+      "loss": 0.2656,
+      "step": 1992
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.918415437776545,
+      "learning_rate": 1.973421970846563e-05,
+      "loss": 0.2379,
+      "step": 1993
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9143249901978743,
+      "learning_rate": 1.9733842389467334e-05,
+      "loss": 0.254,
+      "step": 1994
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8228313878161293,
+      "learning_rate": 1.9733464806438007e-05,
+      "loss": 0.2356,
+      "step": 1995
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9175875302693642,
+      "learning_rate": 1.97330869593879e-05,
+      "loss": 0.2135,
+      "step": 1996
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1814645469727503,
+      "learning_rate": 1.973270884832726e-05,
+      "loss": 0.2161,
+      "step": 1997
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9231461656642026,
+      "learning_rate": 1.9732330473266347e-05,
+      "loss": 0.2537,
+      "step": 1998
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9796736831409919,
+      "learning_rate": 1.9731951834215414e-05,
+      "loss": 0.2183,
+      "step": 1999
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9496265829684779,
+      "learning_rate": 1.973157293118474e-05,
+      "loss": 0.2479,
+      "step": 2000
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9231151077653849,
+      "learning_rate": 1.9731193764184603e-05,
+      "loss": 0.222,
+      "step": 2001
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0860127244202018,
+      "learning_rate": 1.9730814333225285e-05,
+      "loss": 0.2175,
+      "step": 2002
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.179552345473745,
+      "learning_rate": 1.9730434638317076e-05,
+      "loss": 0.2203,
+      "step": 2003
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1933311940151612,
+      "learning_rate": 1.9730054679470278e-05,
+      "loss": 0.2185,
+      "step": 2004
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8347201140855571,
+      "learning_rate": 1.97296744566952e-05,
+      "loss": 0.2103,
+      "step": 2005
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4454228072145152,
+      "learning_rate": 1.9729293970002146e-05,
+      "loss": 0.2386,
+      "step": 2006
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.095418832067566,
+      "learning_rate": 1.972891321940145e-05,
+      "loss": 0.236,
+      "step": 2007
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9516989176459635,
+      "learning_rate": 1.9728532204903433e-05,
+      "loss": 0.2302,
+      "step": 2008
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0057330914000895,
+      "learning_rate": 1.972815092651843e-05,
+      "loss": 0.2372,
+      "step": 2009
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.895760637885882,
+      "learning_rate": 1.9727769384256784e-05,
+      "loss": 0.2467,
+      "step": 2010
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9447309607983849,
+      "learning_rate": 1.972738757812884e-05,
+      "loss": 0.2336,
+      "step": 2011
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8281939870261643,
+      "learning_rate": 1.972700550814496e-05,
+      "loss": 0.2278,
+      "step": 2012
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9350882822101689,
+      "learning_rate": 1.9726623174315513e-05,
+      "loss": 0.2353,
+      "step": 2013
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9463149137617953,
+      "learning_rate": 1.9726240576650856e-05,
+      "loss": 0.2321,
+      "step": 2014
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9948397774746942,
+      "learning_rate": 1.9725857715161375e-05,
+      "loss": 0.2411,
+      "step": 2015
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2647259419814574,
+      "learning_rate": 1.9725474589857456e-05,
+      "loss": 0.2351,
+      "step": 2016
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9016649755178238,
+      "learning_rate": 1.972509120074949e-05,
+      "loss": 0.2386,
+      "step": 2017
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1877608427174775,
+      "learning_rate": 1.9724707547847873e-05,
+      "loss": 0.2469,
+      "step": 2018
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7100428965467536,
+      "learning_rate": 1.9724323631163016e-05,
+      "loss": 0.2349,
+      "step": 2019
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.345463433391392,
+      "learning_rate": 1.972393945070533e-05,
+      "loss": 0.2513,
+      "step": 2020
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9091717965875171,
+      "learning_rate": 1.972355500648524e-05,
+      "loss": 0.233,
+      "step": 2021
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.223819582299652,
+      "learning_rate": 1.9723170298513166e-05,
+      "loss": 0.2505,
+      "step": 2022
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2243075830635146,
+      "learning_rate": 1.9722785326799554e-05,
+      "loss": 0.2262,
+      "step": 2023
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0350825147530838,
+      "learning_rate": 1.9722400091354837e-05,
+      "loss": 0.2277,
+      "step": 2024
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.935339865872896,
+      "learning_rate": 1.9722014592189472e-05,
+      "loss": 0.22,
+      "step": 2025
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.166453007673294,
+      "learning_rate": 1.972162882931391e-05,
+      "loss": 0.2548,
+      "step": 2026
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0351107766845962,
+      "learning_rate": 1.9721242802738615e-05,
+      "loss": 0.2426,
+      "step": 2027
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.084600790162855,
+      "learning_rate": 1.9720856512474065e-05,
+      "loss": 0.2137,
+      "step": 2028
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9560170541274214,
+      "learning_rate": 1.972046995853073e-05,
+      "loss": 0.2629,
+      "step": 2029
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4103631271603914,
+      "learning_rate": 1.9720083140919097e-05,
+      "loss": 0.2184,
+      "step": 2030
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4870866452091123,
+      "learning_rate": 1.9719696059649665e-05,
+      "loss": 0.2314,
+      "step": 2031
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0031746692829822,
+      "learning_rate": 1.9719308714732924e-05,
+      "loss": 0.2556,
+      "step": 2032
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0804349871082373,
+      "learning_rate": 1.9718921106179384e-05,
+      "loss": 0.2393,
+      "step": 2033
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.029145379261695,
+      "learning_rate": 1.9718533233999565e-05,
+      "loss": 0.2449,
+      "step": 2034
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.075889733451575,
+      "learning_rate": 1.9718145098203977e-05,
+      "loss": 0.2061,
+      "step": 2035
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5370899898173875,
+      "learning_rate": 1.971775669880316e-05,
+      "loss": 0.2207,
+      "step": 2036
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9389617118447757,
+      "learning_rate": 1.971736803580764e-05,
+      "loss": 0.2453,
+      "step": 2037
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9431192525993586,
+      "learning_rate": 1.9716979109227965e-05,
+      "loss": 0.2128,
+      "step": 2038
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8151755033716819,
+      "learning_rate": 1.9716589919074682e-05,
+      "loss": 0.2378,
+      "step": 2039
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9130866798995204,
+      "learning_rate": 1.9716200465358352e-05,
+      "loss": 0.2421,
+      "step": 2040
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0000040861444954,
+      "learning_rate": 1.971581074808953e-05,
+      "loss": 0.2559,
+      "step": 2041
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0485492882545504,
+      "learning_rate": 1.9715420767278794e-05,
+      "loss": 0.2318,
+      "step": 2042
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.137323578167587,
+      "learning_rate": 1.9715030522936724e-05,
+      "loss": 0.2508,
+      "step": 2043
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2110860738142555,
+      "learning_rate": 1.9714640015073902e-05,
+      "loss": 0.234,
+      "step": 2044
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9126273522332989,
+      "learning_rate": 1.9714249243700916e-05,
+      "loss": 0.2512,
+      "step": 2045
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.105935543821545,
+      "learning_rate": 1.9713858208828376e-05,
+      "loss": 0.2222,
+      "step": 2046
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.176499937899214,
+      "learning_rate": 1.971346691046688e-05,
+      "loss": 0.2465,
+      "step": 2047
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.044320635285446,
+      "learning_rate": 1.971307534862705e-05,
+      "loss": 0.2773,
+      "step": 2048
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8622995262128048,
+      "learning_rate": 1.9712683523319498e-05,
+      "loss": 0.2372,
+      "step": 2049
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9419349078412376,
+      "learning_rate": 1.9712291434554858e-05,
+      "loss": 0.2445,
+      "step": 2050
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.051102010160456,
+      "learning_rate": 1.9711899082343763e-05,
+      "loss": 0.2449,
+      "step": 2051
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9448524300694392,
+      "learning_rate": 1.971150646669686e-05,
+      "loss": 0.2452,
+      "step": 2052
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.216094189344481,
+      "learning_rate": 1.9711113587624795e-05,
+      "loss": 0.2247,
+      "step": 2053
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.063800136560177,
+      "learning_rate": 1.9710720445138225e-05,
+      "loss": 0.2196,
+      "step": 2054
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6769989193035522,
+      "learning_rate": 1.9710327039247814e-05,
+      "loss": 0.2664,
+      "step": 2055
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2891307743333376,
+      "learning_rate": 1.9709933369964235e-05,
+      "loss": 0.2428,
+      "step": 2056
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8979671748669272,
+      "learning_rate": 1.970953943729816e-05,
+      "loss": 0.2286,
+      "step": 2057
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3718627427597734,
+      "learning_rate": 1.9709145241260283e-05,
+      "loss": 0.2354,
+      "step": 2058
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4770658344351117,
+      "learning_rate": 1.9708750781861294e-05,
+      "loss": 0.2556,
+      "step": 2059
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5425129152295496,
+      "learning_rate": 1.970835605911189e-05,
+      "loss": 0.2405,
+      "step": 2060
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0673330540386734,
+      "learning_rate": 1.970796107302278e-05,
+      "loss": 0.2262,
+      "step": 2061
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2216615188163307,
+      "learning_rate": 1.970756582360468e-05,
+      "loss": 0.2431,
+      "step": 2062
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.370798987347387,
+      "learning_rate": 1.9707170310868303e-05,
+      "loss": 0.2266,
+      "step": 2063
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.357861573424895,
+      "learning_rate": 1.9706774534824387e-05,
+      "loss": 0.2137,
+      "step": 2064
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.991296929358314,
+      "learning_rate": 1.9706378495483664e-05,
+      "loss": 0.2243,
+      "step": 2065
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1184402264270916,
+      "learning_rate": 1.9705982192856874e-05,
+      "loss": 0.2171,
+      "step": 2066
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0273223493964336,
+      "learning_rate": 1.9705585626954772e-05,
+      "loss": 0.23,
+      "step": 2067
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0079792072394853,
+      "learning_rate": 1.9705188797788108e-05,
+      "loss": 0.2138,
+      "step": 2068
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7765221583486901,
+      "learning_rate": 1.9704791705367653e-05,
+      "loss": 0.2036,
+      "step": 2069
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4568975454390984,
+      "learning_rate": 1.9704394349704174e-05,
+      "loss": 0.2434,
+      "step": 2070
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.221785001763761,
+      "learning_rate": 1.970399673080845e-05,
+      "loss": 0.2401,
+      "step": 2071
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0248697227267982,
+      "learning_rate": 1.970359884869126e-05,
+      "loss": 0.2156,
+      "step": 2072
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.10514229232445,
+      "learning_rate": 1.9703200703363415e-05,
+      "loss": 0.2445,
+      "step": 2073
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2233075570824092,
+      "learning_rate": 1.9702802294835695e-05,
+      "loss": 0.2647,
+      "step": 2074
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.411425527125798,
+      "learning_rate": 1.9702403623118918e-05,
+      "loss": 0.2264,
+      "step": 2075
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.255491000596999,
+      "learning_rate": 1.970200468822389e-05,
+      "loss": 0.2211,
+      "step": 2076
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1531607324724602,
+      "learning_rate": 1.970160549016144e-05,
+      "loss": 0.2471,
+      "step": 2077
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1173493303154376,
+      "learning_rate": 1.9701206028942398e-05,
+      "loss": 0.2495,
+      "step": 2078
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1526086421139528,
+      "learning_rate": 1.970080630457759e-05,
+      "loss": 0.2303,
+      "step": 2079
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9464962515330188,
+      "learning_rate": 1.970040631707786e-05,
+      "loss": 0.2603,
+      "step": 2080
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8871094596594084,
+      "learning_rate": 1.9700006066454066e-05,
+      "loss": 0.2512,
+      "step": 2081
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0292996944262494,
+      "learning_rate": 1.9699605552717056e-05,
+      "loss": 0.2385,
+      "step": 2082
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1987567719919687,
+      "learning_rate": 1.96992047758777e-05,
+      "loss": 0.2412,
+      "step": 2083
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8475327484829858,
+      "learning_rate": 1.9698803735946867e-05,
+      "loss": 0.2352,
+      "step": 2084
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9053438157379965,
+      "learning_rate": 1.9698402432935432e-05,
+      "loss": 0.2431,
+      "step": 2085
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.336674943053752,
+      "learning_rate": 1.9698000866854284e-05,
+      "loss": 0.2356,
+      "step": 2086
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2977002169496847,
+      "learning_rate": 1.9697599037714315e-05,
+      "loss": 0.2374,
+      "step": 2087
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2799090702528029,
+      "learning_rate": 1.9697196945526427e-05,
+      "loss": 0.2252,
+      "step": 2088
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9752193191002771,
+      "learning_rate": 1.969679459030152e-05,
+      "loss": 0.2678,
+      "step": 2089
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8762395610925323,
+      "learning_rate": 1.9696391972050516e-05,
+      "loss": 0.2144,
+      "step": 2090
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1460100146302559,
+      "learning_rate": 1.969598909078433e-05,
+      "loss": 0.2508,
+      "step": 2091
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9559822932848991,
+      "learning_rate": 1.969558594651389e-05,
+      "loss": 0.2487,
+      "step": 2092
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9839613710855772,
+      "learning_rate": 1.9695182539250138e-05,
+      "loss": 0.2392,
+      "step": 2093
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8264913114980849,
+      "learning_rate": 1.969477886900401e-05,
+      "loss": 0.2213,
+      "step": 2094
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0003343905023958,
+      "learning_rate": 1.9694374935786457e-05,
+      "loss": 0.2235,
+      "step": 2095
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5516384795930434,
+      "learning_rate": 1.9693970739608437e-05,
+      "loss": 0.2463,
+      "step": 2096
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0876525839249633,
+      "learning_rate": 1.9693566280480914e-05,
+      "loss": 0.226,
+      "step": 2097
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.346822529145253,
+      "learning_rate": 1.9693161558414856e-05,
+      "loss": 0.2393,
+      "step": 2098
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5812948122941315,
+      "learning_rate": 1.9692756573421246e-05,
+      "loss": 0.2685,
+      "step": 2099
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9825131769766344,
+      "learning_rate": 1.9692351325511066e-05,
+      "loss": 0.2434,
+      "step": 2100
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.517270664330921,
+      "learning_rate": 1.9691945814695306e-05,
+      "loss": 0.2442,
+      "step": 2101
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5284733760701465,
+      "learning_rate": 1.9691540040984972e-05,
+      "loss": 0.2366,
+      "step": 2102
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7611185683006056,
+      "learning_rate": 1.9691134004391064e-05,
+      "loss": 0.2571,
+      "step": 2103
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2502143870501974,
+      "learning_rate": 1.9690727704924598e-05,
+      "loss": 0.2439,
+      "step": 2104
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.155788006118132,
+      "learning_rate": 1.9690321142596602e-05,
+      "loss": 0.2368,
+      "step": 2105
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0104694476849565,
+      "learning_rate": 1.968991431741809e-05,
+      "loss": 0.2437,
+      "step": 2106
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3811992101020232,
+      "learning_rate": 1.968950722940011e-05,
+      "loss": 0.2269,
+      "step": 2107
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0196354555505611,
+      "learning_rate": 1.9689099878553698e-05,
+      "loss": 0.2316,
+      "step": 2108
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.701387510634705,
+      "learning_rate": 1.9688692264889905e-05,
+      "loss": 0.2166,
+      "step": 2109
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8829947026926955,
+      "learning_rate": 1.9688284388419784e-05,
+      "loss": 0.2411,
+      "step": 2110
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1844084066745377,
+      "learning_rate": 1.9687876249154402e-05,
+      "loss": 0.2682,
+      "step": 2111
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0692862669633376,
+      "learning_rate": 1.9687467847104834e-05,
+      "loss": 0.2247,
+      "step": 2112
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.931930597121846,
+      "learning_rate": 1.9687059182282152e-05,
+      "loss": 0.2397,
+      "step": 2113
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9933543404165622,
+      "learning_rate": 1.968665025469744e-05,
+      "loss": 0.2026,
+      "step": 2114
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2298239063976584,
+      "learning_rate": 1.9686241064361792e-05,
+      "loss": 0.2491,
+      "step": 2115
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0308486801387686,
+      "learning_rate": 1.9685831611286312e-05,
+      "loss": 0.2167,
+      "step": 2116
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.232081656908198,
+      "learning_rate": 1.96854218954821e-05,
+      "loss": 0.2442,
+      "step": 2117
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1575287552055014,
+      "learning_rate": 1.9685011916960276e-05,
+      "loss": 0.2477,
+      "step": 2118
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1704875982681098,
+      "learning_rate": 1.9684601675731952e-05,
+      "loss": 0.2328,
+      "step": 2119
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0464516245169295,
+      "learning_rate": 1.9684191171808262e-05,
+      "loss": 0.2445,
+      "step": 2120
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.238441161592338,
+      "learning_rate": 1.968378040520034e-05,
+      "loss": 0.2454,
+      "step": 2121
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1120056639856326,
+      "learning_rate": 1.9683369375919325e-05,
+      "loss": 0.2353,
+      "step": 2122
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4210318912864548,
+      "learning_rate": 1.9682958083976374e-05,
+      "loss": 0.2377,
+      "step": 2123
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9675502607652713,
+      "learning_rate": 1.9682546529382635e-05,
+      "loss": 0.2382,
+      "step": 2124
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8799019495856499,
+      "learning_rate": 1.968213471214927e-05,
+      "loss": 0.2327,
+      "step": 2125
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1715411569228944,
+      "learning_rate": 1.968172263228746e-05,
+      "loss": 0.237,
+      "step": 2126
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9671697534047214,
+      "learning_rate": 1.9681310289808377e-05,
+      "loss": 0.2416,
+      "step": 2127
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0001367934628327,
+      "learning_rate": 1.9680897684723205e-05,
+      "loss": 0.2347,
+      "step": 2128
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8494973958552655,
+      "learning_rate": 1.9680484817043134e-05,
+      "loss": 0.2301,
+      "step": 2129
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9240069516003688,
+      "learning_rate": 1.9680071686779368e-05,
+      "loss": 0.2379,
+      "step": 2130
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3136456921955122,
+      "learning_rate": 1.9679658293943112e-05,
+      "loss": 0.2417,
+      "step": 2131
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0592213424376784,
+      "learning_rate": 1.9679244638545572e-05,
+      "loss": 0.2174,
+      "step": 2132
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.910805874421009,
+      "learning_rate": 1.967883072059798e-05,
+      "loss": 0.237,
+      "step": 2133
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8839617866905263,
+      "learning_rate": 1.9678416540111557e-05,
+      "loss": 0.2356,
+      "step": 2134
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2152257087176324,
+      "learning_rate": 1.9678002097097537e-05,
+      "loss": 0.2349,
+      "step": 2135
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1044155337772443,
+      "learning_rate": 1.9677587391567164e-05,
+      "loss": 0.2365,
+      "step": 2136
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7787980072399576,
+      "learning_rate": 1.967717242353169e-05,
+      "loss": 0.2265,
+      "step": 2137
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3616430182996238,
+      "learning_rate": 1.9676757193002363e-05,
+      "loss": 0.2746,
+      "step": 2138
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1228586924105723,
+      "learning_rate": 1.9676341699990452e-05,
+      "loss": 0.2186,
+      "step": 2139
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2125802259045932,
+      "learning_rate": 1.9675925944507226e-05,
+      "loss": 0.2519,
+      "step": 2140
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9444977302475596,
+      "learning_rate": 1.9675509926563964e-05,
+      "loss": 0.2566,
+      "step": 2141
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5317024122818927,
+      "learning_rate": 1.9675093646171947e-05,
+      "loss": 0.2401,
+      "step": 2142
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7667854821812148,
+      "learning_rate": 1.967467710334247e-05,
+      "loss": 0.2389,
+      "step": 2143
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.189612959109022,
+      "learning_rate": 1.9674260298086825e-05,
+      "loss": 0.2352,
+      "step": 2144
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.992641156412239,
+      "learning_rate": 1.967384323041633e-05,
+      "loss": 0.225,
+      "step": 2145
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9343197806222634,
+      "learning_rate": 1.9673425900342286e-05,
+      "loss": 0.2178,
+      "step": 2146
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1016381303751634,
+      "learning_rate": 1.9673008307876017e-05,
+      "loss": 0.2271,
+      "step": 2147
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9690005406542,
+      "learning_rate": 1.9672590453028855e-05,
+      "loss": 0.228,
+      "step": 2148
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.235254056286498,
+      "learning_rate": 1.967217233581213e-05,
+      "loss": 0.2428,
+      "step": 2149
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.109454195598941,
+      "learning_rate": 1.9671753956237187e-05,
+      "loss": 0.2327,
+      "step": 2150
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5720857648078095,
+      "learning_rate": 1.9671335314315365e-05,
+      "loss": 0.2501,
+      "step": 2151
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0473294245674567,
+      "learning_rate": 1.967091641005803e-05,
+      "loss": 0.222,
+      "step": 2152
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2244939085245228,
+      "learning_rate": 1.967049724347654e-05,
+      "loss": 0.2284,
+      "step": 2153
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3558829537666295,
+      "learning_rate": 1.967007781458227e-05,
+      "loss": 0.2353,
+      "step": 2154
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9497548159666054,
+      "learning_rate": 1.966965812338659e-05,
+      "loss": 0.2122,
+      "step": 2155
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9943932518361649,
+      "learning_rate": 1.9669238169900886e-05,
+      "loss": 0.2103,
+      "step": 2156
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.593225948018636,
+      "learning_rate": 1.966881795413655e-05,
+      "loss": 0.2542,
+      "step": 2157
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.282690775096383,
+      "learning_rate": 1.9668397476104983e-05,
+      "loss": 0.2286,
+      "step": 2158
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.101149696753262,
+      "learning_rate": 1.966797673581759e-05,
+      "loss": 0.2089,
+      "step": 2159
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.905210614202573,
+      "learning_rate": 1.966755573328578e-05,
+      "loss": 0.23,
+      "step": 2160
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0378262137010232,
+      "learning_rate": 1.9667134468520974e-05,
+      "loss": 0.2432,
+      "step": 2161
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0797024746109964,
+      "learning_rate": 1.96667129415346e-05,
+      "loss": 0.2282,
+      "step": 2162
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2276995992881126,
+      "learning_rate": 1.966629115233809e-05,
+      "loss": 0.211,
+      "step": 2163
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0352266908341288,
+      "learning_rate": 1.9665869100942888e-05,
+      "loss": 0.2472,
+      "step": 2164
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.417599642847786,
+      "learning_rate": 1.9665446787360444e-05,
+      "loss": 0.2273,
+      "step": 2165
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9920449774927103,
+      "learning_rate": 1.9665024211602208e-05,
+      "loss": 0.242,
+      "step": 2166
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2296252775505276,
+      "learning_rate": 1.9664601373679644e-05,
+      "loss": 0.2491,
+      "step": 2167
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0402566138277567,
+      "learning_rate": 1.966417827360422e-05,
+      "loss": 0.2284,
+      "step": 2168
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1053617884173759,
+      "learning_rate": 1.9663754911387414e-05,
+      "loss": 0.2385,
+      "step": 2169
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1119095497921,
+      "learning_rate": 1.9663331287040713e-05,
+      "loss": 0.2539,
+      "step": 2170
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0983948450880103,
+      "learning_rate": 1.9662907400575606e-05,
+      "loss": 0.2137,
+      "step": 2171
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.269949230715848,
+      "learning_rate": 1.9662483252003585e-05,
+      "loss": 0.2293,
+      "step": 2172
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0564708933309246,
+      "learning_rate": 1.9662058841336164e-05,
+      "loss": 0.2217,
+      "step": 2173
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1216048180385836,
+      "learning_rate": 1.966163416858485e-05,
+      "loss": 0.2748,
+      "step": 2174
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.508310528343341,
+      "learning_rate": 1.9661209233761167e-05,
+      "loss": 0.227,
+      "step": 2175
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.094002964335264,
+      "learning_rate": 1.9660784036876636e-05,
+      "loss": 0.2422,
+      "step": 2176
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2065998099809236,
+      "learning_rate": 1.9660358577942788e-05,
+      "loss": 0.2282,
+      "step": 2177
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5059749594559086,
+      "learning_rate": 1.965993285697117e-05,
+      "loss": 0.2618,
+      "step": 2178
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0542248630235662,
+      "learning_rate": 1.965950687397333e-05,
+      "loss": 0.227,
+      "step": 2179
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1505226952579715,
+      "learning_rate": 1.965908062896082e-05,
+      "loss": 0.2321,
+      "step": 2180
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9613747888101222,
+      "learning_rate": 1.96586541219452e-05,
+      "loss": 0.2532,
+      "step": 2181
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1209519540120025,
+      "learning_rate": 1.9658227352938044e-05,
+      "loss": 0.215,
+      "step": 2182
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1901700934240698,
+      "learning_rate": 1.9657800321950925e-05,
+      "loss": 0.233,
+      "step": 2183
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0703733820154298,
+      "learning_rate": 1.9657373028995427e-05,
+      "loss": 0.2294,
+      "step": 2184
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0694757472302956,
+      "learning_rate": 1.965694547408314e-05,
+      "loss": 0.2351,
+      "step": 2185
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.06619999016574,
+      "learning_rate": 1.9656517657225658e-05,
+      "loss": 0.2091,
+      "step": 2186
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2948971683345574,
+      "learning_rate": 1.9656089578434595e-05,
+      "loss": 0.2643,
+      "step": 2187
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.16157269655789,
+      "learning_rate": 1.9655661237721554e-05,
+      "loss": 0.2298,
+      "step": 2188
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4539573594884465,
+      "learning_rate": 1.9655232635098157e-05,
+      "loss": 0.2289,
+      "step": 2189
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7694359425810158,
+      "learning_rate": 1.965480377057603e-05,
+      "loss": 0.1937,
+      "step": 2190
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3928943982351816,
+      "learning_rate": 1.96543746441668e-05,
+      "loss": 0.2257,
+      "step": 2191
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2322222758893566,
+      "learning_rate": 1.965394525588212e-05,
+      "loss": 0.2518,
+      "step": 2192
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2909786558614489,
+      "learning_rate": 1.9653515605733625e-05,
+      "loss": 0.2674,
+      "step": 2193
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.610007776399509,
+      "learning_rate": 1.9653085693732976e-05,
+      "loss": 0.24,
+      "step": 2194
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.224363132147249,
+      "learning_rate": 1.965265551989183e-05,
+      "loss": 0.2476,
+      "step": 2195
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.105382530444532,
+      "learning_rate": 1.965222508422186e-05,
+      "loss": 0.2573,
+      "step": 2196
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.146983000583085,
+      "learning_rate": 1.9651794386734743e-05,
+      "loss": 0.2428,
+      "step": 2197
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1987218803478044,
+      "learning_rate": 1.965136342744215e-05,
+      "loss": 0.2421,
+      "step": 2198
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3590799693199993,
+      "learning_rate": 1.9650932206355786e-05,
+      "loss": 0.2656,
+      "step": 2199
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9900346782252395,
+      "learning_rate": 1.9650500723487335e-05,
+      "loss": 0.2067,
+      "step": 2200
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.077951461086788,
+      "learning_rate": 1.9650068978848512e-05,
+      "loss": 0.2341,
+      "step": 2201
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.907756752533414,
+      "learning_rate": 1.964963697245102e-05,
+      "loss": 0.2265,
+      "step": 2202
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2688186973976037,
+      "learning_rate": 1.964920470430658e-05,
+      "loss": 0.2449,
+      "step": 2203
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1829663405600277,
+      "learning_rate": 1.964877217442692e-05,
+      "loss": 0.2643,
+      "step": 2204
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8793853900121442,
+      "learning_rate": 1.964833938282377e-05,
+      "loss": 0.2251,
+      "step": 2205
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2663396192955865,
+      "learning_rate": 1.9647906329508866e-05,
+      "loss": 0.2563,
+      "step": 2206
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.086276962256967,
+      "learning_rate": 1.9647473014493958e-05,
+      "loss": 0.2063,
+      "step": 2207
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6805919171172707,
+      "learning_rate": 1.9647039437790802e-05,
+      "loss": 0.2422,
+      "step": 2208
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3180450409740612,
+      "learning_rate": 1.9646605599411155e-05,
+      "loss": 0.2218,
+      "step": 2209
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0908244445799435,
+      "learning_rate": 1.964617149936679e-05,
+      "loss": 0.2358,
+      "step": 2210
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3992407176335435,
+      "learning_rate": 1.9645737137669473e-05,
+      "loss": 0.2655,
+      "step": 2211
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2630430210609305,
+      "learning_rate": 1.9645302514330994e-05,
+      "loss": 0.2403,
+      "step": 2212
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1339372584391034,
+      "learning_rate": 1.9644867629363137e-05,
+      "loss": 0.2247,
+      "step": 2213
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1341369372627903,
+      "learning_rate": 1.9644432482777703e-05,
+      "loss": 0.2674,
+      "step": 2214
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3731513614446267,
+      "learning_rate": 1.964399707458649e-05,
+      "loss": 0.2195,
+      "step": 2215
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3995092356589611,
+      "learning_rate": 1.9643561404801317e-05,
+      "loss": 0.2371,
+      "step": 2216
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0985880612220564,
+      "learning_rate": 1.9643125473433992e-05,
+      "loss": 0.2136,
+      "step": 2217
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8997296488322348,
+      "learning_rate": 1.9642689280496347e-05,
+      "loss": 0.1975,
+      "step": 2218
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7606210335459154,
+      "learning_rate": 1.9642252826000206e-05,
+      "loss": 0.2158,
+      "step": 2219
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9434073289543887,
+      "learning_rate": 1.9641816109957415e-05,
+      "loss": 0.2276,
+      "step": 2220
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9260696285730123,
+      "learning_rate": 1.9641379132379822e-05,
+      "loss": 0.2486,
+      "step": 2221
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.058546288904667,
+      "learning_rate": 1.964094189327927e-05,
+      "loss": 0.2338,
+      "step": 2222
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.394988426693415,
+      "learning_rate": 1.9640504392667626e-05,
+      "loss": 0.2476,
+      "step": 2223
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0985601759324406,
+      "learning_rate": 1.9640066630556756e-05,
+      "loss": 0.2159,
+      "step": 2224
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2145342696137336,
+      "learning_rate": 1.9639628606958535e-05,
+      "loss": 0.2311,
+      "step": 2225
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4288512822935417,
+      "learning_rate": 1.9639190321884842e-05,
+      "loss": 0.2285,
+      "step": 2226
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1520142737740142,
+      "learning_rate": 1.9638751775347568e-05,
+      "loss": 0.2225,
+      "step": 2227
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0531822751852433,
+      "learning_rate": 1.963831296735861e-05,
+      "loss": 0.2425,
+      "step": 2228
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.359747733426235,
+      "learning_rate": 1.9637873897929866e-05,
+      "loss": 0.2537,
+      "step": 2229
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.126644132650638,
+      "learning_rate": 1.9637434567073246e-05,
+      "loss": 0.2411,
+      "step": 2230
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5801327871398765,
+      "learning_rate": 1.9636994974800673e-05,
+      "loss": 0.2442,
+      "step": 2231
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4688906285868015,
+      "learning_rate": 1.9636555121124063e-05,
+      "loss": 0.2321,
+      "step": 2232
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4452794250821794,
+      "learning_rate": 1.963611500605535e-05,
+      "loss": 0.2282,
+      "step": 2233
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2114241613636223,
+      "learning_rate": 1.963567462960648e-05,
+      "loss": 0.2344,
+      "step": 2234
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.213601910804405,
+      "learning_rate": 1.963523399178939e-05,
+      "loss": 0.2561,
+      "step": 2235
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1851182812942889,
+      "learning_rate": 1.963479309261603e-05,
+      "loss": 0.2364,
+      "step": 2236
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8871984432827353,
+      "learning_rate": 1.9634351932098364e-05,
+      "loss": 0.2364,
+      "step": 2237
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2435776107199095,
+      "learning_rate": 1.9633910510248357e-05,
+      "loss": 0.2381,
+      "step": 2238
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3475281308085763,
+      "learning_rate": 1.9633468827077986e-05,
+      "loss": 0.2418,
+      "step": 2239
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.239713466612954,
+      "learning_rate": 1.9633026882599228e-05,
+      "loss": 0.2504,
+      "step": 2240
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2328229260264705,
+      "learning_rate": 1.963258467682407e-05,
+      "loss": 0.241,
+      "step": 2241
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0341545828230583,
+      "learning_rate": 1.9632142209764514e-05,
+      "loss": 0.2174,
+      "step": 2242
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2902668946694982,
+      "learning_rate": 1.963169948143255e-05,
+      "loss": 0.2275,
+      "step": 2243
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0121788933691918,
+      "learning_rate": 1.9631256491840197e-05,
+      "loss": 0.2229,
+      "step": 2244
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9654980378203284,
+      "learning_rate": 1.9630813240999468e-05,
+      "loss": 0.2423,
+      "step": 2245
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3478278298336825,
+      "learning_rate": 1.963036972892238e-05,
+      "loss": 0.2736,
+      "step": 2246
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2691420401208928,
+      "learning_rate": 1.962992595562098e-05,
+      "loss": 0.2059,
+      "step": 2247
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.374703279576238,
+      "learning_rate": 1.9629481921107287e-05,
+      "loss": 0.2613,
+      "step": 2248
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0652932569776843,
+      "learning_rate": 1.9629037625393352e-05,
+      "loss": 0.2316,
+      "step": 2249
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5137256334450104,
+      "learning_rate": 1.962859306849123e-05,
+      "loss": 0.2392,
+      "step": 2250
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.107016666508369,
+      "learning_rate": 1.962814825041298e-05,
+      "loss": 0.2253,
+      "step": 2251
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9550128240410778,
+      "learning_rate": 1.962770317117066e-05,
+      "loss": 0.1973,
+      "step": 2252
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.297841502121561,
+      "learning_rate": 1.9627257830776352e-05,
+      "loss": 0.2174,
+      "step": 2253
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1093582584200514,
+      "learning_rate": 1.9626812229242128e-05,
+      "loss": 0.2306,
+      "step": 2254
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2245798746882979,
+      "learning_rate": 1.962636636658008e-05,
+      "loss": 0.2535,
+      "step": 2255
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.644044637919812,
+      "learning_rate": 1.9625920242802302e-05,
+      "loss": 0.2554,
+      "step": 2256
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1361609839406392,
+      "learning_rate": 1.962547385792089e-05,
+      "loss": 0.2492,
+      "step": 2257
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6662903840032885,
+      "learning_rate": 1.962502721194796e-05,
+      "loss": 0.224,
+      "step": 2258
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9270518048299475,
+      "learning_rate": 1.962458030489562e-05,
+      "loss": 0.2125,
+      "step": 2259
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9755561228878188,
+      "learning_rate": 1.9624133136775998e-05,
+      "loss": 0.2293,
+      "step": 2260
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8168211138034966,
+      "learning_rate": 1.962368570760122e-05,
+      "loss": 0.2527,
+      "step": 2261
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.481777245879265,
+      "learning_rate": 1.9623238017383426e-05,
+      "loss": 0.2372,
+      "step": 2262
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9735772662420452,
+      "learning_rate": 1.9622790066134754e-05,
+      "loss": 0.2282,
+      "step": 2263
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1837794954838359,
+      "learning_rate": 1.962234185386736e-05,
+      "loss": 0.2151,
+      "step": 2264
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0788263704931076,
+      "learning_rate": 1.9621893380593398e-05,
+      "loss": 0.2346,
+      "step": 2265
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3895384690518788,
+      "learning_rate": 1.9621444646325036e-05,
+      "loss": 0.2514,
+      "step": 2266
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1628169725919841,
+      "learning_rate": 1.9620995651074443e-05,
+      "loss": 0.2242,
+      "step": 2267
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.168404660899102,
+      "learning_rate": 1.9620546394853802e-05,
+      "loss": 0.2227,
+      "step": 2268
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1351391787739238,
+      "learning_rate": 1.9620096877675294e-05,
+      "loss": 0.2563,
+      "step": 2269
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0035468703101598,
+      "learning_rate": 1.9619647099551118e-05,
+      "loss": 0.2242,
+      "step": 2270
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0863642017242183,
+      "learning_rate": 1.9619197060493465e-05,
+      "loss": 0.2245,
+      "step": 2271
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3521259808783213,
+      "learning_rate": 1.9618746760514554e-05,
+      "loss": 0.2247,
+      "step": 2272
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1789057952904218,
+      "learning_rate": 1.9618296199626594e-05,
+      "loss": 0.2412,
+      "step": 2273
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1695044838914121,
+      "learning_rate": 1.9617845377841804e-05,
+      "loss": 0.2295,
+      "step": 2274
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.911670320372682,
+      "learning_rate": 1.9617394295172415e-05,
+      "loss": 0.2248,
+      "step": 2275
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4974238095285362,
+      "learning_rate": 1.9616942951630668e-05,
+      "loss": 0.2576,
+      "step": 2276
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.065884469776496,
+      "learning_rate": 1.9616491347228793e-05,
+      "loss": 0.2129,
+      "step": 2277
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0689950684727134,
+      "learning_rate": 1.961603948197905e-05,
+      "loss": 0.2123,
+      "step": 2278
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.122557763242407,
+      "learning_rate": 1.9615587355893693e-05,
+      "loss": 0.2124,
+      "step": 2279
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6997662112278131,
+      "learning_rate": 1.9615134968984984e-05,
+      "loss": 0.2229,
+      "step": 2280
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.427459018774372,
+      "learning_rate": 1.96146823212652e-05,
+      "loss": 0.2395,
+      "step": 2281
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.174838610488931,
+      "learning_rate": 1.961422941274661e-05,
+      "loss": 0.2395,
+      "step": 2282
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0444712650544752,
+      "learning_rate": 1.9613776243441507e-05,
+      "loss": 0.2293,
+      "step": 2283
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9928779696890382,
+      "learning_rate": 1.9613322813362182e-05,
+      "loss": 0.2452,
+      "step": 2284
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9898668843506018,
+      "learning_rate": 1.961286912252093e-05,
+      "loss": 0.236,
+      "step": 2285
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0320698498409984,
+      "learning_rate": 1.961241517093006e-05,
+      "loss": 0.2426,
+      "step": 2286
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9456851193220972,
+      "learning_rate": 1.9611960958601886e-05,
+      "loss": 0.2317,
+      "step": 2287
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0427893908436772,
+      "learning_rate": 1.9611506485548728e-05,
+      "loss": 0.2243,
+      "step": 2288
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.990790934611057,
+      "learning_rate": 1.9611051751782915e-05,
+      "loss": 0.2276,
+      "step": 2289
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4299898481089397,
+      "learning_rate": 1.961059675731678e-05,
+      "loss": 0.2421,
+      "step": 2290
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0567116203220435,
+      "learning_rate": 1.9610141502162662e-05,
+      "loss": 0.2549,
+      "step": 2291
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9813611728893015,
+      "learning_rate": 1.9609685986332918e-05,
+      "loss": 0.2272,
+      "step": 2292
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2665166548155724,
+      "learning_rate": 1.9609230209839894e-05,
+      "loss": 0.2406,
+      "step": 2293
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3882608563003989,
+      "learning_rate": 1.9608774172695964e-05,
+      "loss": 0.2168,
+      "step": 2294
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2389235543966157,
+      "learning_rate": 1.9608317874913484e-05,
+      "loss": 0.2293,
+      "step": 2295
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.867329992139809,
+      "learning_rate": 1.9607861316504848e-05,
+      "loss": 0.2149,
+      "step": 2296
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9020334001335076,
+      "learning_rate": 1.9607404497482422e-05,
+      "loss": 0.2277,
+      "step": 2297
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.845846595246084,
+      "learning_rate": 1.9606947417858614e-05,
+      "loss": 0.251,
+      "step": 2298
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1287125708824222,
+      "learning_rate": 1.960649007764581e-05,
+      "loss": 0.2455,
+      "step": 2299
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9295412239995064,
+      "learning_rate": 1.960603247685642e-05,
+      "loss": 0.2402,
+      "step": 2300
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9073561280730207,
+      "learning_rate": 1.9605574615502857e-05,
+      "loss": 0.2625,
+      "step": 2301
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.561672595584736,
+      "learning_rate": 1.9605116493597544e-05,
+      "loss": 0.2539,
+      "step": 2302
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9185494380688095,
+      "learning_rate": 1.96046581111529e-05,
+      "loss": 0.2491,
+      "step": 2303
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9986931205874869,
+      "learning_rate": 1.9604199468181363e-05,
+      "loss": 0.2366,
+      "step": 2304
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9453153360183175,
+      "learning_rate": 1.960374056469537e-05,
+      "loss": 0.2174,
+      "step": 2305
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.427772384150711,
+      "learning_rate": 1.9603281400707378e-05,
+      "loss": 0.2388,
+      "step": 2306
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0900731948657933,
+      "learning_rate": 1.9602821976229835e-05,
+      "loss": 0.23,
+      "step": 2307
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9054697290691388,
+      "learning_rate": 1.96023622912752e-05,
+      "loss": 0.2297,
+      "step": 2308
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9245203886194943,
+      "learning_rate": 1.9601902345855944e-05,
+      "loss": 0.2433,
+      "step": 2309
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0519878918731516,
+      "learning_rate": 1.9601442139984548e-05,
+      "loss": 0.2548,
+      "step": 2310
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8339937473761487,
+      "learning_rate": 1.9600981673673488e-05,
+      "loss": 0.2238,
+      "step": 2311
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.178375488144858,
+      "learning_rate": 1.9600520946935263e-05,
+      "loss": 0.221,
+      "step": 2312
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8533874274139382,
+      "learning_rate": 1.9600059959782364e-05,
+      "loss": 0.2291,
+      "step": 2313
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9109953738205394,
+      "learning_rate": 1.9599598712227294e-05,
+      "loss": 0.2276,
+      "step": 2314
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8551206349574578,
+      "learning_rate": 1.9599137204282566e-05,
+      "loss": 0.2404,
+      "step": 2315
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9342971710035325,
+      "learning_rate": 1.95986754359607e-05,
+      "loss": 0.2176,
+      "step": 2316
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0563711413685022,
+      "learning_rate": 1.959821340727422e-05,
+      "loss": 0.2443,
+      "step": 2317
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9026438168465336,
+      "learning_rate": 1.9597751118235662e-05,
+      "loss": 0.2257,
+      "step": 2318
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0318289234723361,
+      "learning_rate": 1.9597288568857563e-05,
+      "loss": 0.219,
+      "step": 2319
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8967950538630484,
+      "learning_rate": 1.9596825759152466e-05,
+      "loss": 0.2617,
+      "step": 2320
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9702418393564891,
+      "learning_rate": 1.959636268913293e-05,
+      "loss": 0.2492,
+      "step": 2321
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.306990740051603,
+      "learning_rate": 1.9595899358811515e-05,
+      "loss": 0.2268,
+      "step": 2322
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8623225432994753,
+      "learning_rate": 1.9595435768200785e-05,
+      "loss": 0.2351,
+      "step": 2323
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8927352070901556,
+      "learning_rate": 1.9594971917313323e-05,
+      "loss": 0.2814,
+      "step": 2324
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1475126857496665,
+      "learning_rate": 1.9594507806161703e-05,
+      "loss": 0.2089,
+      "step": 2325
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9149296536113013,
+      "learning_rate": 1.9594043434758515e-05,
+      "loss": 0.2321,
+      "step": 2326
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.198714270590255,
+      "learning_rate": 1.959357880311636e-05,
+      "loss": 0.2626,
+      "step": 2327
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0938019109764625,
+      "learning_rate": 1.9593113911247836e-05,
+      "loss": 0.2646,
+      "step": 2328
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4410644951371105,
+      "learning_rate": 1.9592648759165555e-05,
+      "loss": 0.2555,
+      "step": 2329
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1039407282923293,
+      "learning_rate": 1.9592183346882135e-05,
+      "loss": 0.2521,
+      "step": 2330
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6247005333228324,
+      "learning_rate": 1.95917176744102e-05,
+      "loss": 0.2366,
+      "step": 2331
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4229636128215575,
+      "learning_rate": 1.9591251741762384e-05,
+      "loss": 0.2175,
+      "step": 2332
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9608883015159911,
+      "learning_rate": 1.959078554895132e-05,
+      "loss": 0.2267,
+      "step": 2333
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8501715708850301,
+      "learning_rate": 1.959031909598966e-05,
+      "loss": 0.2308,
+      "step": 2334
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8588300419178304,
+      "learning_rate": 1.958985238289005e-05,
+      "loss": 0.218,
+      "step": 2335
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1299837382980906,
+      "learning_rate": 1.9589385409665152e-05,
+      "loss": 0.2247,
+      "step": 2336
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8643912599959122,
+      "learning_rate": 1.9588918176327632e-05,
+      "loss": 0.2627,
+      "step": 2337
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0214914522161034,
+      "learning_rate": 1.9588450682890167e-05,
+      "loss": 0.2192,
+      "step": 2338
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8877732443442983,
+      "learning_rate": 1.9587982929365434e-05,
+      "loss": 0.2435,
+      "step": 2339
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2859468661713847,
+      "learning_rate": 1.9587514915766124e-05,
+      "loss": 0.2073,
+      "step": 2340
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0567286604363095,
+      "learning_rate": 1.958704664210493e-05,
+      "loss": 0.2664,
+      "step": 2341
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9992382427909487,
+      "learning_rate": 1.9586578108394555e-05,
+      "loss": 0.2252,
+      "step": 2342
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9196579061142152,
+      "learning_rate": 1.9586109314647705e-05,
+      "loss": 0.228,
+      "step": 2343
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.166095192504243,
+      "learning_rate": 1.9585640260877102e-05,
+      "loss": 0.2525,
+      "step": 2344
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8412404902739583,
+      "learning_rate": 1.958517094709546e-05,
+      "loss": 0.2287,
+      "step": 2345
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.919152691293656,
+      "learning_rate": 1.9584701373315523e-05,
+      "loss": 0.2087,
+      "step": 2346
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3970527229544,
+      "learning_rate": 1.9584231539550012e-05,
+      "loss": 0.2386,
+      "step": 2347
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8814552434609396,
+      "learning_rate": 1.9583761445811686e-05,
+      "loss": 0.204,
+      "step": 2348
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1366633521260399,
+      "learning_rate": 1.9583291092113283e-05,
+      "loss": 0.2125,
+      "step": 2349
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5839696990479637,
+      "learning_rate": 1.958282047846757e-05,
+      "loss": 0.2572,
+      "step": 2350
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.989785096957875,
+      "learning_rate": 1.9582349604887313e-05,
+      "loss": 0.2419,
+      "step": 2351
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.551925396859981,
+      "learning_rate": 1.958187847138528e-05,
+      "loss": 0.2099,
+      "step": 2352
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0599349357850392,
+      "learning_rate": 1.958140707797425e-05,
+      "loss": 0.2289,
+      "step": 2353
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9251893046227032,
+      "learning_rate": 1.9580935424667015e-05,
+      "loss": 0.2462,
+      "step": 2354
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0352904888157413,
+      "learning_rate": 1.9580463511476365e-05,
+      "loss": 0.2442,
+      "step": 2355
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8960283950787143,
+      "learning_rate": 1.95799913384151e-05,
+      "loss": 0.2167,
+      "step": 2356
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0395096607347472,
+      "learning_rate": 1.9579518905496032e-05,
+      "loss": 0.2407,
+      "step": 2357
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9843402187850503,
+      "learning_rate": 1.9579046212731968e-05,
+      "loss": 0.2396,
+      "step": 2358
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.112271692780199,
+      "learning_rate": 1.957857326013574e-05,
+      "loss": 0.2292,
+      "step": 2359
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.221454329725577,
+      "learning_rate": 1.9578100047720164e-05,
+      "loss": 0.2175,
+      "step": 2360
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9484425312273643,
+      "learning_rate": 1.957762657549809e-05,
+      "loss": 0.2285,
+      "step": 2361
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9496247993199408,
+      "learning_rate": 1.957715284348235e-05,
+      "loss": 0.2407,
+      "step": 2362
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0254007460980028,
+      "learning_rate": 1.95766788516858e-05,
+      "loss": 0.2231,
+      "step": 2363
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.896010356490778,
+      "learning_rate": 1.9576204600121293e-05,
+      "loss": 0.2314,
+      "step": 2364
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9778267316564238,
+      "learning_rate": 1.9575730088801696e-05,
+      "loss": 0.2544,
+      "step": 2365
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0418675619291378,
+      "learning_rate": 1.957525531773988e-05,
+      "loss": 0.2347,
+      "step": 2366
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.34539811818473,
+      "learning_rate": 1.9574780286948724e-05,
+      "loss": 0.2461,
+      "step": 2367
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2117159081752875,
+      "learning_rate": 1.957430499644111e-05,
+      "loss": 0.2365,
+      "step": 2368
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9916311588570211,
+      "learning_rate": 1.9573829446229935e-05,
+      "loss": 0.2423,
+      "step": 2369
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8642744609344306,
+      "learning_rate": 1.9573353636328094e-05,
+      "loss": 0.1956,
+      "step": 2370
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2779774017391479,
+      "learning_rate": 1.9572877566748495e-05,
+      "loss": 0.2349,
+      "step": 2371
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9530889379498755,
+      "learning_rate": 1.957240123750405e-05,
+      "loss": 0.2795,
+      "step": 2372
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.63061203732391,
+      "learning_rate": 1.9571924648607684e-05,
+      "loss": 0.2606,
+      "step": 2373
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.206120338499393,
+      "learning_rate": 1.9571447800072318e-05,
+      "loss": 0.2514,
+      "step": 2374
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8251362601351317,
+      "learning_rate": 1.957097069191089e-05,
+      "loss": 0.2104,
+      "step": 2375
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0358611536294893,
+      "learning_rate": 1.9570493324136344e-05,
+      "loss": 0.2304,
+      "step": 2376
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9166959393585152,
+      "learning_rate": 1.9570015696761623e-05,
+      "loss": 0.2572,
+      "step": 2377
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0968226275193573,
+      "learning_rate": 1.9569537809799687e-05,
+      "loss": 0.2106,
+      "step": 2378
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.284618905986818,
+      "learning_rate": 1.9569059663263498e-05,
+      "loss": 0.218,
+      "step": 2379
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.07367940810242,
+      "learning_rate": 1.9568581257166025e-05,
+      "loss": 0.2238,
+      "step": 2380
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9241633450880841,
+      "learning_rate": 1.9568102591520246e-05,
+      "loss": 0.2443,
+      "step": 2381
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8693666074144997,
+      "learning_rate": 1.956762366633914e-05,
+      "loss": 0.2408,
+      "step": 2382
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.152812884267505,
+      "learning_rate": 1.956714448163571e-05,
+      "loss": 0.2606,
+      "step": 2383
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0120887204947118,
+      "learning_rate": 1.9566665037422937e-05,
+      "loss": 0.2273,
+      "step": 2384
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8850621950766456,
+      "learning_rate": 1.9566185333713835e-05,
+      "loss": 0.2189,
+      "step": 2385
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0453797190650638,
+      "learning_rate": 1.956570537052142e-05,
+      "loss": 0.2307,
+      "step": 2386
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.966999828439038,
+      "learning_rate": 1.9565225147858704e-05,
+      "loss": 0.2481,
+      "step": 2387
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.795290828022784,
+      "learning_rate": 1.9564744665738714e-05,
+      "loss": 0.2282,
+      "step": 2388
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9236235449239979,
+      "learning_rate": 1.9564263924174488e-05,
+      "loss": 0.2359,
+      "step": 2389
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.987110655783361,
+      "learning_rate": 1.9563782923179063e-05,
+      "loss": 0.2457,
+      "step": 2390
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9355041913405805,
+      "learning_rate": 1.9563301662765482e-05,
+      "loss": 0.225,
+      "step": 2391
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8777824597160528,
+      "learning_rate": 1.9562820142946808e-05,
+      "loss": 0.2172,
+      "step": 2392
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9367037353854338,
+      "learning_rate": 1.9562338363736095e-05,
+      "loss": 0.2321,
+      "step": 2393
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6404557075320965,
+      "learning_rate": 1.9561856325146414e-05,
+      "loss": 0.2112,
+      "step": 2394
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0918037055500003,
+      "learning_rate": 1.956137402719084e-05,
+      "loss": 0.226,
+      "step": 2395
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0405002814531292,
+      "learning_rate": 1.9560891469882457e-05,
+      "loss": 0.2411,
+      "step": 2396
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0119293437402799,
+      "learning_rate": 1.9560408653234352e-05,
+      "loss": 0.2338,
+      "step": 2397
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0379855418069062,
+      "learning_rate": 1.9559925577259622e-05,
+      "loss": 0.2481,
+      "step": 2398
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1508134685563278,
+      "learning_rate": 1.9559442241971373e-05,
+      "loss": 0.2326,
+      "step": 2399
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8364922526982728,
+      "learning_rate": 1.955895864738271e-05,
+      "loss": 0.2106,
+      "step": 2400
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0375308864937787,
+      "learning_rate": 1.955847479350675e-05,
+      "loss": 0.2316,
+      "step": 2401
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0535751080622453,
+      "learning_rate": 1.955799068035663e-05,
+      "loss": 0.2389,
+      "step": 2402
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.928196115611736,
+      "learning_rate": 1.955750630794547e-05,
+      "loss": 0.231,
+      "step": 2403
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9919358376481566,
+      "learning_rate": 1.955702167628641e-05,
+      "loss": 0.2104,
+      "step": 2404
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0893302608613895,
+      "learning_rate": 1.9556536785392598e-05,
+      "loss": 0.2239,
+      "step": 2405
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5843172922704782,
+      "learning_rate": 1.9556051635277184e-05,
+      "loss": 0.2484,
+      "step": 2406
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.188746173343023,
+      "learning_rate": 1.9555566225953333e-05,
+      "loss": 0.2239,
+      "step": 2407
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8501232211498138,
+      "learning_rate": 1.9555080557434206e-05,
+      "loss": 0.2239,
+      "step": 2408
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6514133365172499,
+      "learning_rate": 1.955459462973298e-05,
+      "loss": 0.2575,
+      "step": 2409
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8867170258575549,
+      "learning_rate": 1.9554108442862836e-05,
+      "loss": 0.2368,
+      "step": 2410
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1563166041868196,
+      "learning_rate": 1.955362199683696e-05,
+      "loss": 0.2319,
+      "step": 2411
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0302301616491012,
+      "learning_rate": 1.9553135291668548e-05,
+      "loss": 0.2467,
+      "step": 2412
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1775203268994718,
+      "learning_rate": 1.95526483273708e-05,
+      "loss": 0.248,
+      "step": 2413
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2472244643540031,
+      "learning_rate": 1.9552161103956932e-05,
+      "loss": 0.2336,
+      "step": 2414
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9942684568484321,
+      "learning_rate": 1.955167362144015e-05,
+      "loss": 0.2312,
+      "step": 2415
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2827573628017537,
+      "learning_rate": 1.955118587983368e-05,
+      "loss": 0.2337,
+      "step": 2416
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0779477429174507,
+      "learning_rate": 1.9550697879150757e-05,
+      "loss": 0.2528,
+      "step": 2417
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9287186494476223,
+      "learning_rate": 1.9550209619404616e-05,
+      "loss": 0.2146,
+      "step": 2418
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9780419354412737,
+      "learning_rate": 1.9549721100608494e-05,
+      "loss": 0.2442,
+      "step": 2419
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0299340666660326,
+      "learning_rate": 1.954923232277565e-05,
+      "loss": 0.2229,
+      "step": 2420
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9339979389685626,
+      "learning_rate": 1.954874328591934e-05,
+      "loss": 0.2441,
+      "step": 2421
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9663170606947568,
+      "learning_rate": 1.9548253990052833e-05,
+      "loss": 0.2303,
+      "step": 2422
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0915466600904569,
+      "learning_rate": 1.9547764435189395e-05,
+      "loss": 0.2477,
+      "step": 2423
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.575099114186478,
+      "learning_rate": 1.9547274621342303e-05,
+      "loss": 0.2472,
+      "step": 2424
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9211825722405153,
+      "learning_rate": 1.9546784548524852e-05,
+      "loss": 0.2182,
+      "step": 2425
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6693874263644388,
+      "learning_rate": 1.954629421675033e-05,
+      "loss": 0.2339,
+      "step": 2426
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.560856670186122,
+      "learning_rate": 1.954580362603204e-05,
+      "loss": 0.244,
+      "step": 2427
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3346896255835428,
+      "learning_rate": 1.954531277638328e-05,
+      "loss": 0.2254,
+      "step": 2428
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9549915898367675,
+      "learning_rate": 1.954482166781738e-05,
+      "loss": 0.2545,
+      "step": 2429
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1193417924601143,
+      "learning_rate": 1.9544330300347655e-05,
+      "loss": 0.2506,
+      "step": 2430
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1071622856945116,
+      "learning_rate": 1.9543838673987424e-05,
+      "loss": 0.2501,
+      "step": 2431
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0377912267602951,
+      "learning_rate": 1.9543346788750032e-05,
+      "loss": 0.2332,
+      "step": 2432
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.880835359230946,
+      "learning_rate": 1.9542854644648824e-05,
+      "loss": 0.2402,
+      "step": 2433
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.934264549762433,
+      "learning_rate": 1.954236224169714e-05,
+      "loss": 0.2234,
+      "step": 2434
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.946717586900411,
+      "learning_rate": 1.9541869579908343e-05,
+      "loss": 0.2546,
+      "step": 2435
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0749355807516685,
+      "learning_rate": 1.9541376659295796e-05,
+      "loss": 0.2423,
+      "step": 2436
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0938551679100832,
+      "learning_rate": 1.9540883479872863e-05,
+      "loss": 0.2355,
+      "step": 2437
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9996299987531758,
+      "learning_rate": 1.954039004165293e-05,
+      "loss": 0.2184,
+      "step": 2438
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9625907051201483,
+      "learning_rate": 1.953989634464938e-05,
+      "loss": 0.2111,
+      "step": 2439
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9399174909270136,
+      "learning_rate": 1.9539402388875598e-05,
+      "loss": 0.2269,
+      "step": 2440
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.119626479040853,
+      "learning_rate": 1.9538908174344994e-05,
+      "loss": 0.2429,
+      "step": 2441
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0974321999890257,
+      "learning_rate": 1.9538413701070964e-05,
+      "loss": 0.2383,
+      "step": 2442
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2114085536835701,
+      "learning_rate": 1.9537918969066923e-05,
+      "loss": 0.2101,
+      "step": 2443
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6206815067868514,
+      "learning_rate": 1.953742397834629e-05,
+      "loss": 0.2238,
+      "step": 2444
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1385008062544557,
+      "learning_rate": 1.9536928728922496e-05,
+      "loss": 0.1996,
+      "step": 2445
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9357832841400486,
+      "learning_rate": 1.953643322080897e-05,
+      "loss": 0.2372,
+      "step": 2446
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9088175364992762,
+      "learning_rate": 1.9535937454019155e-05,
+      "loss": 0.2301,
+      "step": 2447
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0977441998849506,
+      "learning_rate": 1.9535441428566496e-05,
+      "loss": 0.2296,
+      "step": 2448
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0568730184364237,
+      "learning_rate": 1.9534945144464452e-05,
+      "loss": 0.2048,
+      "step": 2449
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7694547827815719,
+      "learning_rate": 1.953444860172648e-05,
+      "loss": 0.2341,
+      "step": 2450
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9535818827060624,
+      "learning_rate": 1.9533951800366052e-05,
+      "loss": 0.2231,
+      "step": 2451
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0852235238905874,
+      "learning_rate": 1.9533454740396645e-05,
+      "loss": 0.2207,
+      "step": 2452
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2273715146657684,
+      "learning_rate": 1.953295742183174e-05,
+      "loss": 0.2329,
+      "step": 2453
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0468236296362416,
+      "learning_rate": 1.9532459844684824e-05,
+      "loss": 0.2469,
+      "step": 2454
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.888539113561735,
+      "learning_rate": 1.9531962008969396e-05,
+      "loss": 0.213,
+      "step": 2455
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.901360077521206,
+      "learning_rate": 1.953146391469896e-05,
+      "loss": 0.2335,
+      "step": 2456
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0327000960606898,
+      "learning_rate": 1.953096556188703e-05,
+      "loss": 0.2289,
+      "step": 2457
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9485176494343361,
+      "learning_rate": 1.9530466950547118e-05,
+      "loss": 0.2263,
+      "step": 2458
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1105931610889015,
+      "learning_rate": 1.9529968080692753e-05,
+      "loss": 0.2316,
+      "step": 2459
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9852744012552714,
+      "learning_rate": 1.9529468952337468e-05,
+      "loss": 0.2566,
+      "step": 2460
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.037940345459345,
+      "learning_rate": 1.9528969565494792e-05,
+      "loss": 0.2414,
+      "step": 2461
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.376993818426987,
+      "learning_rate": 1.9528469920178287e-05,
+      "loss": 0.2499,
+      "step": 2462
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8866653946404849,
+      "learning_rate": 1.9527970016401493e-05,
+      "loss": 0.2225,
+      "step": 2463
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0459832052160578,
+      "learning_rate": 1.9527469854177973e-05,
+      "loss": 0.2262,
+      "step": 2464
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9970960002026228,
+      "learning_rate": 1.9526969433521298e-05,
+      "loss": 0.2127,
+      "step": 2465
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.031139726816588,
+      "learning_rate": 1.9526468754445035e-05,
+      "loss": 0.239,
+      "step": 2466
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9082242265666298,
+      "learning_rate": 1.9525967816962775e-05,
+      "loss": 0.2135,
+      "step": 2467
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0929122867480856,
+      "learning_rate": 1.9525466621088093e-05,
+      "loss": 0.2361,
+      "step": 2468
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.843789342979408,
+      "learning_rate": 1.95249651668346e-05,
+      "loss": 0.2289,
+      "step": 2469
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3704990673651014,
+      "learning_rate": 1.952446345421588e-05,
+      "loss": 0.254,
+      "step": 2470
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.861102949292528,
+      "learning_rate": 1.9523961483245552e-05,
+      "loss": 0.2082,
+      "step": 2471
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9862142239764253,
+      "learning_rate": 1.9523459253937233e-05,
+      "loss": 0.2239,
+      "step": 2472
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9025994750433144,
+      "learning_rate": 1.9522956766304543e-05,
+      "loss": 0.2438,
+      "step": 2473
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0347478039910887,
+      "learning_rate": 1.9522454020361116e-05,
+      "loss": 0.2403,
+      "step": 2474
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.268028229249795,
+      "learning_rate": 1.9521951016120582e-05,
+      "loss": 0.2541,
+      "step": 2475
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9898697278724068,
+      "learning_rate": 1.952144775359659e-05,
+      "loss": 0.2636,
+      "step": 2476
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9029126837260496,
+      "learning_rate": 1.9520944232802793e-05,
+      "loss": 0.2285,
+      "step": 2477
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4451648613384407,
+      "learning_rate": 1.9520440453752842e-05,
+      "loss": 0.2512,
+      "step": 2478
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3173582760945324,
+      "learning_rate": 1.951993641646041e-05,
+      "loss": 0.2275,
+      "step": 2479
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1085719825276406,
+      "learning_rate": 1.951943212093916e-05,
+      "loss": 0.2278,
+      "step": 2480
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0384753106293034,
+      "learning_rate": 1.951892756720278e-05,
+      "loss": 0.2648,
+      "step": 2481
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.100632928939612,
+      "learning_rate": 1.9518422755264947e-05,
+      "loss": 0.2049,
+      "step": 2482
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1089547410194291,
+      "learning_rate": 1.9517917685139365e-05,
+      "loss": 0.2381,
+      "step": 2483
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1286418385854793,
+      "learning_rate": 1.9517412356839727e-05,
+      "loss": 0.2673,
+      "step": 2484
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7957571895575681,
+      "learning_rate": 1.951690677037974e-05,
+      "loss": 0.2002,
+      "step": 2485
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2857267543737927,
+      "learning_rate": 1.9516400925773118e-05,
+      "loss": 0.2252,
+      "step": 2486
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9023184796103526,
+      "learning_rate": 1.9515894823033584e-05,
+      "loss": 0.2185,
+      "step": 2487
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1830975982529575,
+      "learning_rate": 1.9515388462174868e-05,
+      "loss": 0.261,
+      "step": 2488
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9656013811551463,
+      "learning_rate": 1.95148818432107e-05,
+      "loss": 0.2556,
+      "step": 2489
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9506680880228655,
+      "learning_rate": 1.9514374966154826e-05,
+      "loss": 0.2196,
+      "step": 2490
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0597012016818077,
+      "learning_rate": 1.951386783102099e-05,
+      "loss": 0.2247,
+      "step": 2491
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9004544786269002,
+      "learning_rate": 1.9513360437822957e-05,
+      "loss": 0.2284,
+      "step": 2492
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0277177890445819,
+      "learning_rate": 1.9512852786574483e-05,
+      "loss": 0.2368,
+      "step": 2493
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8764913099954449,
+      "learning_rate": 1.951234487728934e-05,
+      "loss": 0.2349,
+      "step": 2494
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0231867249114779,
+      "learning_rate": 1.9511836709981306e-05,
+      "loss": 0.2336,
+      "step": 2495
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1317906311257189,
+      "learning_rate": 1.951132828466416e-05,
+      "loss": 0.2067,
+      "step": 2496
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0525609819493371,
+      "learning_rate": 1.95108196013517e-05,
+      "loss": 0.2466,
+      "step": 2497
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3414958850247156,
+      "learning_rate": 1.951031066005772e-05,
+      "loss": 0.2371,
+      "step": 2498
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9807391428552752,
+      "learning_rate": 1.9509801460796027e-05,
+      "loss": 0.2522,
+      "step": 2499
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1634163941213589,
+      "learning_rate": 1.950929200358043e-05,
+      "loss": 0.2359,
+      "step": 2500
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.858070387539989,
+      "learning_rate": 1.9508782288424754e-05,
+      "loss": 0.2278,
+      "step": 2501
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.968181824391059,
+      "learning_rate": 1.950827231534282e-05,
+      "loss": 0.2317,
+      "step": 2502
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1859112241043284,
+      "learning_rate": 1.950776208434846e-05,
+      "loss": 0.2141,
+      "step": 2503
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1055241908293456,
+      "learning_rate": 1.9507251595455524e-05,
+      "loss": 0.233,
+      "step": 2504
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1720307433593404,
+      "learning_rate": 1.9506740848677845e-05,
+      "loss": 0.2295,
+      "step": 2505
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9640362423866915,
+      "learning_rate": 1.9506229844029283e-05,
+      "loss": 0.2212,
+      "step": 2506
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1471940629954807,
+      "learning_rate": 1.95057185815237e-05,
+      "loss": 0.2176,
+      "step": 2507
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2772531433878958,
+      "learning_rate": 1.9505207061174966e-05,
+      "loss": 0.245,
+      "step": 2508
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.024101923763926,
+      "learning_rate": 1.9504695282996953e-05,
+      "loss": 0.2446,
+      "step": 2509
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1602529558305126,
+      "learning_rate": 1.9504183247003544e-05,
+      "loss": 0.2296,
+      "step": 2510
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1835773861670562,
+      "learning_rate": 1.9503670953208628e-05,
+      "loss": 0.2329,
+      "step": 2511
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9684880160608303,
+      "learning_rate": 1.9503158401626098e-05,
+      "loss": 0.2482,
+      "step": 2512
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3001737299396119,
+      "learning_rate": 1.950264559226986e-05,
+      "loss": 0.2391,
+      "step": 2513
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.056275760297049,
+      "learning_rate": 1.9502132525153826e-05,
+      "loss": 0.2292,
+      "step": 2514
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8285779120395363,
+      "learning_rate": 1.950161920029191e-05,
+      "loss": 0.2096,
+      "step": 2515
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2265740847100444,
+      "learning_rate": 1.9501105617698034e-05,
+      "loss": 0.2525,
+      "step": 2516
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.168129362633706,
+      "learning_rate": 1.9500591777386134e-05,
+      "loss": 0.214,
+      "step": 2517
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3091587397759643,
+      "learning_rate": 1.9500077679370145e-05,
+      "loss": 0.265,
+      "step": 2518
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1204506664591425,
+      "learning_rate": 1.949956332366401e-05,
+      "loss": 0.2612,
+      "step": 2519
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.064079571330691,
+      "learning_rate": 1.9499048710281686e-05,
+      "loss": 0.2295,
+      "step": 2520
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1114983928013271,
+      "learning_rate": 1.949853383923713e-05,
+      "loss": 0.2359,
+      "step": 2521
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1622974627712408,
+      "learning_rate": 1.9498018710544306e-05,
+      "loss": 0.2127,
+      "step": 2522
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2721738857153555,
+      "learning_rate": 1.9497503324217188e-05,
+      "loss": 0.2444,
+      "step": 2523
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2230204085056002,
+      "learning_rate": 1.9496987680269755e-05,
+      "loss": 0.2441,
+      "step": 2524
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1287902286094293,
+      "learning_rate": 1.9496471778715996e-05,
+      "loss": 0.2071,
+      "step": 2525
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0371896187150622,
+      "learning_rate": 1.94959556195699e-05,
+      "loss": 0.2307,
+      "step": 2526
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.019653166099954,
+      "learning_rate": 1.9495439202845478e-05,
+      "loss": 0.2605,
+      "step": 2527
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2402110275780256,
+      "learning_rate": 1.9494922528556727e-05,
+      "loss": 0.2256,
+      "step": 2528
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9884338951179454,
+      "learning_rate": 1.9494405596717664e-05,
+      "loss": 0.2233,
+      "step": 2529
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0533813644665326,
+      "learning_rate": 1.949388840734232e-05,
+      "loss": 0.2037,
+      "step": 2530
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.967803989425464,
+      "learning_rate": 1.949337096044471e-05,
+      "loss": 0.2355,
+      "step": 2531
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1704508502228437,
+      "learning_rate": 1.949285325603888e-05,
+      "loss": 0.2442,
+      "step": 2532
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8687769631305765,
+      "learning_rate": 1.9492335294138868e-05,
+      "loss": 0.2589,
+      "step": 2533
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0410902336408165,
+      "learning_rate": 1.9491817074758727e-05,
+      "loss": 0.2424,
+      "step": 2534
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0019116803639123,
+      "learning_rate": 1.949129859791251e-05,
+      "loss": 0.2116,
+      "step": 2535
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0089129827853796,
+      "learning_rate": 1.9490779863614284e-05,
+      "loss": 0.2288,
+      "step": 2536
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9476886305903411,
+      "learning_rate": 1.9490260871878114e-05,
+      "loss": 0.2283,
+      "step": 2537
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7249136953545823,
+      "learning_rate": 1.9489741622718087e-05,
+      "loss": 0.2329,
+      "step": 2538
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6606994488067524,
+      "learning_rate": 1.9489222116148278e-05,
+      "loss": 0.245,
+      "step": 2539
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1349112515004496,
+      "learning_rate": 1.948870235218279e-05,
+      "loss": 0.217,
+      "step": 2540
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0864289977825945,
+      "learning_rate": 1.9488182330835706e-05,
+      "loss": 0.249,
+      "step": 2541
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0162834015559579,
+      "learning_rate": 1.9487662052121145e-05,
+      "loss": 0.2613,
+      "step": 2542
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3802411208694967,
+      "learning_rate": 1.9487141516053214e-05,
+      "loss": 0.2406,
+      "step": 2543
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3959880188571363,
+      "learning_rate": 1.9486620722646036e-05,
+      "loss": 0.2466,
+      "step": 2544
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9696196695677537,
+      "learning_rate": 1.948609967191373e-05,
+      "loss": 0.2519,
+      "step": 2545
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1169183926233577,
+      "learning_rate": 1.9485578363870438e-05,
+      "loss": 0.2221,
+      "step": 2546
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9685353962725984,
+      "learning_rate": 1.9485056798530296e-05,
+      "loss": 0.2147,
+      "step": 2547
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3676253959318463,
+      "learning_rate": 1.9484534975907454e-05,
+      "loss": 0.2331,
+      "step": 2548
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.072106777885036,
+      "learning_rate": 1.9484012896016064e-05,
+      "loss": 0.2405,
+      "step": 2549
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9564555833352242,
+      "learning_rate": 1.948349055887029e-05,
+      "loss": 0.2367,
+      "step": 2550
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1139384170444573,
+      "learning_rate": 1.9482967964484297e-05,
+      "loss": 0.2386,
+      "step": 2551
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8949825457977273,
+      "learning_rate": 1.9482445112872265e-05,
+      "loss": 0.2142,
+      "step": 2552
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9924360109156986,
+      "learning_rate": 1.948192200404837e-05,
+      "loss": 0.2308,
+      "step": 2553
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.213906686434998,
+      "learning_rate": 1.948139863802681e-05,
+      "loss": 0.2162,
+      "step": 2554
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.245261071851514,
+      "learning_rate": 1.9480875014821776e-05,
+      "loss": 0.2611,
+      "step": 2555
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9617311900940123,
+      "learning_rate": 1.9480351134447466e-05,
+      "loss": 0.2092,
+      "step": 2556
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8336642002205943,
+      "learning_rate": 1.94798269969181e-05,
+      "loss": 0.2081,
+      "step": 2557
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2352999776362503,
+      "learning_rate": 1.947930260224789e-05,
+      "loss": 0.2334,
+      "step": 2558
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2133376413858712,
+      "learning_rate": 1.9478777950451063e-05,
+      "loss": 0.2486,
+      "step": 2559
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8660840115714117,
+      "learning_rate": 1.9478253041541848e-05,
+      "loss": 0.2115,
+      "step": 2560
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0302407585810627,
+      "learning_rate": 1.9477727875534483e-05,
+      "loss": 0.238,
+      "step": 2561
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.998569291011515,
+      "learning_rate": 1.9477202452443217e-05,
+      "loss": 0.226,
+      "step": 2562
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2341549907489868,
+      "learning_rate": 1.9476676772282297e-05,
+      "loss": 0.2366,
+      "step": 2563
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9984474900319265,
+      "learning_rate": 1.9476150835065983e-05,
+      "loss": 0.2373,
+      "step": 2564
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8639329130273585,
+      "learning_rate": 1.9475624640808542e-05,
+      "loss": 0.2377,
+      "step": 2565
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8980352825774708,
+      "learning_rate": 1.9475098189524253e-05,
+      "loss": 0.2344,
+      "step": 2566
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0722782379588431,
+      "learning_rate": 1.9474571481227385e-05,
+      "loss": 0.2404,
+      "step": 2567
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0137257435535263,
+      "learning_rate": 1.947404451593223e-05,
+      "loss": 0.2399,
+      "step": 2568
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.063657778639185,
+      "learning_rate": 1.9473517293653084e-05,
+      "loss": 0.2402,
+      "step": 2569
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9111470386340367,
+      "learning_rate": 1.947298981440425e-05,
+      "loss": 0.2421,
+      "step": 2570
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0021174540309534,
+      "learning_rate": 1.947246207820003e-05,
+      "loss": 0.2228,
+      "step": 2571
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1306991183837498,
+      "learning_rate": 1.947193408505474e-05,
+      "loss": 0.2297,
+      "step": 2572
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9252103940645041,
+      "learning_rate": 1.9471405834982702e-05,
+      "loss": 0.2478,
+      "step": 2573
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3628088679806831,
+      "learning_rate": 1.947087732799825e-05,
+      "loss": 0.243,
+      "step": 2574
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9693549537628366,
+      "learning_rate": 1.947034856411571e-05,
+      "loss": 0.2315,
+      "step": 2575
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9421192104463941,
+      "learning_rate": 1.9469819543349433e-05,
+      "loss": 0.2135,
+      "step": 2576
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8127772980411968,
+      "learning_rate": 1.9469290265713767e-05,
+      "loss": 0.2272,
+      "step": 2577
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8785917986766272,
+      "learning_rate": 1.9468760731223065e-05,
+      "loss": 0.2169,
+      "step": 2578
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8235476076278131,
+      "learning_rate": 1.9468230939891695e-05,
+      "loss": 0.2154,
+      "step": 2579
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1338475304907736,
+      "learning_rate": 1.9467700891734027e-05,
+      "loss": 0.2246,
+      "step": 2580
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.088495536191225,
+      "learning_rate": 1.9467170586764436e-05,
+      "loss": 0.238,
+      "step": 2581
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.508331904771354,
+      "learning_rate": 1.946664002499731e-05,
+      "loss": 0.2506,
+      "step": 2582
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3740198237235974,
+      "learning_rate": 1.9466109206447036e-05,
+      "loss": 0.2363,
+      "step": 2583
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0443872256988116,
+      "learning_rate": 1.9465578131128017e-05,
+      "loss": 0.2265,
+      "step": 2584
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9073842275417658,
+      "learning_rate": 1.9465046799054657e-05,
+      "loss": 0.2197,
+      "step": 2585
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6196237127154511,
+      "learning_rate": 1.9464515210241368e-05,
+      "loss": 0.2237,
+      "step": 2586
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1295037581852974,
+      "learning_rate": 1.9463983364702567e-05,
+      "loss": 0.2357,
+      "step": 2587
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8859703033393711,
+      "learning_rate": 1.9463451262452685e-05,
+      "loss": 0.2127,
+      "step": 2588
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0225681338883512,
+      "learning_rate": 1.946291890350615e-05,
+      "loss": 0.2223,
+      "step": 2589
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9560884298829112,
+      "learning_rate": 1.946238628787741e-05,
+      "loss": 0.1999,
+      "step": 2590
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3680393661657546,
+      "learning_rate": 1.9461853415580902e-05,
+      "loss": 0.2471,
+      "step": 2591
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9219507253318284,
+      "learning_rate": 1.9461320286631088e-05,
+      "loss": 0.2394,
+      "step": 2592
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9532444945302694,
+      "learning_rate": 1.946078690104243e-05,
+      "loss": 0.231,
+      "step": 2593
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8964117763557078,
+      "learning_rate": 1.946025325882939e-05,
+      "loss": 0.2198,
+      "step": 2594
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9022765936891728,
+      "learning_rate": 1.945971936000645e-05,
+      "loss": 0.2021,
+      "step": 2595
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0377148811164694,
+      "learning_rate": 1.945918520458808e-05,
+      "loss": 0.2258,
+      "step": 2596
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8926644487193754,
+      "learning_rate": 1.9458650792588784e-05,
+      "loss": 0.2272,
+      "step": 2597
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2339397203295959,
+      "learning_rate": 1.945811612402305e-05,
+      "loss": 0.2071,
+      "step": 2598
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1447126510888062,
+      "learning_rate": 1.945758119890538e-05,
+      "loss": 0.2384,
+      "step": 2599
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.94796043499078,
+      "learning_rate": 1.9457046017250283e-05,
+      "loss": 0.2329,
+      "step": 2600
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7726117349573447,
+      "learning_rate": 1.9456510579072282e-05,
+      "loss": 0.2293,
+      "step": 2601
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1989169212832826,
+      "learning_rate": 1.94559748843859e-05,
+      "loss": 0.2463,
+      "step": 2602
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.074481968333032,
+      "learning_rate": 1.9455438933205662e-05,
+      "loss": 0.2264,
+      "step": 2603
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9435469677018979,
+      "learning_rate": 1.945490272554611e-05,
+      "loss": 0.2463,
+      "step": 2604
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1295581894773257,
+      "learning_rate": 1.9454366261421786e-05,
+      "loss": 0.2323,
+      "step": 2605
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8943028767051312,
+      "learning_rate": 1.9453829540847243e-05,
+      "loss": 0.2174,
+      "step": 2606
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2457900588078497,
+      "learning_rate": 1.9453292563837043e-05,
+      "loss": 0.2574,
+      "step": 2607
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8224583557257457,
+      "learning_rate": 1.9452755330405745e-05,
+      "loss": 0.2154,
+      "step": 2608
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0649357865229008,
+      "learning_rate": 1.9452217840567927e-05,
+      "loss": 0.2397,
+      "step": 2609
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.925172871095016,
+      "learning_rate": 1.9451680094338163e-05,
+      "loss": 0.2485,
+      "step": 2610
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3335832249417243,
+      "learning_rate": 1.9451142091731045e-05,
+      "loss": 0.2324,
+      "step": 2611
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9363316898690303,
+      "learning_rate": 1.9450603832761165e-05,
+      "loss": 0.2601,
+      "step": 2612
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0677907449571924,
+      "learning_rate": 1.945006531744312e-05,
+      "loss": 0.2292,
+      "step": 2613
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3032820403922316,
+      "learning_rate": 1.9449526545791523e-05,
+      "loss": 0.2523,
+      "step": 2614
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9162782130809316,
+      "learning_rate": 1.9448987517820982e-05,
+      "loss": 0.2226,
+      "step": 2615
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.025744133538097,
+      "learning_rate": 1.944844823354612e-05,
+      "loss": 0.2215,
+      "step": 2616
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9187446041207256,
+      "learning_rate": 1.944790869298157e-05,
+      "loss": 0.2469,
+      "step": 2617
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.869231477206988,
+      "learning_rate": 1.9447368896141958e-05,
+      "loss": 0.2336,
+      "step": 2618
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1103999027549547,
+      "learning_rate": 1.9446828843041933e-05,
+      "loss": 0.2437,
+      "step": 2619
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0309658164029158,
+      "learning_rate": 1.9446288533696145e-05,
+      "loss": 0.2207,
+      "step": 2620
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0021576454715904,
+      "learning_rate": 1.9445747968119246e-05,
+      "loss": 0.2232,
+      "step": 2621
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0926465258868414,
+      "learning_rate": 1.9445207146325894e-05,
+      "loss": 0.2556,
+      "step": 2622
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.67810028770524,
+      "learning_rate": 1.9444666068330772e-05,
+      "loss": 0.208,
+      "step": 2623
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8767727670984775,
+      "learning_rate": 1.9444124734148543e-05,
+      "loss": 0.226,
+      "step": 2624
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2002140113970021,
+      "learning_rate": 1.9443583143793904e-05,
+      "loss": 0.2337,
+      "step": 2625
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0739431790098095,
+      "learning_rate": 1.9443041297281536e-05,
+      "loss": 0.2426,
+      "step": 2626
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2736818974157589,
+      "learning_rate": 1.9442499194626138e-05,
+      "loss": 0.2387,
+      "step": 2627
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1784427170299878,
+      "learning_rate": 1.9441956835842416e-05,
+      "loss": 0.2219,
+      "step": 2628
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1776402200746279,
+      "learning_rate": 1.9441414220945083e-05,
+      "loss": 0.2268,
+      "step": 2629
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0166012698350608,
+      "learning_rate": 1.9440871349948856e-05,
+      "loss": 0.2448,
+      "step": 2630
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1629080894535453,
+      "learning_rate": 1.9440328222868457e-05,
+      "loss": 0.2375,
+      "step": 2631
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5773238347578813,
+      "learning_rate": 1.9439784839718627e-05,
+      "loss": 0.2214,
+      "step": 2632
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8701043666547783,
+      "learning_rate": 1.94392412005141e-05,
+      "loss": 0.2403,
+      "step": 2633
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9530883571583211,
+      "learning_rate": 1.943869730526962e-05,
+      "loss": 0.2313,
+      "step": 2634
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2905919957477667,
+      "learning_rate": 1.9438153153999942e-05,
+      "loss": 0.2338,
+      "step": 2635
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3297130904337844,
+      "learning_rate": 1.9437608746719828e-05,
+      "loss": 0.2568,
+      "step": 2636
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1409616790457267,
+      "learning_rate": 1.943706408344404e-05,
+      "loss": 0.2696,
+      "step": 2637
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.165365917351629,
+      "learning_rate": 1.9436519164187363e-05,
+      "loss": 0.2398,
+      "step": 2638
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6898833986979496,
+      "learning_rate": 1.9435973988964564e-05,
+      "loss": 0.2401,
+      "step": 2639
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6757623447359549,
+      "learning_rate": 1.943542855779044e-05,
+      "loss": 0.2353,
+      "step": 2640
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0909672887562558,
+      "learning_rate": 1.9434882870679783e-05,
+      "loss": 0.2225,
+      "step": 2641
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.286445863100236,
+      "learning_rate": 1.9434336927647397e-05,
+      "loss": 0.216,
+      "step": 2642
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1677363817710942,
+      "learning_rate": 1.9433790728708085e-05,
+      "loss": 0.2267,
+      "step": 2643
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0503125887814437,
+      "learning_rate": 1.943324427387667e-05,
+      "loss": 0.242,
+      "step": 2644
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.007415443963795,
+      "learning_rate": 1.9432697563167974e-05,
+      "loss": 0.2323,
+      "step": 2645
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3820066306983543,
+      "learning_rate": 1.9432150596596818e-05,
+      "loss": 0.2487,
+      "step": 2646
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0342232451505902,
+      "learning_rate": 1.9431603374178048e-05,
+      "loss": 0.2394,
+      "step": 2647
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5428237143556407,
+      "learning_rate": 1.94310558959265e-05,
+      "loss": 0.2356,
+      "step": 2648
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.67906044044975,
+      "learning_rate": 1.943050816185703e-05,
+      "loss": 0.2331,
+      "step": 2649
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0458855765510422,
+      "learning_rate": 1.9429960171984496e-05,
+      "loss": 0.2103,
+      "step": 2650
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9721893985186519,
+      "learning_rate": 1.9429411926323756e-05,
+      "loss": 0.2038,
+      "step": 2651
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0903666819111881,
+      "learning_rate": 1.942886342488969e-05,
+      "loss": 0.2179,
+      "step": 2652
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5080139281275566,
+      "learning_rate": 1.9428314667697166e-05,
+      "loss": 0.2317,
+      "step": 2653
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6690664079020918,
+      "learning_rate": 1.9427765654761078e-05,
+      "loss": 0.2292,
+      "step": 2654
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3396452269621446,
+      "learning_rate": 1.9427216386096313e-05,
+      "loss": 0.2412,
+      "step": 2655
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5807328497654407,
+      "learning_rate": 1.942666686171777e-05,
+      "loss": 0.2357,
+      "step": 2656
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.190131146360561,
+      "learning_rate": 1.9426117081640356e-05,
+      "loss": 0.2241,
+      "step": 2657
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9877783524716975,
+      "learning_rate": 1.9425567045878983e-05,
+      "loss": 0.2049,
+      "step": 2658
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6005142246281456,
+      "learning_rate": 1.942501675444857e-05,
+      "loss": 0.203,
+      "step": 2659
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7675592499022506,
+      "learning_rate": 1.942446620736405e-05,
+      "loss": 0.2501,
+      "step": 2660
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3275111289487056,
+      "learning_rate": 1.942391540464035e-05,
+      "loss": 0.2689,
+      "step": 2661
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0635503262207242,
+      "learning_rate": 1.942336434629241e-05,
+      "loss": 0.2241,
+      "step": 2662
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5034106101267861,
+      "learning_rate": 1.9422813032335183e-05,
+      "loss": 0.2364,
+      "step": 2663
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1863865222873722,
+      "learning_rate": 1.942226146278362e-05,
+      "loss": 0.2401,
+      "step": 2664
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2804977196215885,
+      "learning_rate": 1.9421709637652683e-05,
+      "loss": 0.2521,
+      "step": 2665
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3243453835764538,
+      "learning_rate": 1.9421157556957335e-05,
+      "loss": 0.2466,
+      "step": 2666
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.114670908193584,
+      "learning_rate": 1.9420605220712563e-05,
+      "loss": 0.2216,
+      "step": 2667
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9544414132223076,
+      "learning_rate": 1.942005262893334e-05,
+      "loss": 0.2341,
+      "step": 2668
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0708571897346173,
+      "learning_rate": 1.9419499781634655e-05,
+      "loss": 0.2332,
+      "step": 2669
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8900595168867483,
+      "learning_rate": 1.9418946678831507e-05,
+      "loss": 0.2251,
+      "step": 2670
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.129854489633797,
+      "learning_rate": 1.9418393320538898e-05,
+      "loss": 0.2257,
+      "step": 2671
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7338804200785907,
+      "learning_rate": 1.9417839706771842e-05,
+      "loss": 0.2239,
+      "step": 2672
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9958041310863106,
+      "learning_rate": 1.941728583754535e-05,
+      "loss": 0.2195,
+      "step": 2673
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3733645464339348,
+      "learning_rate": 1.9416731712874446e-05,
+      "loss": 0.2589,
+      "step": 2674
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9909583269793304,
+      "learning_rate": 1.9416177332774162e-05,
+      "loss": 0.2301,
+      "step": 2675
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0248646828945818,
+      "learning_rate": 1.941562269725954e-05,
+      "loss": 0.2144,
+      "step": 2676
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.008259417023293,
+      "learning_rate": 1.9415067806345618e-05,
+      "loss": 0.2314,
+      "step": 2677
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0543469075480103,
+      "learning_rate": 1.9414512660047447e-05,
+      "loss": 0.2312,
+      "step": 2678
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1781005557537256,
+      "learning_rate": 1.9413957258380096e-05,
+      "loss": 0.2802,
+      "step": 2679
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4001122745244066,
+      "learning_rate": 1.9413401601358616e-05,
+      "loss": 0.2215,
+      "step": 2680
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3713329500791462,
+      "learning_rate": 1.9412845688998088e-05,
+      "loss": 0.2685,
+      "step": 2681
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.886757575995524,
+      "learning_rate": 1.941228952131359e-05,
+      "loss": 0.223,
+      "step": 2682
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.22683645359689,
+      "learning_rate": 1.9411733098320206e-05,
+      "loss": 0.2672,
+      "step": 2683
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5315058602311509,
+      "learning_rate": 1.941117642003303e-05,
+      "loss": 0.2429,
+      "step": 2684
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1110857691169762,
+      "learning_rate": 1.9410619486467165e-05,
+      "loss": 0.2305,
+      "step": 2685
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.258428468415178,
+      "learning_rate": 1.941006229763771e-05,
+      "loss": 0.2534,
+      "step": 2686
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0637285182214002,
+      "learning_rate": 1.9409504853559785e-05,
+      "loss": 0.2362,
+      "step": 2687
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9076212147606959,
+      "learning_rate": 1.9408947154248513e-05,
+      "loss": 0.2272,
+      "step": 2688
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3770753242493052,
+      "learning_rate": 1.9408389199719014e-05,
+      "loss": 0.2113,
+      "step": 2689
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9711446101427936,
+      "learning_rate": 1.940783098998643e-05,
+      "loss": 0.1977,
+      "step": 2690
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.140062040259186,
+      "learning_rate": 1.9407272525065898e-05,
+      "loss": 0.2167,
+      "step": 2691
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1803908714893359,
+      "learning_rate": 1.9406713804972565e-05,
+      "loss": 0.2327,
+      "step": 2692
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8373370160514211,
+      "learning_rate": 1.940615482972159e-05,
+      "loss": 0.2331,
+      "step": 2693
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7754881930206292,
+      "learning_rate": 1.9405595599328135e-05,
+      "loss": 0.2258,
+      "step": 2694
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0043781465210622,
+      "learning_rate": 1.940503611380737e-05,
+      "loss": 0.2151,
+      "step": 2695
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8156833311388194,
+      "learning_rate": 1.9404476373174464e-05,
+      "loss": 0.2327,
+      "step": 2696
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0110757172867997,
+      "learning_rate": 1.940391637744461e-05,
+      "loss": 0.208,
+      "step": 2697
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.267672729778902,
+      "learning_rate": 1.9403356126632992e-05,
+      "loss": 0.2283,
+      "step": 2698
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0261197599948402,
+      "learning_rate": 1.9402795620754804e-05,
+      "loss": 0.2338,
+      "step": 2699
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.329068764323991,
+      "learning_rate": 1.9402234859825257e-05,
+      "loss": 0.2397,
+      "step": 2700
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1760870358990558,
+      "learning_rate": 1.940167384385956e-05,
+      "loss": 0.2614,
+      "step": 2701
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3730668241267234,
+      "learning_rate": 1.9401112572872925e-05,
+      "loss": 0.2195,
+      "step": 2702
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2577938917534763,
+      "learning_rate": 1.9400551046880585e-05,
+      "loss": 0.2342,
+      "step": 2703
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9975635494089065,
+      "learning_rate": 1.9399989265897764e-05,
+      "loss": 0.2163,
+      "step": 2704
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9901243552258541,
+      "learning_rate": 1.9399427229939704e-05,
+      "loss": 0.2293,
+      "step": 2705
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1380070535962723,
+      "learning_rate": 1.939886493902165e-05,
+      "loss": 0.2385,
+      "step": 2706
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8633683152623378,
+      "learning_rate": 1.9398302393158853e-05,
+      "loss": 0.214,
+      "step": 2707
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3079801081816878,
+      "learning_rate": 1.939773959236657e-05,
+      "loss": 0.246,
+      "step": 2708
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4099778762372743,
+      "learning_rate": 1.9397176536660074e-05,
+      "loss": 0.2494,
+      "step": 2709
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3922786096799684,
+      "learning_rate": 1.939661322605463e-05,
+      "loss": 0.2666,
+      "step": 2710
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2057784347406664,
+      "learning_rate": 1.9396049660565525e-05,
+      "loss": 0.2277,
+      "step": 2711
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6404060910870868,
+      "learning_rate": 1.939548584020804e-05,
+      "loss": 0.2288,
+      "step": 2712
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9016999206905085,
+      "learning_rate": 1.9394921764997475e-05,
+      "loss": 0.1847,
+      "step": 2713
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.162953170396771,
+      "learning_rate": 1.939435743494912e-05,
+      "loss": 0.2327,
+      "step": 2714
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1998203345890663,
+      "learning_rate": 1.9393792850078294e-05,
+      "loss": 0.2468,
+      "step": 2715
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3223511057603083,
+      "learning_rate": 1.9393228010400303e-05,
+      "loss": 0.2008,
+      "step": 2716
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2970323911231219,
+      "learning_rate": 1.9392662915930476e-05,
+      "loss": 0.252,
+      "step": 2717
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9859759934321387,
+      "learning_rate": 1.9392097566684132e-05,
+      "loss": 0.2326,
+      "step": 2718
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5788317329342427,
+      "learning_rate": 1.9391531962676614e-05,
+      "loss": 0.2219,
+      "step": 2719
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4030978767199909,
+      "learning_rate": 1.939096610392326e-05,
+      "loss": 0.2487,
+      "step": 2720
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0004712075775568,
+      "learning_rate": 1.939039999043942e-05,
+      "loss": 0.2151,
+      "step": 2721
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9521128417870931,
+      "learning_rate": 1.938983362224045e-05,
+      "loss": 0.2365,
+      "step": 2722
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8453050326579191,
+      "learning_rate": 1.9389266999341717e-05,
+      "loss": 0.228,
+      "step": 2723
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9700950773605047,
+      "learning_rate": 1.938870012175858e-05,
+      "loss": 0.2038,
+      "step": 2724
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3553509430700317,
+      "learning_rate": 1.9388132989506422e-05,
+      "loss": 0.2297,
+      "step": 2725
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9893932786413834,
+      "learning_rate": 1.938756560260063e-05,
+      "loss": 0.2173,
+      "step": 2726
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.219016122502431,
+      "learning_rate": 1.938699796105659e-05,
+      "loss": 0.2452,
+      "step": 2727
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1359466605415534,
+      "learning_rate": 1.93864300648897e-05,
+      "loss": 0.2274,
+      "step": 2728
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9213735652946478,
+      "learning_rate": 1.9385861914115365e-05,
+      "loss": 0.217,
+      "step": 2729
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0375854513096163,
+      "learning_rate": 1.9385293508748994e-05,
+      "loss": 0.2354,
+      "step": 2730
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1904506309150737,
+      "learning_rate": 1.9384724848806007e-05,
+      "loss": 0.2345,
+      "step": 2731
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0897334514227217,
+      "learning_rate": 1.938415593430183e-05,
+      "loss": 0.2282,
+      "step": 2732
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3949774336848564,
+      "learning_rate": 1.938358676525189e-05,
+      "loss": 0.2158,
+      "step": 2733
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2083550955177704,
+      "learning_rate": 1.938301734167163e-05,
+      "loss": 0.2311,
+      "step": 2734
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1414932076518072,
+      "learning_rate": 1.9382447663576495e-05,
+      "loss": 0.2166,
+      "step": 2735
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2718276826173287,
+      "learning_rate": 1.9381877730981938e-05,
+      "loss": 0.2304,
+      "step": 2736
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8383903990044212,
+      "learning_rate": 1.9381307543903416e-05,
+      "loss": 0.2392,
+      "step": 2737
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1569915189018203,
+      "learning_rate": 1.93807371023564e-05,
+      "loss": 0.2274,
+      "step": 2738
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4849041251151516,
+      "learning_rate": 1.9380166406356357e-05,
+      "loss": 0.2647,
+      "step": 2739
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9769699732654332,
+      "learning_rate": 1.9379595455918773e-05,
+      "loss": 0.2233,
+      "step": 2740
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1559082444358126,
+      "learning_rate": 1.937902425105913e-05,
+      "loss": 0.225,
+      "step": 2741
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9116840164315465,
+      "learning_rate": 1.9378452791792924e-05,
+      "loss": 0.2057,
+      "step": 2742
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0057407085044325,
+      "learning_rate": 1.937788107813566e-05,
+      "loss": 0.2108,
+      "step": 2743
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0099264312055576,
+      "learning_rate": 1.937730911010284e-05,
+      "loss": 0.2116,
+      "step": 2744
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4555501133886994,
+      "learning_rate": 1.9376736887709982e-05,
+      "loss": 0.216,
+      "step": 2745
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0396223279709482,
+      "learning_rate": 1.9376164410972604e-05,
+      "loss": 0.229,
+      "step": 2746
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1136116076973643,
+      "learning_rate": 1.9375591679906242e-05,
+      "loss": 0.2507,
+      "step": 2747
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8975991818619756,
+      "learning_rate": 1.937501869452642e-05,
+      "loss": 0.2231,
+      "step": 2748
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5743246579158048,
+      "learning_rate": 1.937444545484869e-05,
+      "loss": 0.2339,
+      "step": 2749
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.469957685828871,
+      "learning_rate": 1.9373871960888594e-05,
+      "loss": 0.2389,
+      "step": 2750
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0671056971589354,
+      "learning_rate": 1.9373298212661697e-05,
+      "loss": 0.2364,
+      "step": 2751
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.88066037520169,
+      "learning_rate": 1.9372724210183552e-05,
+      "loss": 0.2524,
+      "step": 2752
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.109280660148114,
+      "learning_rate": 1.9372149953469733e-05,
+      "loss": 0.2065,
+      "step": 2753
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4632192072371528,
+      "learning_rate": 1.937157544253582e-05,
+      "loss": 0.2408,
+      "step": 2754
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9689206571329406,
+      "learning_rate": 1.9371000677397393e-05,
+      "loss": 0.2171,
+      "step": 2755
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.250810549389868,
+      "learning_rate": 1.9370425658070043e-05,
+      "loss": 0.25,
+      "step": 2756
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9785658371917585,
+      "learning_rate": 1.936985038456937e-05,
+      "loss": 0.2174,
+      "step": 2757
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9513477898582572,
+      "learning_rate": 1.936927485691097e-05,
+      "loss": 0.2498,
+      "step": 2758
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9823602970822934,
+      "learning_rate": 1.9368699075110467e-05,
+      "loss": 0.2077,
+      "step": 2759
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.01832212753293,
+      "learning_rate": 1.9368123039183468e-05,
+      "loss": 0.2258,
+      "step": 2760
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0482144964607565,
+      "learning_rate": 1.9367546749145605e-05,
+      "loss": 0.2374,
+      "step": 2761
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1170462467104787,
+      "learning_rate": 1.9366970205012508e-05,
+      "loss": 0.218,
+      "step": 2762
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2372709983607768,
+      "learning_rate": 1.9366393406799813e-05,
+      "loss": 0.2195,
+      "step": 2763
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.606539485103394,
+      "learning_rate": 1.9365816354523167e-05,
+      "loss": 0.2416,
+      "step": 2764
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.08146758226909,
+      "learning_rate": 1.9365239048198227e-05,
+      "loss": 0.2265,
+      "step": 2765
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9354277640682458,
+      "learning_rate": 1.9364661487840645e-05,
+      "loss": 0.2565,
+      "step": 2766
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.194024150331995,
+      "learning_rate": 1.9364083673466094e-05,
+      "loss": 0.2313,
+      "step": 2767
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.208037331414121,
+      "learning_rate": 1.9363505605090243e-05,
+      "loss": 0.2331,
+      "step": 2768
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.654984133731084,
+      "learning_rate": 1.9362927282728774e-05,
+      "loss": 0.244,
+      "step": 2769
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1251978450679647,
+      "learning_rate": 1.9362348706397374e-05,
+      "loss": 0.2296,
+      "step": 2770
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.089003515036792,
+      "learning_rate": 1.9361769876111734e-05,
+      "loss": 0.2047,
+      "step": 2771
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0686045742352648,
+      "learning_rate": 1.936119079188756e-05,
+      "loss": 0.2195,
+      "step": 2772
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.912428797587506,
+      "learning_rate": 1.936061145374056e-05,
+      "loss": 0.2152,
+      "step": 2773
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2551911812214693,
+      "learning_rate": 1.936003186168644e-05,
+      "loss": 0.2467,
+      "step": 2774
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9804946097373898,
+      "learning_rate": 1.935945201574093e-05,
+      "loss": 0.2328,
+      "step": 2775
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2927497100162066,
+      "learning_rate": 1.9358871915919754e-05,
+      "loss": 0.2347,
+      "step": 2776
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2909319229029048,
+      "learning_rate": 1.935829156223865e-05,
+      "loss": 0.2266,
+      "step": 2777
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1159629090785874,
+      "learning_rate": 1.935771095471336e-05,
+      "loss": 0.2208,
+      "step": 2778
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3930226613031322,
+      "learning_rate": 1.935713009335963e-05,
+      "loss": 0.2331,
+      "step": 2779
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0452858269400926,
+      "learning_rate": 1.9356548978193216e-05,
+      "loss": 0.2501,
+      "step": 2780
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9458649340006015,
+      "learning_rate": 1.9355967609229886e-05,
+      "loss": 0.2338,
+      "step": 2781
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2348517530609575,
+      "learning_rate": 1.9355385986485406e-05,
+      "loss": 0.2253,
+      "step": 2782
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5834193611796006,
+      "learning_rate": 1.935480410997555e-05,
+      "loss": 0.2621,
+      "step": 2783
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8904407124477265,
+      "learning_rate": 1.9354221979716107e-05,
+      "loss": 0.2173,
+      "step": 2784
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9469089622240592,
+      "learning_rate": 1.9353639595722863e-05,
+      "loss": 0.2164,
+      "step": 2785
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7623120361802205,
+      "learning_rate": 1.9353056958011613e-05,
+      "loss": 0.2241,
+      "step": 2786
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.343534262397146,
+      "learning_rate": 1.935247406659817e-05,
+      "loss": 0.2292,
+      "step": 2787
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.9677492958050955,
+      "learning_rate": 1.935189092149834e-05,
+      "loss": 0.237,
+      "step": 2788
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0513218979485803,
+      "learning_rate": 1.9351307522727936e-05,
+      "loss": 0.2477,
+      "step": 2789
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0806439607517602,
+      "learning_rate": 1.935072387030279e-05,
+      "loss": 0.2478,
+      "step": 2790
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0885677043023188,
+      "learning_rate": 1.9350139964238732e-05,
+      "loss": 0.2577,
+      "step": 2791
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9436811681196573,
+      "learning_rate": 1.9349555804551598e-05,
+      "loss": 0.2148,
+      "step": 2792
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4197577391194087,
+      "learning_rate": 1.9348971391257235e-05,
+      "loss": 0.2472,
+      "step": 2793
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0348878799009253,
+      "learning_rate": 1.9348386724371495e-05,
+      "loss": 0.2477,
+      "step": 2794
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1077434643852102,
+      "learning_rate": 1.9347801803910236e-05,
+      "loss": 0.2239,
+      "step": 2795
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3548339541610674,
+      "learning_rate": 1.9347216629889326e-05,
+      "loss": 0.224,
+      "step": 2796
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7495935883278109,
+      "learning_rate": 1.9346631202324638e-05,
+      "loss": 0.2304,
+      "step": 2797
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.101603564197832,
+      "learning_rate": 1.9346045521232048e-05,
+      "loss": 0.2365,
+      "step": 2798
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1903832241877375,
+      "learning_rate": 1.9345459586627448e-05,
+      "loss": 0.2371,
+      "step": 2799
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1802158056744014,
+      "learning_rate": 1.9344873398526733e-05,
+      "loss": 0.2418,
+      "step": 2800
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.503865653521154,
+      "learning_rate": 1.934428695694579e-05,
+      "loss": 0.2316,
+      "step": 2801
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.801483119089678,
+      "learning_rate": 1.9343700261900543e-05,
+      "loss": 0.2447,
+      "step": 2802
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.688265735472681,
+      "learning_rate": 1.9343113313406893e-05,
+      "loss": 0.2316,
+      "step": 2803
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.522621184047923,
+      "learning_rate": 1.9342526111480772e-05,
+      "loss": 0.2275,
+      "step": 2804
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6548636105769403,
+      "learning_rate": 1.9341938656138097e-05,
+      "loss": 0.2378,
+      "step": 2805
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.149292127360877,
+      "learning_rate": 1.9341350947394812e-05,
+      "loss": 0.2148,
+      "step": 2806
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0514337747612676,
+      "learning_rate": 1.9340762985266853e-05,
+      "loss": 0.2271,
+      "step": 2807
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0551929436991787,
+      "learning_rate": 1.9340174769770173e-05,
+      "loss": 0.2237,
+      "step": 2808
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3379299345155393,
+      "learning_rate": 1.933958630092072e-05,
+      "loss": 0.2207,
+      "step": 2809
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5583520618210185,
+      "learning_rate": 1.9338997578734466e-05,
+      "loss": 0.2106,
+      "step": 2810
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4166387778203358,
+      "learning_rate": 1.9338408603227374e-05,
+      "loss": 0.2488,
+      "step": 2811
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2264068153220962,
+      "learning_rate": 1.9337819374415422e-05,
+      "loss": 0.2566,
+      "step": 2812
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1433870801860593,
+      "learning_rate": 1.933722989231459e-05,
+      "loss": 0.2185,
+      "step": 2813
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8592350172417055,
+      "learning_rate": 1.933664015694087e-05,
+      "loss": 0.2098,
+      "step": 2814
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8753480279923114,
+      "learning_rate": 1.933605016831026e-05,
+      "loss": 0.2096,
+      "step": 2815
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0621829098299027,
+      "learning_rate": 1.933545992643876e-05,
+      "loss": 0.2262,
+      "step": 2816
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.294220489663147,
+      "learning_rate": 1.9334869431342386e-05,
+      "loss": 0.236,
+      "step": 2817
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.387910567461303,
+      "learning_rate": 1.9334278683037145e-05,
+      "loss": 0.227,
+      "step": 2818
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.102746476766718,
+      "learning_rate": 1.9333687681539073e-05,
+      "loss": 0.2447,
+      "step": 2819
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8837216385130431,
+      "learning_rate": 1.9333096426864194e-05,
+      "loss": 0.2374,
+      "step": 2820
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0590666816622536,
+      "learning_rate": 1.9332504919028548e-05,
+      "loss": 0.2266,
+      "step": 2821
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2261448440402158,
+      "learning_rate": 1.9331913158048175e-05,
+      "loss": 0.2126,
+      "step": 2822
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3830268899911844,
+      "learning_rate": 1.9331321143939134e-05,
+      "loss": 0.1973,
+      "step": 2823
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2466959722982942,
+      "learning_rate": 1.933072887671748e-05,
+      "loss": 0.2186,
+      "step": 2824
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1776701511053869,
+      "learning_rate": 1.933013635639928e-05,
+      "loss": 0.2299,
+      "step": 2825
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.079444331175208,
+      "learning_rate": 1.93295435830006e-05,
+      "loss": 0.243,
+      "step": 2826
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9769968665319854,
+      "learning_rate": 1.9328950556537523e-05,
+      "loss": 0.2288,
+      "step": 2827
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.233898058764008,
+      "learning_rate": 1.932835727702614e-05,
+      "loss": 0.2327,
+      "step": 2828
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9367566360481667,
+      "learning_rate": 1.9327763744482536e-05,
+      "loss": 0.2117,
+      "step": 2829
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3818642138448198,
+      "learning_rate": 1.9327169958922813e-05,
+      "loss": 0.2181,
+      "step": 2830
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2256836948803158,
+      "learning_rate": 1.932657592036308e-05,
+      "loss": 0.2203,
+      "step": 2831
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.104918269812186,
+      "learning_rate": 1.9325981628819448e-05,
+      "loss": 0.2184,
+      "step": 2832
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4675592439172866,
+      "learning_rate": 1.9325387084308036e-05,
+      "loss": 0.2331,
+      "step": 2833
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.220652819519909,
+      "learning_rate": 1.9324792286844977e-05,
+      "loss": 0.2333,
+      "step": 2834
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2065214985533652,
+      "learning_rate": 1.9324197236446397e-05,
+      "loss": 0.2497,
+      "step": 2835
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0117364195372431,
+      "learning_rate": 1.932360193312844e-05,
+      "loss": 0.2159,
+      "step": 2836
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1569651541855233,
+      "learning_rate": 1.9323006376907253e-05,
+      "loss": 0.2282,
+      "step": 2837
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1876368415555707,
+      "learning_rate": 1.9322410567798996e-05,
+      "loss": 0.2212,
+      "step": 2838
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3740372088877062,
+      "learning_rate": 1.932181450581982e-05,
+      "loss": 0.2384,
+      "step": 2839
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1016021890316428,
+      "learning_rate": 1.9321218190985906e-05,
+      "loss": 0.222,
+      "step": 2840
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9683116783883369,
+      "learning_rate": 1.9320621623313416e-05,
+      "loss": 0.2128,
+      "step": 2841
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1316210342046384,
+      "learning_rate": 1.932002480281854e-05,
+      "loss": 0.2403,
+      "step": 2842
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9264831837106831,
+      "learning_rate": 1.9319427729517467e-05,
+      "loss": 0.2175,
+      "step": 2843
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.03768617926261,
+      "learning_rate": 1.9318830403426388e-05,
+      "loss": 0.2477,
+      "step": 2844
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2202169153802127,
+      "learning_rate": 1.9318232824561507e-05,
+      "loss": 0.2141,
+      "step": 2845
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8568755411257019,
+      "learning_rate": 1.9317634992939034e-05,
+      "loss": 0.2263,
+      "step": 2846
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0483047904842462,
+      "learning_rate": 1.931703690857519e-05,
+      "loss": 0.2382,
+      "step": 2847
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.329723964272318,
+      "learning_rate": 1.9316438571486188e-05,
+      "loss": 0.1943,
+      "step": 2848
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3214312045060064,
+      "learning_rate": 1.9315839981688267e-05,
+      "loss": 0.2255,
+      "step": 2849
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6453793440023385,
+      "learning_rate": 1.931524113919766e-05,
+      "loss": 0.2587,
+      "step": 2850
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2808218991790954,
+      "learning_rate": 1.931464204403061e-05,
+      "loss": 0.2426,
+      "step": 2851
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1321376470227988,
+      "learning_rate": 1.931404269620337e-05,
+      "loss": 0.2192,
+      "step": 2852
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8540458805915087,
+      "learning_rate": 1.9313443095732197e-05,
+      "loss": 0.2372,
+      "step": 2853
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.629983136852726,
+      "learning_rate": 1.9312843242633354e-05,
+      "loss": 0.2371,
+      "step": 2854
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9867139867736547,
+      "learning_rate": 1.931224313692311e-05,
+      "loss": 0.2338,
+      "step": 2855
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0737634184425395,
+      "learning_rate": 1.9311642778617742e-05,
+      "loss": 0.231,
+      "step": 2856
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.03320041635422,
+      "learning_rate": 1.931104216773354e-05,
+      "loss": 0.2224,
+      "step": 2857
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1224550744055295,
+      "learning_rate": 1.9310441304286794e-05,
+      "loss": 0.2247,
+      "step": 2858
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9518778508395871,
+      "learning_rate": 1.9309840188293803e-05,
+      "loss": 0.2337,
+      "step": 2859
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9511248974571589,
+      "learning_rate": 1.930923881977087e-05,
+      "loss": 0.2395,
+      "step": 2860
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1447901689296558,
+      "learning_rate": 1.9308637198734307e-05,
+      "loss": 0.22,
+      "step": 2861
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1499640522355206,
+      "learning_rate": 1.9308035325200436e-05,
+      "loss": 0.242,
+      "step": 2862
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3407711277898409,
+      "learning_rate": 1.9307433199185582e-05,
+      "loss": 0.2421,
+      "step": 2863
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0156304338906854,
+      "learning_rate": 1.9306830820706074e-05,
+      "loss": 0.2197,
+      "step": 2864
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.974378473587215,
+      "learning_rate": 1.9306228189778255e-05,
+      "loss": 0.2275,
+      "step": 2865
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0145603045608926,
+      "learning_rate": 1.930562530641847e-05,
+      "loss": 0.1958,
+      "step": 2866
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0589342389369067,
+      "learning_rate": 1.9305022170643077e-05,
+      "loss": 0.2191,
+      "step": 2867
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0494184592272218,
+      "learning_rate": 1.9304418782468427e-05,
+      "loss": 0.2211,
+      "step": 2868
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.826301537596647,
+      "learning_rate": 1.9303815141910894e-05,
+      "loss": 0.2075,
+      "step": 2869
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8674197245882275,
+      "learning_rate": 1.930321124898685e-05,
+      "loss": 0.2213,
+      "step": 2870
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.28917194083532,
+      "learning_rate": 1.930260710371268e-05,
+      "loss": 0.2432,
+      "step": 2871
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9229131007885853,
+      "learning_rate": 1.9302002706104762e-05,
+      "loss": 0.2072,
+      "step": 2872
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9223149186191164,
+      "learning_rate": 1.9301398056179493e-05,
+      "loss": 0.2302,
+      "step": 2873
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6633490926322958,
+      "learning_rate": 1.930079315395328e-05,
+      "loss": 0.2199,
+      "step": 2874
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9742814069706348,
+      "learning_rate": 1.930018799944253e-05,
+      "loss": 0.2316,
+      "step": 2875
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1254691217888606,
+      "learning_rate": 1.929958259266365e-05,
+      "loss": 0.2259,
+      "step": 2876
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8419765118945145,
+      "learning_rate": 1.9298976933633068e-05,
+      "loss": 0.2367,
+      "step": 2877
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0637810663668488,
+      "learning_rate": 1.929837102236721e-05,
+      "loss": 0.2029,
+      "step": 2878
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1969913921398376,
+      "learning_rate": 1.9297764858882516e-05,
+      "loss": 0.219,
+      "step": 2879
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9500843843503507,
+      "learning_rate": 1.929715844319542e-05,
+      "loss": 0.204,
+      "step": 2880
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9159934453202914,
+      "learning_rate": 1.9296551775322383e-05,
+      "loss": 0.2314,
+      "step": 2881
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9819107607663788,
+      "learning_rate": 1.9295944855279853e-05,
+      "loss": 0.2259,
+      "step": 2882
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0649590229898775,
+      "learning_rate": 1.9295337683084292e-05,
+      "loss": 0.1982,
+      "step": 2883
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2130419725924906,
+      "learning_rate": 1.929473025875217e-05,
+      "loss": 0.23,
+      "step": 2884
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9468141626402687,
+      "learning_rate": 1.9294122582299964e-05,
+      "loss": 0.2211,
+      "step": 2885
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0795576693617677,
+      "learning_rate": 1.929351465374416e-05,
+      "loss": 0.2469,
+      "step": 2886
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7962182832585174,
+      "learning_rate": 1.9292906473101246e-05,
+      "loss": 0.1996,
+      "step": 2887
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9714002340563536,
+      "learning_rate": 1.929229804038772e-05,
+      "loss": 0.2417,
+      "step": 2888
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.09410087899835,
+      "learning_rate": 1.9291689355620088e-05,
+      "loss": 0.2343,
+      "step": 2889
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7838595149854972,
+      "learning_rate": 1.9291080418814852e-05,
+      "loss": 0.1895,
+      "step": 2890
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.274523284116608,
+      "learning_rate": 1.9290471229988536e-05,
+      "loss": 0.2156,
+      "step": 2891
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9507973660783633,
+      "learning_rate": 1.9289861789157666e-05,
+      "loss": 0.237,
+      "step": 2892
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1444280596229304,
+      "learning_rate": 1.9289252096338767e-05,
+      "loss": 0.2345,
+      "step": 2893
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8911921538107122,
+      "learning_rate": 1.928864215154838e-05,
+      "loss": 0.2183,
+      "step": 2894
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9689915619745783,
+      "learning_rate": 1.928803195480305e-05,
+      "loss": 0.2347,
+      "step": 2895
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1021079330412795,
+      "learning_rate": 1.9287421506119332e-05,
+      "loss": 0.2305,
+      "step": 2896
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0599028799692833,
+      "learning_rate": 1.9286810805513774e-05,
+      "loss": 0.2447,
+      "step": 2897
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0162405743349308,
+      "learning_rate": 1.9286199853002956e-05,
+      "loss": 0.2046,
+      "step": 2898
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1035504592713803,
+      "learning_rate": 1.928558864860344e-05,
+      "loss": 0.2172,
+      "step": 2899
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.303054013613804,
+      "learning_rate": 1.9284977192331807e-05,
+      "loss": 0.2312,
+      "step": 2900
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2167740655890367,
+      "learning_rate": 1.9284365484204645e-05,
+      "loss": 0.2389,
+      "step": 2901
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3669197391194035,
+      "learning_rate": 1.9283753524238542e-05,
+      "loss": 0.2293,
+      "step": 2902
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1895888740056975,
+      "learning_rate": 1.92831413124501e-05,
+      "loss": 0.2241,
+      "step": 2903
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2522969542985063,
+      "learning_rate": 1.9282528848855925e-05,
+      "loss": 0.2601,
+      "step": 2904
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.302544121030843,
+      "learning_rate": 1.9281916133472636e-05,
+      "loss": 0.1939,
+      "step": 2905
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0861778850757935,
+      "learning_rate": 1.9281303166316846e-05,
+      "loss": 0.2684,
+      "step": 2906
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2882850775483696,
+      "learning_rate": 1.928068994740518e-05,
+      "loss": 0.2334,
+      "step": 2907
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2321169609007878,
+      "learning_rate": 1.9280076476754276e-05,
+      "loss": 0.2164,
+      "step": 2908
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6514277681396714,
+      "learning_rate": 1.9279462754380774e-05,
+      "loss": 0.2389,
+      "step": 2909
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1270106887925726,
+      "learning_rate": 1.927884878030132e-05,
+      "loss": 0.2225,
+      "step": 2910
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8445104404776635,
+      "learning_rate": 1.927823455453257e-05,
+      "loss": 0.2282,
+      "step": 2911
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1310180099501284,
+      "learning_rate": 1.9277620077091184e-05,
+      "loss": 0.2357,
+      "step": 2912
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1734832772791386,
+      "learning_rate": 1.927700534799383e-05,
+      "loss": 0.2436,
+      "step": 2913
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.055450159546421,
+      "learning_rate": 1.927639036725718e-05,
+      "loss": 0.2293,
+      "step": 2914
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.235235898756223,
+      "learning_rate": 1.927577513489792e-05,
+      "loss": 0.2186,
+      "step": 2915
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0584009839980613,
+      "learning_rate": 1.927515965093273e-05,
+      "loss": 0.2259,
+      "step": 2916
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0460374119343292,
+      "learning_rate": 1.9274543915378315e-05,
+      "loss": 0.2276,
+      "step": 2917
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.670885330919633,
+      "learning_rate": 1.927392792825137e-05,
+      "loss": 0.2071,
+      "step": 2918
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0249029467969002,
+      "learning_rate": 1.927331168956861e-05,
+      "loss": 0.2114,
+      "step": 2919
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9630012806325837,
+      "learning_rate": 1.9272695199346743e-05,
+      "loss": 0.2633,
+      "step": 2920
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0913292371471266,
+      "learning_rate": 1.92720784576025e-05,
+      "loss": 0.2235,
+      "step": 2921
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8716918819429021,
+      "learning_rate": 1.92714614643526e-05,
+      "loss": 0.2295,
+      "step": 2922
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1952869315038404,
+      "learning_rate": 1.9270844219613785e-05,
+      "loss": 0.2343,
+      "step": 2923
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9901067479541756,
+      "learning_rate": 1.9270226723402798e-05,
+      "loss": 0.2148,
+      "step": 2924
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9465534365045836,
+      "learning_rate": 1.926960897573639e-05,
+      "loss": 0.2189,
+      "step": 2925
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9050720063547557,
+      "learning_rate": 1.926899097663131e-05,
+      "loss": 0.205,
+      "step": 2926
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3984753431326877,
+      "learning_rate": 1.926837272610433e-05,
+      "loss": 0.2278,
+      "step": 2927
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.331648313954839,
+      "learning_rate": 1.9267754224172216e-05,
+      "loss": 0.2252,
+      "step": 2928
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3474581608803735,
+      "learning_rate": 1.926713547085174e-05,
+      "loss": 0.2242,
+      "step": 2929
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1764702056906664,
+      "learning_rate": 1.9266516466159697e-05,
+      "loss": 0.2275,
+      "step": 2930
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9102442675036209,
+      "learning_rate": 1.9265897210112868e-05,
+      "loss": 0.209,
+      "step": 2931
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9579038746913898,
+      "learning_rate": 1.9265277702728058e-05,
+      "loss": 0.2238,
+      "step": 2932
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.087611433992817,
+      "learning_rate": 1.9264657944022063e-05,
+      "loss": 0.2219,
+      "step": 2933
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9513234172795997,
+      "learning_rate": 1.92640379340117e-05,
+      "loss": 0.2293,
+      "step": 2934
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.035375087455176,
+      "learning_rate": 1.9263417672713786e-05,
+      "loss": 0.2013,
+      "step": 2935
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9867731544458642,
+      "learning_rate": 1.926279716014514e-05,
+      "loss": 0.2239,
+      "step": 2936
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.004772357407247,
+      "learning_rate": 1.92621763963226e-05,
+      "loss": 0.2428,
+      "step": 2937
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2468846047363114,
+      "learning_rate": 1.9261555381263003e-05,
+      "loss": 0.264,
+      "step": 2938
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3487744229263021,
+      "learning_rate": 1.926093411498319e-05,
+      "loss": 0.2162,
+      "step": 2939
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2469162410816559,
+      "learning_rate": 1.926031259750002e-05,
+      "loss": 0.2244,
+      "step": 2940
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1165749330292827,
+      "learning_rate": 1.9259690828830345e-05,
+      "loss": 0.2609,
+      "step": 2941
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.879048795841088,
+      "learning_rate": 1.925906880899104e-05,
+      "loss": 0.2225,
+      "step": 2942
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1138683990180442,
+      "learning_rate": 1.9258446537998964e-05,
+      "loss": 0.209,
+      "step": 2943
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.563574900308686,
+      "learning_rate": 1.9257824015871005e-05,
+      "loss": 0.2385,
+      "step": 2944
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8906959033595399,
+      "learning_rate": 1.9257201242624045e-05,
+      "loss": 0.2098,
+      "step": 2945
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.030668698944916,
+      "learning_rate": 1.925657821827498e-05,
+      "loss": 0.2218,
+      "step": 2946
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.096503438378198,
+      "learning_rate": 1.9255954942840706e-05,
+      "loss": 0.2421,
+      "step": 2947
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2449681975950357,
+      "learning_rate": 1.9255331416338134e-05,
+      "loss": 0.2259,
+      "step": 2948
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.067845526085949,
+      "learning_rate": 1.9254707638784174e-05,
+      "loss": 0.2051,
+      "step": 2949
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.645151745527325,
+      "learning_rate": 1.9254083610195745e-05,
+      "loss": 0.2271,
+      "step": 2950
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.440117632626847,
+      "learning_rate": 1.9253459330589776e-05,
+      "loss": 0.2204,
+      "step": 2951
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.294112671200197,
+      "learning_rate": 1.9252834799983197e-05,
+      "loss": 0.2179,
+      "step": 2952
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0002684462624296,
+      "learning_rate": 1.9252210018392957e-05,
+      "loss": 0.2398,
+      "step": 2953
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9685935114284663,
+      "learning_rate": 1.9251584985835996e-05,
+      "loss": 0.2193,
+      "step": 2954
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0508821576329554,
+      "learning_rate": 1.9250959702329268e-05,
+      "loss": 0.2246,
+      "step": 2955
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8610381375921086,
+      "learning_rate": 1.9250334167889737e-05,
+      "loss": 0.2236,
+      "step": 2956
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.260320103814766,
+      "learning_rate": 1.9249708382534372e-05,
+      "loss": 0.2235,
+      "step": 2957
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.070880397566758,
+      "learning_rate": 1.924908234628014e-05,
+      "loss": 0.2206,
+      "step": 2958
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8849634135287198,
+      "learning_rate": 1.9248456059144028e-05,
+      "loss": 0.2156,
+      "step": 2959
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8334624823803712,
+      "learning_rate": 1.9247829521143023e-05,
+      "loss": 0.2195,
+      "step": 2960
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9487501038077363,
+      "learning_rate": 1.924720273229412e-05,
+      "loss": 0.2473,
+      "step": 2961
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.294794933510296,
+      "learning_rate": 1.9246575692614323e-05,
+      "loss": 0.2442,
+      "step": 2962
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2417909542707162,
+      "learning_rate": 1.9245948402120634e-05,
+      "loss": 0.2306,
+      "step": 2963
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0910392732978642,
+      "learning_rate": 1.9245320860830075e-05,
+      "loss": 0.2041,
+      "step": 2964
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9316213997274377,
+      "learning_rate": 1.9244693068759668e-05,
+      "loss": 0.2398,
+      "step": 2965
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3031353937967904,
+      "learning_rate": 1.9244065025926434e-05,
+      "loss": 0.2096,
+      "step": 2966
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7483928647007118,
+      "learning_rate": 1.9243436732347418e-05,
+      "loss": 0.2105,
+      "step": 2967
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2147163505835121,
+      "learning_rate": 1.9242808188039658e-05,
+      "loss": 0.242,
+      "step": 2968
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9165976744588308,
+      "learning_rate": 1.92421793930202e-05,
+      "loss": 0.2287,
+      "step": 2969
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.129164002317354,
+      "learning_rate": 1.924155034730611e-05,
+      "loss": 0.2114,
+      "step": 2970
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8668190231085972,
+      "learning_rate": 1.924092105091444e-05,
+      "loss": 0.219,
+      "step": 2971
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8988161489801656,
+      "learning_rate": 1.9240291503862266e-05,
+      "loss": 0.2329,
+      "step": 2972
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5519469682120905,
+      "learning_rate": 1.9239661706166663e-05,
+      "loss": 0.2372,
+      "step": 2973
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9317913944468382,
+      "learning_rate": 1.9239031657844718e-05,
+      "loss": 0.2376,
+      "step": 2974
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0039810802977571,
+      "learning_rate": 1.9238401358913513e-05,
+      "loss": 0.2426,
+      "step": 2975
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8406615489980125,
+      "learning_rate": 1.923777080939015e-05,
+      "loss": 0.2081,
+      "step": 2976
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9565145582443978,
+      "learning_rate": 1.9237140009291733e-05,
+      "loss": 0.2146,
+      "step": 2977
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9123321433259106,
+      "learning_rate": 1.9236508958635372e-05,
+      "loss": 0.2132,
+      "step": 2978
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0533861024137494,
+      "learning_rate": 1.923587765743818e-05,
+      "loss": 0.1994,
+      "step": 2979
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2296651007941344,
+      "learning_rate": 1.923524610571729e-05,
+      "loss": 0.2353,
+      "step": 2980
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0783244560851948,
+      "learning_rate": 1.9234614303489823e-05,
+      "loss": 0.2487,
+      "step": 2981
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1787683750063183,
+      "learning_rate": 1.9233982250772927e-05,
+      "loss": 0.247,
+      "step": 2982
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0130160636525272,
+      "learning_rate": 1.9233349947583735e-05,
+      "loss": 0.2336,
+      "step": 2983
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4574101701417077,
+      "learning_rate": 1.923271739393941e-05,
+      "loss": 0.2048,
+      "step": 2984
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3197654603803377,
+      "learning_rate": 1.9232084589857103e-05,
+      "loss": 0.2364,
+      "step": 2985
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.207157479828292,
+      "learning_rate": 1.9231451535353977e-05,
+      "loss": 0.2134,
+      "step": 2986
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9671518118956477,
+      "learning_rate": 1.9230818230447207e-05,
+      "loss": 0.209,
+      "step": 2987
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.954083161294547,
+      "learning_rate": 1.9230184675153974e-05,
+      "loss": 0.2262,
+      "step": 2988
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0068100034817131,
+      "learning_rate": 1.9229550869491456e-05,
+      "loss": 0.2074,
+      "step": 2989
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2304033656255684,
+      "learning_rate": 1.9228916813476855e-05,
+      "loss": 0.2365,
+      "step": 2990
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1480477035501895,
+      "learning_rate": 1.922828250712736e-05,
+      "loss": 0.2176,
+      "step": 2991
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6072046871345662,
+      "learning_rate": 1.9227647950460184e-05,
+      "loss": 0.2173,
+      "step": 2992
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.510713170555769,
+      "learning_rate": 1.9227013143492534e-05,
+      "loss": 0.2311,
+      "step": 2993
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.607032952016177,
+      "learning_rate": 1.922637808624163e-05,
+      "loss": 0.2301,
+      "step": 2994
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2996906089038012,
+      "learning_rate": 1.92257427787247e-05,
+      "loss": 0.2596,
+      "step": 2995
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0050341574660457,
+      "learning_rate": 1.922510722095898e-05,
+      "loss": 0.2361,
+      "step": 2996
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9414520504409785,
+      "learning_rate": 1.92244714129617e-05,
+      "loss": 0.2298,
+      "step": 2997
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9724450849054851,
+      "learning_rate": 1.9223835354750117e-05,
+      "loss": 0.2228,
+      "step": 2998
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9975235707500363,
+      "learning_rate": 1.9223199046341477e-05,
+      "loss": 0.2471,
+      "step": 2999
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9784518821562274,
+      "learning_rate": 1.922256248775304e-05,
+      "loss": 0.222,
+      "step": 3000
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8459640070005651,
+      "learning_rate": 1.9221925679002076e-05,
+      "loss": 0.2217,
+      "step": 3001
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.255243443188498,
+      "learning_rate": 1.9221288620105857e-05,
+      "loss": 0.2356,
+      "step": 3002
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8893090922997737,
+      "learning_rate": 1.9220651311081666e-05,
+      "loss": 0.2279,
+      "step": 3003
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9402338401958679,
+      "learning_rate": 1.922001375194678e-05,
+      "loss": 0.2093,
+      "step": 3004
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8455121896699624,
+      "learning_rate": 1.9219375942718508e-05,
+      "loss": 0.2226,
+      "step": 3005
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2765300051535464,
+      "learning_rate": 1.921873788341414e-05,
+      "loss": 0.2542,
+      "step": 3006
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2140638684126275,
+      "learning_rate": 1.9218099574050985e-05,
+      "loss": 0.2534,
+      "step": 3007
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0750642375996193,
+      "learning_rate": 1.9217461014646362e-05,
+      "loss": 0.2138,
+      "step": 3008
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.246919525779567,
+      "learning_rate": 1.9216822205217586e-05,
+      "loss": 0.2293,
+      "step": 3009
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0217795188677212,
+      "learning_rate": 1.9216183145781984e-05,
+      "loss": 0.2211,
+      "step": 3010
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.587256853484376,
+      "learning_rate": 1.92155438363569e-05,
+      "loss": 0.2676,
+      "step": 3011
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1552003644377378,
+      "learning_rate": 1.9214904276959664e-05,
+      "loss": 0.2202,
+      "step": 3012
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.110685839383147,
+      "learning_rate": 1.921426446760763e-05,
+      "loss": 0.2331,
+      "step": 3013
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7931555285092697,
+      "learning_rate": 1.9213624408318155e-05,
+      "loss": 0.2201,
+      "step": 3014
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2477592588381938,
+      "learning_rate": 1.9212984099108594e-05,
+      "loss": 0.2276,
+      "step": 3015
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4367511021452408,
+      "learning_rate": 1.921234353999632e-05,
+      "loss": 0.2284,
+      "step": 3016
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.306302205604074,
+      "learning_rate": 1.921170273099871e-05,
+      "loss": 0.2194,
+      "step": 3017
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0782994679740574,
+      "learning_rate": 1.921106167213314e-05,
+      "loss": 0.2372,
+      "step": 3018
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0505196944402984,
+      "learning_rate": 1.9210420363417e-05,
+      "loss": 0.2144,
+      "step": 3019
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9353288519979084,
+      "learning_rate": 1.920977880486769e-05,
+      "loss": 0.2283,
+      "step": 3020
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.974558823252189,
+      "learning_rate": 1.920913699650261e-05,
+      "loss": 0.217,
+      "step": 3021
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9563677813824926,
+      "learning_rate": 1.920849493833917e-05,
+      "loss": 0.2466,
+      "step": 3022
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9999446931006459,
+      "learning_rate": 1.9207852630394782e-05,
+      "loss": 0.1926,
+      "step": 3023
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1905647435368694,
+      "learning_rate": 1.920721007268687e-05,
+      "loss": 0.211,
+      "step": 3024
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9547156412821325,
+      "learning_rate": 1.9206567265232867e-05,
+      "loss": 0.2114,
+      "step": 3025
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7980185646523026,
+      "learning_rate": 1.920592420805021e-05,
+      "loss": 0.2466,
+      "step": 3026
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.160882813189863,
+      "learning_rate": 1.9205280901156332e-05,
+      "loss": 0.203,
+      "step": 3027
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.482553328977277,
+      "learning_rate": 1.9204637344568694e-05,
+      "loss": 0.2092,
+      "step": 3028
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8412106230735823,
+      "learning_rate": 1.920399353830475e-05,
+      "loss": 0.2172,
+      "step": 3029
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.029983017735956,
+      "learning_rate": 1.920334948238196e-05,
+      "loss": 0.2332,
+      "step": 3030
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9229551534640521,
+      "learning_rate": 1.9202705176817794e-05,
+      "loss": 0.2158,
+      "step": 3031
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1423176753684943,
+      "learning_rate": 1.920206062162973e-05,
+      "loss": 0.2294,
+      "step": 3032
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9970226971429255,
+      "learning_rate": 1.9201415816835254e-05,
+      "loss": 0.2271,
+      "step": 3033
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1127263130100777,
+      "learning_rate": 1.9200770762451854e-05,
+      "loss": 0.2265,
+      "step": 3034
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9070022790029587,
+      "learning_rate": 1.9200125458497025e-05,
+      "loss": 0.226,
+      "step": 3035
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4140524654210664,
+      "learning_rate": 1.9199479904988277e-05,
+      "loss": 0.2193,
+      "step": 3036
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2310860599371727,
+      "learning_rate": 1.9198834101943115e-05,
+      "loss": 0.2257,
+      "step": 3037
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2545161274463466,
+      "learning_rate": 1.9198188049379055e-05,
+      "loss": 0.2317,
+      "step": 3038
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0034022129693965,
+      "learning_rate": 1.919754174731363e-05,
+      "loss": 0.235,
+      "step": 3039
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.114243885076107,
+      "learning_rate": 1.9196895195764363e-05,
+      "loss": 0.243,
+      "step": 3040
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0156746704279098,
+      "learning_rate": 1.9196248394748794e-05,
+      "loss": 0.2241,
+      "step": 3041
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0509811949541532,
+      "learning_rate": 1.919560134428447e-05,
+      "loss": 0.2409,
+      "step": 3042
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9396601478975011,
+      "learning_rate": 1.919495404438894e-05,
+      "loss": 0.2226,
+      "step": 3043
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.312189831656218,
+      "learning_rate": 1.919430649507976e-05,
+      "loss": 0.2228,
+      "step": 3044
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1669179726287418,
+      "learning_rate": 1.9193658696374498e-05,
+      "loss": 0.2098,
+      "step": 3045
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.053822503469039,
+      "learning_rate": 1.9193010648290725e-05,
+      "loss": 0.219,
+      "step": 3046
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.897292649110749,
+      "learning_rate": 1.919236235084602e-05,
+      "loss": 0.2268,
+      "step": 3047
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1825640096018137,
+      "learning_rate": 1.9191713804057965e-05,
+      "loss": 0.2277,
+      "step": 3048
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0258482915835343,
+      "learning_rate": 1.9191065007944153e-05,
+      "loss": 0.229,
+      "step": 3049
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0150255181566332,
+      "learning_rate": 1.9190415962522186e-05,
+      "loss": 0.2122,
+      "step": 3050
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.221461841386964,
+      "learning_rate": 1.9189766667809667e-05,
+      "loss": 0.2288,
+      "step": 3051
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0912579290474678,
+      "learning_rate": 1.9189117123824208e-05,
+      "loss": 0.2157,
+      "step": 3052
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.092173960673062,
+      "learning_rate": 1.9188467330583428e-05,
+      "loss": 0.2197,
+      "step": 3053
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2105441945424267,
+      "learning_rate": 1.918781728810495e-05,
+      "loss": 0.2308,
+      "step": 3054
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0786894300997212,
+      "learning_rate": 1.9187166996406413e-05,
+      "loss": 0.2213,
+      "step": 3055
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0532621519315912,
+      "learning_rate": 1.918651645550545e-05,
+      "loss": 0.2421,
+      "step": 3056
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9811524305129725,
+      "learning_rate": 1.9185865665419708e-05,
+      "loss": 0.2341,
+      "step": 3057
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0652674203246557,
+      "learning_rate": 1.9185214626166845e-05,
+      "loss": 0.2249,
+      "step": 3058
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9858359011987853,
+      "learning_rate": 1.9184563337764516e-05,
+      "loss": 0.22,
+      "step": 3059
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9134652124007379,
+      "learning_rate": 1.9183911800230384e-05,
+      "loss": 0.2318,
+      "step": 3060
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2143347015760846,
+      "learning_rate": 1.9183260013582126e-05,
+      "loss": 0.2309,
+      "step": 3061
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9738448412568219,
+      "learning_rate": 1.9182607977837424e-05,
+      "loss": 0.2371,
+      "step": 3062
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7966606408722735,
+      "learning_rate": 1.9181955693013962e-05,
+      "loss": 0.2107,
+      "step": 3063
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4371746934422849,
+      "learning_rate": 1.918130315912943e-05,
+      "loss": 0.2405,
+      "step": 3064
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9880342880142227,
+      "learning_rate": 1.9180650376201536e-05,
+      "loss": 0.2291,
+      "step": 3065
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0844528067687451,
+      "learning_rate": 1.917999734424798e-05,
+      "loss": 0.2339,
+      "step": 3066
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.273986948076706,
+      "learning_rate": 1.9179344063286475e-05,
+      "loss": 0.2195,
+      "step": 3067
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8814553758365897,
+      "learning_rate": 1.917869053333475e-05,
+      "loss": 0.2086,
+      "step": 3068
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.077182563935586,
+      "learning_rate": 1.9178036754410518e-05,
+      "loss": 0.2181,
+      "step": 3069
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9425617308554278,
+      "learning_rate": 1.9177382726531527e-05,
+      "loss": 0.2067,
+      "step": 3070
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.013551998686744,
+      "learning_rate": 1.9176728449715506e-05,
+      "loss": 0.2191,
+      "step": 3071
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0285491895566228,
+      "learning_rate": 1.9176073923980212e-05,
+      "loss": 0.2442,
+      "step": 3072
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0493606012843992,
+      "learning_rate": 1.917541914934339e-05,
+      "loss": 0.206,
+      "step": 3073
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2117368635392565,
+      "learning_rate": 1.917476412582281e-05,
+      "loss": 0.2272,
+      "step": 3074
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5386868025963205,
+      "learning_rate": 1.9174108853436234e-05,
+      "loss": 0.2331,
+      "step": 3075
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.311440053988242,
+      "learning_rate": 1.9173453332201436e-05,
+      "loss": 0.2102,
+      "step": 3076
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4581511839772705,
+      "learning_rate": 1.91727975621362e-05,
+      "loss": 0.2173,
+      "step": 3077
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9644674644204856,
+      "learning_rate": 1.917214154325831e-05,
+      "loss": 0.2111,
+      "step": 3078
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1068412843688058,
+      "learning_rate": 1.917148527558556e-05,
+      "loss": 0.2383,
+      "step": 3079
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1228544945994425,
+      "learning_rate": 1.917082875913576e-05,
+      "loss": 0.2257,
+      "step": 3080
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1290992454356428,
+      "learning_rate": 1.9170171993926708e-05,
+      "loss": 0.2302,
+      "step": 3081
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5171743857569169,
+      "learning_rate": 1.9169514979976224e-05,
+      "loss": 0.2399,
+      "step": 3082
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.179035837795795,
+      "learning_rate": 1.9168857717302128e-05,
+      "loss": 0.2181,
+      "step": 3083
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9525048557256401,
+      "learning_rate": 1.9168200205922248e-05,
+      "loss": 0.2206,
+      "step": 3084
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9192591243102435,
+      "learning_rate": 1.916754244585442e-05,
+      "loss": 0.2114,
+      "step": 3085
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2411058008519829,
+      "learning_rate": 1.9166884437116486e-05,
+      "loss": 0.2307,
+      "step": 3086
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4298363741722075,
+      "learning_rate": 1.9166226179726294e-05,
+      "loss": 0.2412,
+      "step": 3087
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.051321479238389,
+      "learning_rate": 1.9165567673701696e-05,
+      "loss": 0.2105,
+      "step": 3088
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1983398040289723,
+      "learning_rate": 1.9164908919060562e-05,
+      "loss": 0.2134,
+      "step": 3089
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9650459782933238,
+      "learning_rate": 1.9164249915820753e-05,
+      "loss": 0.2188,
+      "step": 3090
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1929879231929374,
+      "learning_rate": 1.9163590664000145e-05,
+      "loss": 0.2292,
+      "step": 3091
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9577148201922246,
+      "learning_rate": 1.916293116361663e-05,
+      "loss": 0.2178,
+      "step": 3092
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0158885393736095,
+      "learning_rate": 1.916227141468808e-05,
+      "loss": 0.2073,
+      "step": 3093
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2815901739893645,
+      "learning_rate": 1.9161611417232407e-05,
+      "loss": 0.1999,
+      "step": 3094
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.197536993244301,
+      "learning_rate": 1.9160951171267508e-05,
+      "loss": 0.2276,
+      "step": 3095
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2323433276514397,
+      "learning_rate": 1.9160290676811288e-05,
+      "loss": 0.2112,
+      "step": 3096
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.119380261757278,
+      "learning_rate": 1.9159629933881666e-05,
+      "loss": 0.2015,
+      "step": 3097
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.045749768250887,
+      "learning_rate": 1.915896894249657e-05,
+      "loss": 0.2191,
+      "step": 3098
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9658151735731435,
+      "learning_rate": 1.9158307702673917e-05,
+      "loss": 0.2234,
+      "step": 3099
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9059897342936998,
+      "learning_rate": 1.9157646214431653e-05,
+      "loss": 0.2434,
+      "step": 3100
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1570549601192839,
+      "learning_rate": 1.9156984477787717e-05,
+      "loss": 0.2198,
+      "step": 3101
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.142206977481465,
+      "learning_rate": 1.9156322492760064e-05,
+      "loss": 0.2396,
+      "step": 3102
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4612871008091388,
+      "learning_rate": 1.915566025936664e-05,
+      "loss": 0.2434,
+      "step": 3103
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.161012691350098,
+      "learning_rate": 1.9154997777625418e-05,
+      "loss": 0.2316,
+      "step": 3104
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0674154619263267,
+      "learning_rate": 1.9154335047554364e-05,
+      "loss": 0.2143,
+      "step": 3105
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.420357609574483,
+      "learning_rate": 1.9153672069171454e-05,
+      "loss": 0.222,
+      "step": 3106
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1620397000841887,
+      "learning_rate": 1.9153008842494673e-05,
+      "loss": 0.235,
+      "step": 3107
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1933801258599444,
+      "learning_rate": 1.9152345367542008e-05,
+      "loss": 0.221,
+      "step": 3108
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.92162970400094,
+      "learning_rate": 1.915168164433146e-05,
+      "loss": 0.2049,
+      "step": 3109
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0999983754740923,
+      "learning_rate": 1.9151017672881032e-05,
+      "loss": 0.2288,
+      "step": 3110
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7372094733382606,
+      "learning_rate": 1.915035345320873e-05,
+      "loss": 0.262,
+      "step": 3111
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.158268282961799,
+      "learning_rate": 1.9149688985332575e-05,
+      "loss": 0.2077,
+      "step": 3112
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0296742546521,
+      "learning_rate": 1.914902426927059e-05,
+      "loss": 0.2285,
+      "step": 3113
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8850667462372587,
+      "learning_rate": 1.9148359305040802e-05,
+      "loss": 0.223,
+      "step": 3114
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1648725757139722,
+      "learning_rate": 1.9147694092661254e-05,
+      "loss": 0.222,
+      "step": 3115
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1864890985053547,
+      "learning_rate": 1.914702863214999e-05,
+      "loss": 0.2313,
+      "step": 3116
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9597519082119035,
+      "learning_rate": 1.9146362923525053e-05,
+      "loss": 0.2367,
+      "step": 3117
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0768884920382897,
+      "learning_rate": 1.9145696966804505e-05,
+      "loss": 0.2137,
+      "step": 3118
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2765280153968082,
+      "learning_rate": 1.914503076200641e-05,
+      "loss": 0.2335,
+      "step": 3119
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.891702028568063,
+      "learning_rate": 1.9144364309148842e-05,
+      "loss": 0.2163,
+      "step": 3120
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0401173324989785,
+      "learning_rate": 1.9143697608249873e-05,
+      "loss": 0.2443,
+      "step": 3121
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.150292160815212,
+      "learning_rate": 1.914303065932759e-05,
+      "loss": 0.2442,
+      "step": 3122
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9544448291925663,
+      "learning_rate": 1.9142363462400087e-05,
+      "loss": 0.2108,
+      "step": 3123
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9610639977235427,
+      "learning_rate": 1.914169601748546e-05,
+      "loss": 0.2076,
+      "step": 3124
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0201336974255446,
+      "learning_rate": 1.9141028324601808e-05,
+      "loss": 0.2314,
+      "step": 3125
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.189912911424515,
+      "learning_rate": 1.9140360383767248e-05,
+      "loss": 0.1984,
+      "step": 3126
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2961454181303826,
+      "learning_rate": 1.9139692194999894e-05,
+      "loss": 0.2383,
+      "step": 3127
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4274584800275039,
+      "learning_rate": 1.9139023758317875e-05,
+      "loss": 0.235,
+      "step": 3128
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8855836301343938,
+      "learning_rate": 1.913835507373932e-05,
+      "loss": 0.2156,
+      "step": 3129
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0934329027469938,
+      "learning_rate": 1.9137686141282368e-05,
+      "loss": 0.2233,
+      "step": 3130
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8189168623294839,
+      "learning_rate": 1.9137016960965164e-05,
+      "loss": 0.1943,
+      "step": 3131
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2129777212990402,
+      "learning_rate": 1.9136347532805855e-05,
+      "loss": 0.2486,
+      "step": 3132
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8058648848724157,
+      "learning_rate": 1.9135677856822606e-05,
+      "loss": 0.2119,
+      "step": 3133
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.01784769091552,
+      "learning_rate": 1.9135007933033583e-05,
+      "loss": 0.2201,
+      "step": 3134
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9792483281736258,
+      "learning_rate": 1.913433776145695e-05,
+      "loss": 0.238,
+      "step": 3135
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0007834040170969,
+      "learning_rate": 1.9133667342110887e-05,
+      "loss": 0.1993,
+      "step": 3136
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1539922532606464,
+      "learning_rate": 1.9132996675013583e-05,
+      "loss": 0.2184,
+      "step": 3137
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1508044653257241,
+      "learning_rate": 1.913232576018323e-05,
+      "loss": 0.2313,
+      "step": 3138
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.069508988072164,
+      "learning_rate": 1.9131654597638024e-05,
+      "loss": 0.2287,
+      "step": 3139
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.374642745001945,
+      "learning_rate": 1.9130983187396174e-05,
+      "loss": 0.2417,
+      "step": 3140
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.131383765935505,
+      "learning_rate": 1.9130311529475886e-05,
+      "loss": 0.2367,
+      "step": 3141
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9307771276706998,
+      "learning_rate": 1.9129639623895382e-05,
+      "loss": 0.2307,
+      "step": 3142
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3643681521091495,
+      "learning_rate": 1.9128967470672887e-05,
+      "loss": 0.2675,
+      "step": 3143
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0992636255726658,
+      "learning_rate": 1.9128295069826636e-05,
+      "loss": 0.238,
+      "step": 3144
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8847005171553731,
+      "learning_rate": 1.9127622421374866e-05,
+      "loss": 0.2194,
+      "step": 3145
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2081783271673872,
+      "learning_rate": 1.912694952533582e-05,
+      "loss": 0.2383,
+      "step": 3146
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9796553656595477,
+      "learning_rate": 1.9126276381727752e-05,
+      "loss": 0.2247,
+      "step": 3147
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8756885079427202,
+      "learning_rate": 1.9125602990568925e-05,
+      "loss": 0.1946,
+      "step": 3148
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2317980453228476,
+      "learning_rate": 1.91249293518776e-05,
+      "loss": 0.2202,
+      "step": 3149
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0438633428794166,
+      "learning_rate": 1.9124255465672053e-05,
+      "loss": 0.2118,
+      "step": 3150
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9645029574938986,
+      "learning_rate": 1.9123581331970558e-05,
+      "loss": 0.2419,
+      "step": 3151
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0684720842692546,
+      "learning_rate": 1.9122906950791406e-05,
+      "loss": 0.2385,
+      "step": 3152
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3288111295644136,
+      "learning_rate": 1.9122232322152883e-05,
+      "loss": 0.1971,
+      "step": 3153
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.033806895095567,
+      "learning_rate": 1.91215574460733e-05,
+      "loss": 0.2276,
+      "step": 3154
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0238508770319936,
+      "learning_rate": 1.9120882322570952e-05,
+      "loss": 0.2428,
+      "step": 3155
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.03641331436747,
+      "learning_rate": 1.912020695166416e-05,
+      "loss": 0.2319,
+      "step": 3156
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5371657052324845,
+      "learning_rate": 1.9119531333371233e-05,
+      "loss": 0.2649,
+      "step": 3157
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9765508272616171,
+      "learning_rate": 1.9118855467710507e-05,
+      "loss": 0.1991,
+      "step": 3158
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1342684309424567,
+      "learning_rate": 1.911817935470031e-05,
+      "loss": 0.2457,
+      "step": 3159
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.332260724706468,
+      "learning_rate": 1.9117502994358984e-05,
+      "loss": 0.2313,
+      "step": 3160
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2189330783389225,
+      "learning_rate": 1.9116826386704873e-05,
+      "loss": 0.2363,
+      "step": 3161
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.037243374045091,
+      "learning_rate": 1.9116149531756333e-05,
+      "loss": 0.2506,
+      "step": 3162
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4394206019917581,
+      "learning_rate": 1.9115472429531722e-05,
+      "loss": 0.2078,
+      "step": 3163
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2227594911150255,
+      "learning_rate": 1.911479508004941e-05,
+      "loss": 0.2286,
+      "step": 3164
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8577485490636886,
+      "learning_rate": 1.911411748332776e-05,
+      "loss": 0.2209,
+      "step": 3165
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2941653244360138,
+      "learning_rate": 1.9113439639385164e-05,
+      "loss": 0.2426,
+      "step": 3166
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0202153420555122,
+      "learning_rate": 1.9112761548239996e-05,
+      "loss": 0.2191,
+      "step": 3167
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6811743560837742,
+      "learning_rate": 1.911208320991066e-05,
+      "loss": 0.2217,
+      "step": 3168
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4294763694872188,
+      "learning_rate": 1.9111404624415554e-05,
+      "loss": 0.2437,
+      "step": 3169
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3818248282341945,
+      "learning_rate": 1.9110725791773085e-05,
+      "loss": 0.2393,
+      "step": 3170
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9783655888403653,
+      "learning_rate": 1.911004671200166e-05,
+      "loss": 0.2074,
+      "step": 3171
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3146336646097434,
+      "learning_rate": 1.9109367385119705e-05,
+      "loss": 0.2205,
+      "step": 3172
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8869805809256699,
+      "learning_rate": 1.9108687811145645e-05,
+      "loss": 0.2042,
+      "step": 3173
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0622836441472836,
+      "learning_rate": 1.9108007990097913e-05,
+      "loss": 0.2256,
+      "step": 3174
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9475793238096982,
+      "learning_rate": 1.910732792199495e-05,
+      "loss": 0.2319,
+      "step": 3175
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0748427136473677,
+      "learning_rate": 1.9106647606855203e-05,
+      "loss": 0.2387,
+      "step": 3176
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0525166947996967,
+      "learning_rate": 1.9105967044697125e-05,
+      "loss": 0.2305,
+      "step": 3177
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0904798663674182,
+      "learning_rate": 1.9105286235539178e-05,
+      "loss": 0.2356,
+      "step": 3178
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1674600619281295,
+      "learning_rate": 1.9104605179399827e-05,
+      "loss": 0.2236,
+      "step": 3179
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.090127583465506,
+      "learning_rate": 1.9103923876297544e-05,
+      "loss": 0.2298,
+      "step": 3180
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8327437385704604,
+      "learning_rate": 1.9103242326250815e-05,
+      "loss": 0.2066,
+      "step": 3181
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1699184536903102,
+      "learning_rate": 1.9102560529278122e-05,
+      "loss": 0.234,
+      "step": 3182
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1169478692759198,
+      "learning_rate": 1.910187848539796e-05,
+      "loss": 0.2445,
+      "step": 3183
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8226074400974731,
+      "learning_rate": 1.9101196194628834e-05,
+      "loss": 0.1799,
+      "step": 3184
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9114950516377383,
+      "learning_rate": 1.9100513656989244e-05,
+      "loss": 0.2268,
+      "step": 3185
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9785174645275201,
+      "learning_rate": 1.9099830872497707e-05,
+      "loss": 0.2279,
+      "step": 3186
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8498906357021067,
+      "learning_rate": 1.909914784117274e-05,
+      "loss": 0.1969,
+      "step": 3187
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.16707129988209,
+      "learning_rate": 1.9098464563032878e-05,
+      "loss": 0.243,
+      "step": 3188
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1269165149984086,
+      "learning_rate": 1.9097781038096652e-05,
+      "loss": 0.2092,
+      "step": 3189
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2750144293869767,
+      "learning_rate": 1.9097097266382598e-05,
+      "loss": 0.2094,
+      "step": 3190
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0660217584543055,
+      "learning_rate": 1.909641324790927e-05,
+      "loss": 0.2109,
+      "step": 3191
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.122843501272006,
+      "learning_rate": 1.909572898269522e-05,
+      "loss": 0.2205,
+      "step": 3192
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1216680184980137,
+      "learning_rate": 1.9095044470759004e-05,
+      "loss": 0.2398,
+      "step": 3193
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.32198193611617,
+      "learning_rate": 1.9094359712119192e-05,
+      "loss": 0.2416,
+      "step": 3194
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2047762964034372,
+      "learning_rate": 1.9093674706794363e-05,
+      "loss": 0.2091,
+      "step": 3195
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0298224869937613,
+      "learning_rate": 1.9092989454803094e-05,
+      "loss": 0.2606,
+      "step": 3196
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5457240549213624,
+      "learning_rate": 1.909230395616397e-05,
+      "loss": 0.2052,
+      "step": 3197
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9502655176707185,
+      "learning_rate": 1.909161821089559e-05,
+      "loss": 0.2274,
+      "step": 3198
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.01536728621932,
+      "learning_rate": 1.9090932219016548e-05,
+      "loss": 0.2049,
+      "step": 3199
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2487157303403302,
+      "learning_rate": 1.909024598054546e-05,
+      "loss": 0.2306,
+      "step": 3200
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2078412801534302,
+      "learning_rate": 1.9089559495500934e-05,
+      "loss": 0.2371,
+      "step": 3201
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0483351379753305,
+      "learning_rate": 1.90888727639016e-05,
+      "loss": 0.2389,
+      "step": 3202
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.128869956503603,
+      "learning_rate": 1.908818578576607e-05,
+      "loss": 0.235,
+      "step": 3203
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0102531146450082,
+      "learning_rate": 1.9087498561112992e-05,
+      "loss": 0.211,
+      "step": 3204
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0299810783657681,
+      "learning_rate": 1.9086811089961e-05,
+      "loss": 0.2387,
+      "step": 3205
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9802814387541344,
+      "learning_rate": 1.9086123372328748e-05,
+      "loss": 0.2237,
+      "step": 3206
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9266592538656312,
+      "learning_rate": 1.9085435408234882e-05,
+      "loss": 0.2249,
+      "step": 3207
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9220119641068419,
+      "learning_rate": 1.9084747197698068e-05,
+      "loss": 0.2266,
+      "step": 3208
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1514474631890381,
+      "learning_rate": 1.9084058740736974e-05,
+      "loss": 0.2119,
+      "step": 3209
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1158632704978646,
+      "learning_rate": 1.9083370037370276e-05,
+      "loss": 0.219,
+      "step": 3210
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0141902460312189,
+      "learning_rate": 1.908268108761665e-05,
+      "loss": 0.2253,
+      "step": 3211
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6050271110549044,
+      "learning_rate": 1.9081991891494787e-05,
+      "loss": 0.2286,
+      "step": 3212
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2109911266896056,
+      "learning_rate": 1.908130244902338e-05,
+      "loss": 0.2159,
+      "step": 3213
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3546715824770614,
+      "learning_rate": 1.9080612760221134e-05,
+      "loss": 0.2547,
+      "step": 3214
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7959685045832672,
+      "learning_rate": 1.907992282510675e-05,
+      "loss": 0.2115,
+      "step": 3215
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5087742014255319,
+      "learning_rate": 1.9079232643698947e-05,
+      "loss": 0.2251,
+      "step": 3216
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.924666360606091,
+      "learning_rate": 1.907854221601645e-05,
+      "loss": 0.2053,
+      "step": 3217
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9530973107451134,
+      "learning_rate": 1.9077851542077978e-05,
+      "loss": 0.2252,
+      "step": 3218
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.950666139394692,
+      "learning_rate": 1.9077160621902274e-05,
+      "loss": 0.2505,
+      "step": 3219
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5103529027736473,
+      "learning_rate": 1.9076469455508072e-05,
+      "loss": 0.2258,
+      "step": 3220
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0096711939190677,
+      "learning_rate": 1.9075778042914126e-05,
+      "loss": 0.2177,
+      "step": 3221
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0032885320723282,
+      "learning_rate": 1.9075086384139187e-05,
+      "loss": 0.2226,
+      "step": 3222
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.206005456150721,
+      "learning_rate": 1.9074394479202017e-05,
+      "loss": 0.2112,
+      "step": 3223
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.150609400759136,
+      "learning_rate": 1.9073702328121382e-05,
+      "loss": 0.2051,
+      "step": 3224
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1756724863924575,
+      "learning_rate": 1.9073009930916063e-05,
+      "loss": 0.2261,
+      "step": 3225
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5673489315487212,
+      "learning_rate": 1.9072317287604837e-05,
+      "loss": 0.1954,
+      "step": 3226
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.568033424099215,
+      "learning_rate": 1.907162439820649e-05,
+      "loss": 0.2572,
+      "step": 3227
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9642478242450317,
+      "learning_rate": 1.907093126273982e-05,
+      "loss": 0.2057,
+      "step": 3228
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.618424478919842,
+      "learning_rate": 1.907023788122363e-05,
+      "loss": 0.2433,
+      "step": 3229
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9986359526257098,
+      "learning_rate": 1.906954425367672e-05,
+      "loss": 0.2171,
+      "step": 3230
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1235099739225607,
+      "learning_rate": 1.906885038011791e-05,
+      "loss": 0.2228,
+      "step": 3231
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.154482508979427,
+      "learning_rate": 1.906815626056602e-05,
+      "loss": 0.2396,
+      "step": 3232
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3890043410430053,
+      "learning_rate": 1.9067461895039888e-05,
+      "loss": 0.2621,
+      "step": 3233
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.358399667722323,
+      "learning_rate": 1.906676728355833e-05,
+      "loss": 0.2182,
+      "step": 3234
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0794146703989658,
+      "learning_rate": 1.9066072426140203e-05,
+      "loss": 0.2255,
+      "step": 3235
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9904688704649632,
+      "learning_rate": 1.9065377322804347e-05,
+      "loss": 0.227,
+      "step": 3236
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.558693059373171,
+      "learning_rate": 1.9064681973569622e-05,
+      "loss": 0.2473,
+      "step": 3237
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9947430603467428,
+      "learning_rate": 1.9063986378454884e-05,
+      "loss": 0.2325,
+      "step": 3238
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2265249569014518,
+      "learning_rate": 1.9063290537479004e-05,
+      "loss": 0.2128,
+      "step": 3239
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0611571246887315,
+      "learning_rate": 1.9062594450660857e-05,
+      "loss": 0.2107,
+      "step": 3240
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1814888608068674,
+      "learning_rate": 1.9061898118019326e-05,
+      "loss": 0.2456,
+      "step": 3241
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7326384660130874,
+      "learning_rate": 1.9061201539573292e-05,
+      "loss": 0.2249,
+      "step": 3242
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0426774919663655,
+      "learning_rate": 1.9060504715341654e-05,
+      "loss": 0.2328,
+      "step": 3243
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.318233062490341,
+      "learning_rate": 1.905980764534332e-05,
+      "loss": 0.2281,
+      "step": 3244
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2924652569164976,
+      "learning_rate": 1.9059110329597185e-05,
+      "loss": 0.2243,
+      "step": 3245
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2512314642191262,
+      "learning_rate": 1.9058412768122175e-05,
+      "loss": 0.2308,
+      "step": 3246
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.083747561104402,
+      "learning_rate": 1.9057714960937205e-05,
+      "loss": 0.2267,
+      "step": 3247
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0294046112909658,
+      "learning_rate": 1.9057016908061205e-05,
+      "loss": 0.2287,
+      "step": 3248
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3535822475892574,
+      "learning_rate": 1.905631860951311e-05,
+      "loss": 0.2515,
+      "step": 3249
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7848478161203745,
+      "learning_rate": 1.905562006531186e-05,
+      "loss": 0.2369,
+      "step": 3250
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3252968240520733,
+      "learning_rate": 1.905492127547641e-05,
+      "loss": 0.2383,
+      "step": 3251
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.093520045965069,
+      "learning_rate": 1.9054222240025706e-05,
+      "loss": 0.2449,
+      "step": 3252
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0924413513876863,
+      "learning_rate": 1.905352295897871e-05,
+      "loss": 0.2235,
+      "step": 3253
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1210881431514452,
+      "learning_rate": 1.9052823432354396e-05,
+      "loss": 0.2335,
+      "step": 3254
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2761817272497882,
+      "learning_rate": 1.905212366017173e-05,
+      "loss": 0.2387,
+      "step": 3255
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6487483979669733,
+      "learning_rate": 1.90514236424497e-05,
+      "loss": 0.213,
+      "step": 3256
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0463799652349561,
+      "learning_rate": 1.9050723379207296e-05,
+      "loss": 0.2427,
+      "step": 3257
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9613364989367936,
+      "learning_rate": 1.9050022870463507e-05,
+      "loss": 0.2282,
+      "step": 3258
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0254909889735437,
+      "learning_rate": 1.9049322116237336e-05,
+      "loss": 0.2341,
+      "step": 3259
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4414193791955492,
+      "learning_rate": 1.9048621116547793e-05,
+      "loss": 0.2174,
+      "step": 3260
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5035822960911616,
+      "learning_rate": 1.904791987141389e-05,
+      "loss": 0.2139,
+      "step": 3261
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9153873984869315,
+      "learning_rate": 1.9047218380854652e-05,
+      "loss": 0.2077,
+      "step": 3262
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.204895202902941,
+      "learning_rate": 1.90465166448891e-05,
+      "loss": 0.2203,
+      "step": 3263
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.235198349138494,
+      "learning_rate": 1.9045814663536275e-05,
+      "loss": 0.2553,
+      "step": 3264
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6304314205413217,
+      "learning_rate": 1.9045112436815217e-05,
+      "loss": 0.2391,
+      "step": 3265
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9506320321314072,
+      "learning_rate": 1.904440996474497e-05,
+      "loss": 0.2381,
+      "step": 3266
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1184506110462116,
+      "learning_rate": 1.9043707247344596e-05,
+      "loss": 0.212,
+      "step": 3267
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.058991829044029,
+      "learning_rate": 1.904300428463315e-05,
+      "loss": 0.2115,
+      "step": 3268
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5336373043447653,
+      "learning_rate": 1.9042301076629704e-05,
+      "loss": 0.2197,
+      "step": 3269
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1747866988506088,
+      "learning_rate": 1.9041597623353327e-05,
+      "loss": 0.2236,
+      "step": 3270
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3912454174431848,
+      "learning_rate": 1.9040893924823108e-05,
+      "loss": 0.2098,
+      "step": 3271
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2756488798101397,
+      "learning_rate": 1.9040189981058128e-05,
+      "loss": 0.2524,
+      "step": 3272
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8518406486526697,
+      "learning_rate": 1.9039485792077485e-05,
+      "loss": 0.2237,
+      "step": 3273
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2949126687880619,
+      "learning_rate": 1.903878135790028e-05,
+      "loss": 0.2297,
+      "step": 3274
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.193263514417983,
+      "learning_rate": 1.903807667854562e-05,
+      "loss": 0.2105,
+      "step": 3275
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3225038029048628,
+      "learning_rate": 1.9037371754032618e-05,
+      "loss": 0.2345,
+      "step": 3276
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1300634224130286,
+      "learning_rate": 1.9036666584380398e-05,
+      "loss": 0.1938,
+      "step": 3277
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9816663216784832,
+      "learning_rate": 1.9035961169608087e-05,
+      "loss": 0.247,
+      "step": 3278
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5861016137562751,
+      "learning_rate": 1.9035255509734816e-05,
+      "loss": 0.2081,
+      "step": 3279
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.082690305307336,
+      "learning_rate": 1.903454960477973e-05,
+      "loss": 0.219,
+      "step": 3280
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3516616270012032,
+      "learning_rate": 1.903384345476198e-05,
+      "loss": 0.1982,
+      "step": 3281
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1284573755467393,
+      "learning_rate": 1.9033137059700712e-05,
+      "loss": 0.2256,
+      "step": 3282
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0071641984300603,
+      "learning_rate": 1.9032430419615094e-05,
+      "loss": 0.2387,
+      "step": 3283
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9908369990968336,
+      "learning_rate": 1.9031723534524287e-05,
+      "loss": 0.2043,
+      "step": 3284
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9434326851921663,
+      "learning_rate": 1.903101640444747e-05,
+      "loss": 0.1993,
+      "step": 3285
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9552024564147497,
+      "learning_rate": 1.9030309029403825e-05,
+      "loss": 0.2083,
+      "step": 3286
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7097234946706654,
+      "learning_rate": 1.9029601409412536e-05,
+      "loss": 0.2347,
+      "step": 3287
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2451337932069761,
+      "learning_rate": 1.90288935444928e-05,
+      "loss": 0.2595,
+      "step": 3288
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1152757703933747,
+      "learning_rate": 1.9028185434663814e-05,
+      "loss": 0.221,
+      "step": 3289
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8721790062954682,
+      "learning_rate": 1.902747707994479e-05,
+      "loss": 0.2246,
+      "step": 3290
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9772443493974295,
+      "learning_rate": 1.902676848035494e-05,
+      "loss": 0.2276,
+      "step": 3291
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1658165944676975,
+      "learning_rate": 1.9026059635913484e-05,
+      "loss": 0.2275,
+      "step": 3292
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4208291073865067,
+      "learning_rate": 1.9025350546639654e-05,
+      "loss": 0.2043,
+      "step": 3293
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1546386268942097,
+      "learning_rate": 1.902464121255268e-05,
+      "loss": 0.2295,
+      "step": 3294
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9581823924399043,
+      "learning_rate": 1.90239316336718e-05,
+      "loss": 0.2291,
+      "step": 3295
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9678091509808123,
+      "learning_rate": 1.9023221810016268e-05,
+      "loss": 0.2213,
+      "step": 3296
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9801111140281928,
+      "learning_rate": 1.9022511741605334e-05,
+      "loss": 0.2178,
+      "step": 3297
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2702708802151763,
+      "learning_rate": 1.9021801428458258e-05,
+      "loss": 0.2208,
+      "step": 3298
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9622585836212532,
+      "learning_rate": 1.902109087059431e-05,
+      "loss": 0.2191,
+      "step": 3299
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0074093143700964,
+      "learning_rate": 1.902038006803276e-05,
+      "loss": 0.2198,
+      "step": 3300
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0348778645975119,
+      "learning_rate": 1.9019669020792896e-05,
+      "loss": 0.2068,
+      "step": 3301
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.878794499122924,
+      "learning_rate": 1.9018957728893997e-05,
+      "loss": 0.1931,
+      "step": 3302
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9688383797935949,
+      "learning_rate": 1.901824619235536e-05,
+      "loss": 0.2115,
+      "step": 3303
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1918226103761322,
+      "learning_rate": 1.9017534411196286e-05,
+      "loss": 0.2262,
+      "step": 3304
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.909382788046008,
+      "learning_rate": 1.9016822385436086e-05,
+      "loss": 0.2242,
+      "step": 3305
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8945541647722558,
+      "learning_rate": 1.9016110115094064e-05,
+      "loss": 0.2313,
+      "step": 3306
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9286600500193802,
+      "learning_rate": 1.9015397600189548e-05,
+      "loss": 0.2113,
+      "step": 3307
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0063058750169587,
+      "learning_rate": 1.9014684840741863e-05,
+      "loss": 0.2239,
+      "step": 3308
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1075815585913915,
+      "learning_rate": 1.9013971836770342e-05,
+      "loss": 0.2347,
+      "step": 3309
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9537824608305381,
+      "learning_rate": 1.9013258588294324e-05,
+      "loss": 0.2021,
+      "step": 3310
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3363127593510533,
+      "learning_rate": 1.9012545095333163e-05,
+      "loss": 0.2246,
+      "step": 3311
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0966856534204155,
+      "learning_rate": 1.9011831357906204e-05,
+      "loss": 0.2427,
+      "step": 3312
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0293981595871236,
+      "learning_rate": 1.901111737603281e-05,
+      "loss": 0.2134,
+      "step": 3313
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2481574644414424,
+      "learning_rate": 1.9010403149732347e-05,
+      "loss": 0.2696,
+      "step": 3314
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9090088403265032,
+      "learning_rate": 1.900968867902419e-05,
+      "loss": 0.214,
+      "step": 3315
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.228778878223172,
+      "learning_rate": 1.9008973963927722e-05,
+      "loss": 0.2327,
+      "step": 3316
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6815691699994177,
+      "learning_rate": 1.9008259004462326e-05,
+      "loss": 0.2281,
+      "step": 3317
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0204039327333942,
+      "learning_rate": 1.9007543800647395e-05,
+      "loss": 0.2123,
+      "step": 3318
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9460619920119041,
+      "learning_rate": 1.9006828352502328e-05,
+      "loss": 0.257,
+      "step": 3319
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.270528757519636,
+      "learning_rate": 1.9006112660046535e-05,
+      "loss": 0.2336,
+      "step": 3320
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.407866180437645,
+      "learning_rate": 1.9005396723299426e-05,
+      "loss": 0.2366,
+      "step": 3321
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1886401077079094,
+      "learning_rate": 1.9004680542280423e-05,
+      "loss": 0.2178,
+      "step": 3322
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3281765035685824,
+      "learning_rate": 1.900396411700895e-05,
+      "loss": 0.2115,
+      "step": 3323
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3100309021873207,
+      "learning_rate": 1.9003247447504447e-05,
+      "loss": 0.2379,
+      "step": 3324
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9845075767531712,
+      "learning_rate": 1.900253053378634e-05,
+      "loss": 0.233,
+      "step": 3325
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6700218837971317,
+      "learning_rate": 1.9001813375874093e-05,
+      "loss": 0.2322,
+      "step": 3326
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0300567032531562,
+      "learning_rate": 1.900109597378714e-05,
+      "loss": 0.2444,
+      "step": 3327
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.10901757705295,
+      "learning_rate": 1.900037832754496e-05,
+      "loss": 0.2259,
+      "step": 3328
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9447163714069051,
+      "learning_rate": 1.8999660437167003e-05,
+      "loss": 0.2366,
+      "step": 3329
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.307139230075517,
+      "learning_rate": 1.8998942302672753e-05,
+      "loss": 0.256,
+      "step": 3330
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.371731200627975,
+      "learning_rate": 1.8998223924081683e-05,
+      "loss": 0.2393,
+      "step": 3331
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1905128490939072,
+      "learning_rate": 1.899750530141328e-05,
+      "loss": 0.2536,
+      "step": 3332
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9503692709073092,
+      "learning_rate": 1.8996786434687035e-05,
+      "loss": 0.2219,
+      "step": 3333
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9439331749855291,
+      "learning_rate": 1.8996067323922454e-05,
+      "loss": 0.2323,
+      "step": 3334
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9532307715531738,
+      "learning_rate": 1.8995347969139034e-05,
+      "loss": 0.2252,
+      "step": 3335
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.552957222597281,
+      "learning_rate": 1.8994628370356296e-05,
+      "loss": 0.2267,
+      "step": 3336
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2716668963571127,
+      "learning_rate": 1.8993908527593756e-05,
+      "loss": 0.2372,
+      "step": 3337
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7846699126598632,
+      "learning_rate": 1.899318844087094e-05,
+      "loss": 0.196,
+      "step": 3338
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0129496638676174,
+      "learning_rate": 1.8992468110207376e-05,
+      "loss": 0.2155,
+      "step": 3339
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8783271979901689,
+      "learning_rate": 1.8991747535622607e-05,
+      "loss": 0.2038,
+      "step": 3340
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9169434293805285,
+      "learning_rate": 1.8991026717136182e-05,
+      "loss": 0.2245,
+      "step": 3341
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9595660483435009,
+      "learning_rate": 1.8990305654767646e-05,
+      "loss": 0.2497,
+      "step": 3342
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5644155971430844,
+      "learning_rate": 1.8989584348536563e-05,
+      "loss": 0.201,
+      "step": 3343
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3322471354041998,
+      "learning_rate": 1.8988862798462496e-05,
+      "loss": 0.2064,
+      "step": 3344
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9960819465156675,
+      "learning_rate": 1.8988141004565017e-05,
+      "loss": 0.2317,
+      "step": 3345
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.025625915335841,
+      "learning_rate": 1.8987418966863708e-05,
+      "loss": 0.223,
+      "step": 3346
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9463078148518489,
+      "learning_rate": 1.898669668537815e-05,
+      "loss": 0.2218,
+      "step": 3347
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0451425798298368,
+      "learning_rate": 1.898597416012794e-05,
+      "loss": 0.2416,
+      "step": 3348
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.220357318287464,
+      "learning_rate": 1.898525139113267e-05,
+      "loss": 0.2352,
+      "step": 3349
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9201741657190172,
+      "learning_rate": 1.898452837841195e-05,
+      "loss": 0.1981,
+      "step": 3350
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1030367729470085,
+      "learning_rate": 1.898380512198539e-05,
+      "loss": 0.206,
+      "step": 3351
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.018621044627046,
+      "learning_rate": 1.898308162187261e-05,
+      "loss": 0.208,
+      "step": 3352
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0250380740128227,
+      "learning_rate": 1.898235787809323e-05,
+      "loss": 0.2217,
+      "step": 3353
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0443360909922148,
+      "learning_rate": 1.8981633890666886e-05,
+      "loss": 0.2061,
+      "step": 3354
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.118094985175832,
+      "learning_rate": 1.8980909659613217e-05,
+      "loss": 0.242,
+      "step": 3355
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1719394317035994,
+      "learning_rate": 1.8980185184951864e-05,
+      "loss": 0.2262,
+      "step": 3356
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1119499336976413,
+      "learning_rate": 1.8979460466702483e-05,
+      "loss": 0.2172,
+      "step": 3357
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9199146540350096,
+      "learning_rate": 1.897873550488473e-05,
+      "loss": 0.2207,
+      "step": 3358
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.597157202627985,
+      "learning_rate": 1.8978010299518268e-05,
+      "loss": 0.2236,
+      "step": 3359
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.145656969444609,
+      "learning_rate": 1.897728485062277e-05,
+      "loss": 0.2448,
+      "step": 3360
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.196306172255373,
+      "learning_rate": 1.8976559158217913e-05,
+      "loss": 0.2421,
+      "step": 3361
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0543796183789402,
+      "learning_rate": 1.8975833222323383e-05,
+      "loss": 0.2004,
+      "step": 3362
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1382328752417115,
+      "learning_rate": 1.897510704295887e-05,
+      "loss": 0.2205,
+      "step": 3363
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2594780069794822,
+      "learning_rate": 1.897438062014407e-05,
+      "loss": 0.2492,
+      "step": 3364
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1099254758261405,
+      "learning_rate": 1.897365395389869e-05,
+      "loss": 0.2082,
+      "step": 3365
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.105618387840811,
+      "learning_rate": 1.8972927044242438e-05,
+      "loss": 0.2275,
+      "step": 3366
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2596995074456943,
+      "learning_rate": 1.8972199891195034e-05,
+      "loss": 0.2188,
+      "step": 3367
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1453752628087497,
+      "learning_rate": 1.8971472494776203e-05,
+      "loss": 0.2037,
+      "step": 3368
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.850769972503201,
+      "learning_rate": 1.8970744855005674e-05,
+      "loss": 0.2566,
+      "step": 3369
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.539628585952559,
+      "learning_rate": 1.897001697190318e-05,
+      "loss": 0.2761,
+      "step": 3370
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1468763974627258,
+      "learning_rate": 1.8969288845488473e-05,
+      "loss": 0.2271,
+      "step": 3371
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2157208655996792,
+      "learning_rate": 1.8968560475781297e-05,
+      "loss": 0.2655,
+      "step": 3372
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1547576211357717,
+      "learning_rate": 1.8967831862801414e-05,
+      "loss": 0.213,
+      "step": 3373
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2452954267397751,
+      "learning_rate": 1.8967103006568583e-05,
+      "loss": 0.2269,
+      "step": 3374
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1873167847152575,
+      "learning_rate": 1.8966373907102577e-05,
+      "loss": 0.2443,
+      "step": 3375
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0945146479285706,
+      "learning_rate": 1.8965644564423173e-05,
+      "loss": 0.2098,
+      "step": 3376
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.59138745969418,
+      "learning_rate": 1.8964914978550154e-05,
+      "loss": 0.2093,
+      "step": 3377
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0840798791455932,
+      "learning_rate": 1.896418514950331e-05,
+      "loss": 0.2356,
+      "step": 3378
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.207943842119613,
+      "learning_rate": 1.8963455077302435e-05,
+      "loss": 0.2311,
+      "step": 3379
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1114184292207892,
+      "learning_rate": 1.896272476196734e-05,
+      "loss": 0.2061,
+      "step": 3380
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6099084050038648,
+      "learning_rate": 1.8961994203517822e-05,
+      "loss": 0.2319,
+      "step": 3381
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9832513731356815,
+      "learning_rate": 1.896126340197371e-05,
+      "loss": 0.2509,
+      "step": 3382
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2736609473875997,
+      "learning_rate": 1.896053235735482e-05,
+      "loss": 0.2466,
+      "step": 3383
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.028911036798015,
+      "learning_rate": 1.8959801069680986e-05,
+      "loss": 0.2225,
+      "step": 3384
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.378791568014497,
+      "learning_rate": 1.8959069538972043e-05,
+      "loss": 0.203,
+      "step": 3385
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2369113654728618,
+      "learning_rate": 1.895833776524783e-05,
+      "loss": 0.2125,
+      "step": 3386
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.192780451638109,
+      "learning_rate": 1.89576057485282e-05,
+      "loss": 0.2195,
+      "step": 3387
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3540915753515144,
+      "learning_rate": 1.8956873488833008e-05,
+      "loss": 0.2378,
+      "step": 3388
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.194988330128871,
+      "learning_rate": 1.8956140986182116e-05,
+      "loss": 0.2271,
+      "step": 3389
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.473213780154375,
+      "learning_rate": 1.8955408240595396e-05,
+      "loss": 0.2136,
+      "step": 3390
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8201258083185985,
+      "learning_rate": 1.8954675252092717e-05,
+      "loss": 0.2231,
+      "step": 3391
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8603983223800908,
+      "learning_rate": 1.895394202069397e-05,
+      "loss": 0.2228,
+      "step": 3392
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.894489027911129,
+      "learning_rate": 1.895320854641904e-05,
+      "loss": 0.2111,
+      "step": 3393
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1816876791342712,
+      "learning_rate": 1.8952474829287825e-05,
+      "loss": 0.2426,
+      "step": 3394
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.155318420616888,
+      "learning_rate": 1.895174086932022e-05,
+      "loss": 0.2128,
+      "step": 3395
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9597525424228827,
+      "learning_rate": 1.895100666653614e-05,
+      "loss": 0.2171,
+      "step": 3396
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9415740622946664,
+      "learning_rate": 1.8950272220955497e-05,
+      "loss": 0.2316,
+      "step": 3397
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9863523292989811,
+      "learning_rate": 1.8949537532598213e-05,
+      "loss": 0.2078,
+      "step": 3398
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6101817215152627,
+      "learning_rate": 1.8948802601484224e-05,
+      "loss": 0.2297,
+      "step": 3399
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2594625137553501,
+      "learning_rate": 1.8948067427633456e-05,
+      "loss": 0.2318,
+      "step": 3400
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7816854513161775,
+      "learning_rate": 1.8947332011065853e-05,
+      "loss": 0.1975,
+      "step": 3401
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1061855335081938,
+      "learning_rate": 1.8946596351801363e-05,
+      "loss": 0.2419,
+      "step": 3402
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5434594029241404,
+      "learning_rate": 1.8945860449859945e-05,
+      "loss": 0.2057,
+      "step": 3403
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1371130973679873,
+      "learning_rate": 1.8945124305261555e-05,
+      "loss": 0.2249,
+      "step": 3404
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9609027280934379,
+      "learning_rate": 1.8944387918026162e-05,
+      "loss": 0.2081,
+      "step": 3405
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9246659917659515,
+      "learning_rate": 1.8943651288173743e-05,
+      "loss": 0.2175,
+      "step": 3406
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9675906075566838,
+      "learning_rate": 1.8942914415724275e-05,
+      "loss": 0.2108,
+      "step": 3407
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.167654837285515,
+      "learning_rate": 1.8942177300697753e-05,
+      "loss": 0.2064,
+      "step": 3408
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.30419671209794,
+      "learning_rate": 1.8941439943114162e-05,
+      "loss": 0.1932,
+      "step": 3409
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6847133694504224,
+      "learning_rate": 1.8940702342993512e-05,
+      "loss": 0.2183,
+      "step": 3410
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2953727771799972,
+      "learning_rate": 1.8939964500355806e-05,
+      "loss": 0.2059,
+      "step": 3411
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3106937767502531,
+      "learning_rate": 1.8939226415221054e-05,
+      "loss": 0.205,
+      "step": 3412
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.592550504669995,
+      "learning_rate": 1.893848808760928e-05,
+      "loss": 0.1927,
+      "step": 3413
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4358442707709609,
+      "learning_rate": 1.8937749517540516e-05,
+      "loss": 0.2047,
+      "step": 3414
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2130697178931753,
+      "learning_rate": 1.8937010705034788e-05,
+      "loss": 0.2185,
+      "step": 3415
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.204625251601906,
+      "learning_rate": 1.8936271650112143e-05,
+      "loss": 0.2199,
+      "step": 3416
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.648059762048116,
+      "learning_rate": 1.8935532352792624e-05,
+      "loss": 0.2435,
+      "step": 3417
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5113204599402539,
+      "learning_rate": 1.8934792813096283e-05,
+      "loss": 0.2171,
+      "step": 3418
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.08312601172551,
+      "learning_rate": 1.8934053031043185e-05,
+      "loss": 0.2164,
+      "step": 3419
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8703153818528033,
+      "learning_rate": 1.8933313006653392e-05,
+      "loss": 0.2151,
+      "step": 3420
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9451239356452712,
+      "learning_rate": 1.893257273994698e-05,
+      "loss": 0.2394,
+      "step": 3421
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2083503347871671,
+      "learning_rate": 1.8931832230944026e-05,
+      "loss": 0.2349,
+      "step": 3422
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2617570854350872,
+      "learning_rate": 1.8931091479664622e-05,
+      "loss": 0.2264,
+      "step": 3423
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1220719798890133,
+      "learning_rate": 1.8930350486128855e-05,
+      "loss": 0.2257,
+      "step": 3424
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.210152001777275,
+      "learning_rate": 1.8929609250356827e-05,
+      "loss": 0.2366,
+      "step": 3425
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5213675641639302,
+      "learning_rate": 1.8928867772368644e-05,
+      "loss": 0.2119,
+      "step": 3426
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.161765898603026,
+      "learning_rate": 1.892812605218442e-05,
+      "loss": 0.2024,
+      "step": 3427
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.038853006912805,
+      "learning_rate": 1.8927384089824267e-05,
+      "loss": 0.1975,
+      "step": 3428
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2892374237322095,
+      "learning_rate": 1.8926641885308325e-05,
+      "loss": 0.2194,
+      "step": 3429
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5090525000593058,
+      "learning_rate": 1.8925899438656708e-05,
+      "loss": 0.2007,
+      "step": 3430
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3066340896057012,
+      "learning_rate": 1.892515674988957e-05,
+      "loss": 0.2439,
+      "step": 3431
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6703291899021413,
+      "learning_rate": 1.892441381902705e-05,
+      "loss": 0.2068,
+      "step": 3432
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.470448296567249,
+      "learning_rate": 1.8923670646089303e-05,
+      "loss": 0.218,
+      "step": 3433
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3832623440700667,
+      "learning_rate": 1.8922927231096482e-05,
+      "loss": 0.2101,
+      "step": 3434
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2149799232129805,
+      "learning_rate": 1.892218357406876e-05,
+      "loss": 0.2379,
+      "step": 3435
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2009435208243155,
+      "learning_rate": 1.89214396750263e-05,
+      "loss": 0.2305,
+      "step": 3436
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9612695423719364,
+      "learning_rate": 1.892069553398929e-05,
+      "loss": 0.2457,
+      "step": 3437
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1203701248483504,
+      "learning_rate": 1.8919951150977908e-05,
+      "loss": 0.2138,
+      "step": 3438
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1046369072938969,
+      "learning_rate": 1.8919206526012346e-05,
+      "loss": 0.222,
+      "step": 3439
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0531311533915446,
+      "learning_rate": 1.8918461659112805e-05,
+      "loss": 0.2276,
+      "step": 3440
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.291918364659948,
+      "learning_rate": 1.8917716550299485e-05,
+      "loss": 0.2109,
+      "step": 3441
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5619486000971663,
+      "learning_rate": 1.8916971199592603e-05,
+      "loss": 0.222,
+      "step": 3442
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.11363710814902,
+      "learning_rate": 1.891622560701237e-05,
+      "loss": 0.2221,
+      "step": 3443
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8271041690104084,
+      "learning_rate": 1.8915479772579017e-05,
+      "loss": 0.223,
+      "step": 3444
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1204943235932574,
+      "learning_rate": 1.891473369631277e-05,
+      "loss": 0.2328,
+      "step": 3445
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0413365146492488,
+      "learning_rate": 1.891398737823387e-05,
+      "loss": 0.2104,
+      "step": 3446
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2649950767159048,
+      "learning_rate": 1.8913240818362556e-05,
+      "loss": 0.2119,
+      "step": 3447
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4830297934477925,
+      "learning_rate": 1.8912494016719084e-05,
+      "loss": 0.2071,
+      "step": 3448
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1955636236263514,
+      "learning_rate": 1.8911746973323706e-05,
+      "loss": 0.2253,
+      "step": 3449
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.268887849540954,
+      "learning_rate": 1.8910999688196688e-05,
+      "loss": 0.2314,
+      "step": 3450
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8540019787261037,
+      "learning_rate": 1.8910252161358302e-05,
+      "loss": 0.2186,
+      "step": 3451
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3500922534295132,
+      "learning_rate": 1.8909504392828822e-05,
+      "loss": 0.193,
+      "step": 3452
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1737633956260982,
+      "learning_rate": 1.8908756382628534e-05,
+      "loss": 0.2178,
+      "step": 3453
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5649160231502703,
+      "learning_rate": 1.8908008130777724e-05,
+      "loss": 0.2198,
+      "step": 3454
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.954757815776317,
+      "learning_rate": 1.890725963729669e-05,
+      "loss": 0.2328,
+      "step": 3455
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5228254564372452,
+      "learning_rate": 1.8906510902205736e-05,
+      "loss": 0.2107,
+      "step": 3456
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0442097517235953,
+      "learning_rate": 1.890576192552517e-05,
+      "loss": 0.1938,
+      "step": 3457
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3410754929651154,
+      "learning_rate": 1.890501270727531e-05,
+      "loss": 0.205,
+      "step": 3458
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1872730608250062,
+      "learning_rate": 1.8904263247476478e-05,
+      "loss": 0.2256,
+      "step": 3459
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.085472544504074,
+      "learning_rate": 1.8903513546149e-05,
+      "loss": 0.1979,
+      "step": 3460
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.133249139850197,
+      "learning_rate": 1.8902763603313213e-05,
+      "loss": 0.2256,
+      "step": 3461
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.24180581737015,
+      "learning_rate": 1.8902013418989464e-05,
+      "loss": 0.2318,
+      "step": 3462
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8748478086219742,
+      "learning_rate": 1.89012629931981e-05,
+      "loss": 0.2209,
+      "step": 3463
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6952368589548503,
+      "learning_rate": 1.890051232595947e-05,
+      "loss": 0.2124,
+      "step": 3464
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1286193647776563,
+      "learning_rate": 1.8899761417293944e-05,
+      "loss": 0.214,
+      "step": 3465
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3691731443740385,
+      "learning_rate": 1.8899010267221884e-05,
+      "loss": 0.236,
+      "step": 3466
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.106432902418719,
+      "learning_rate": 1.8898258875763668e-05,
+      "loss": 0.1958,
+      "step": 3467
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.42216513069837,
+      "learning_rate": 1.889750724293968e-05,
+      "loss": 0.2178,
+      "step": 3468
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.039236057239844,
+      "learning_rate": 1.88967553687703e-05,
+      "loss": 0.211,
+      "step": 3469
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1630241938144983,
+      "learning_rate": 1.8896003253275934e-05,
+      "loss": 0.2031,
+      "step": 3470
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.133200488856877,
+      "learning_rate": 1.8895250896476976e-05,
+      "loss": 0.2157,
+      "step": 3471
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4087899695633774,
+      "learning_rate": 1.8894498298393835e-05,
+      "loss": 0.2696,
+      "step": 3472
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7305735648730336,
+      "learning_rate": 1.8893745459046922e-05,
+      "loss": 0.2323,
+      "step": 3473
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.142204157272831,
+      "learning_rate": 1.8892992378456664e-05,
+      "loss": 0.2282,
+      "step": 3474
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1118952952593542,
+      "learning_rate": 1.8892239056643487e-05,
+      "loss": 0.2036,
+      "step": 3475
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0687130740086,
+      "learning_rate": 1.8891485493627823e-05,
+      "loss": 0.2208,
+      "step": 3476
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6734111624578498,
+      "learning_rate": 1.8890731689430108e-05,
+      "loss": 0.2395,
+      "step": 3477
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3438494750974113,
+      "learning_rate": 1.88899776440708e-05,
+      "loss": 0.2236,
+      "step": 3478
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2048291784648975,
+      "learning_rate": 1.8889223357570342e-05,
+      "loss": 0.2327,
+      "step": 3479
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0046820562623493,
+      "learning_rate": 1.88884688299492e-05,
+      "loss": 0.2235,
+      "step": 3480
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5238429554318127,
+      "learning_rate": 1.8887714061227838e-05,
+      "loss": 0.2112,
+      "step": 3481
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.213570258302131,
+      "learning_rate": 1.8886959051426733e-05,
+      "loss": 0.1763,
+      "step": 3482
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0069459742245659,
+      "learning_rate": 1.8886203800566357e-05,
+      "loss": 0.2067,
+      "step": 3483
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.198997190207289,
+      "learning_rate": 1.8885448308667204e-05,
+      "loss": 0.2202,
+      "step": 3484
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1565023181231882,
+      "learning_rate": 1.8884692575749762e-05,
+      "loss": 0.2215,
+      "step": 3485
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2061530708342951,
+      "learning_rate": 1.8883936601834533e-05,
+      "loss": 0.1979,
+      "step": 3486
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0102406613321953,
+      "learning_rate": 1.8883180386942022e-05,
+      "loss": 0.2072,
+      "step": 3487
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6719172697318467,
+      "learning_rate": 1.8882423931092745e-05,
+      "loss": 0.2198,
+      "step": 3488
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.55648843029096,
+      "learning_rate": 1.888166723430721e-05,
+      "loss": 0.2062,
+      "step": 3489
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2375859182455549,
+      "learning_rate": 1.8880910296605956e-05,
+      "loss": 0.1965,
+      "step": 3490
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2731145678375624,
+      "learning_rate": 1.8880153118009505e-05,
+      "loss": 0.2256,
+      "step": 3491
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.164193051691617,
+      "learning_rate": 1.8879395698538402e-05,
+      "loss": 0.2163,
+      "step": 3492
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3166783793707386,
+      "learning_rate": 1.8878638038213186e-05,
+      "loss": 0.2129,
+      "step": 3493
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1521288362770579,
+      "learning_rate": 1.8877880137054413e-05,
+      "loss": 0.2269,
+      "step": 3494
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1986007406716817,
+      "learning_rate": 1.8877121995082638e-05,
+      "loss": 0.223,
+      "step": 3495
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0455207674415665,
+      "learning_rate": 1.887636361231843e-05,
+      "loss": 0.2218,
+      "step": 3496
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.975037927396976,
+      "learning_rate": 1.887560498878236e-05,
+      "loss": 0.205,
+      "step": 3497
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.99183527876214,
+      "learning_rate": 1.8874846124495e-05,
+      "loss": 0.2056,
+      "step": 3498
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3402372514211767,
+      "learning_rate": 1.8874087019476937e-05,
+      "loss": 0.2011,
+      "step": 3499
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3365592978089498,
+      "learning_rate": 1.8873327673748765e-05,
+      "loss": 0.2356,
+      "step": 3500
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4976108619117818,
+      "learning_rate": 1.8872568087331074e-05,
+      "loss": 0.2298,
+      "step": 3501
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0639877022857973,
+      "learning_rate": 1.8871808260244476e-05,
+      "loss": 0.2193,
+      "step": 3502
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.029750538822256,
+      "learning_rate": 1.8871048192509576e-05,
+      "loss": 0.2353,
+      "step": 3503
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5591330622484874,
+      "learning_rate": 1.8870287884147e-05,
+      "loss": 0.2064,
+      "step": 3504
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2796091932736209,
+      "learning_rate": 1.8869527335177354e-05,
+      "loss": 0.2481,
+      "step": 3505
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1622857079794304,
+      "learning_rate": 1.8868766545621286e-05,
+      "loss": 0.2504,
+      "step": 3506
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1477897384360691,
+      "learning_rate": 1.886800551549942e-05,
+      "loss": 0.2014,
+      "step": 3507
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1667763938280227,
+      "learning_rate": 1.88672442448324e-05,
+      "loss": 0.2078,
+      "step": 3508
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.137503241315197,
+      "learning_rate": 1.886648273364089e-05,
+      "loss": 0.2268,
+      "step": 3509
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5931312444608623,
+      "learning_rate": 1.886572098194553e-05,
+      "loss": 0.2399,
+      "step": 3510
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.998589900804111,
+      "learning_rate": 1.8864958989766982e-05,
+      "loss": 0.2177,
+      "step": 3511
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0766991349805863,
+      "learning_rate": 1.8864196757125926e-05,
+      "loss": 0.219,
+      "step": 3512
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2669927981226383,
+      "learning_rate": 1.886343428404303e-05,
+      "loss": 0.2455,
+      "step": 3513
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3821721941809648,
+      "learning_rate": 1.8862671570538983e-05,
+      "loss": 0.2291,
+      "step": 3514
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9922061065377965,
+      "learning_rate": 1.8861908616634465e-05,
+      "loss": 0.2355,
+      "step": 3515
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9647774432744618,
+      "learning_rate": 1.8861145422350175e-05,
+      "loss": 0.2225,
+      "step": 3516
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2450905932652228,
+      "learning_rate": 1.8860381987706815e-05,
+      "loss": 0.1923,
+      "step": 3517
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.174949296255601,
+      "learning_rate": 1.8859618312725097e-05,
+      "loss": 0.2273,
+      "step": 3518
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9185629784663724,
+      "learning_rate": 1.885885439742573e-05,
+      "loss": 0.2421,
+      "step": 3519
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0252428894923662,
+      "learning_rate": 1.8858090241829435e-05,
+      "loss": 0.2276,
+      "step": 3520
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3017122731058193,
+      "learning_rate": 1.8857325845956943e-05,
+      "loss": 0.2324,
+      "step": 3521
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3801729559240783,
+      "learning_rate": 1.8856561209828985e-05,
+      "loss": 0.2102,
+      "step": 3522
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.518892942443358,
+      "learning_rate": 1.8855796333466306e-05,
+      "loss": 0.2125,
+      "step": 3523
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.623769747136324,
+      "learning_rate": 1.8855031216889654e-05,
+      "loss": 0.2158,
+      "step": 3524
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0182322618961068,
+      "learning_rate": 1.8854265860119777e-05,
+      "loss": 0.2311,
+      "step": 3525
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6124376494565882,
+      "learning_rate": 1.8853500263177438e-05,
+      "loss": 0.243,
+      "step": 3526
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3650878655188263,
+      "learning_rate": 1.8852734426083407e-05,
+      "loss": 0.2351,
+      "step": 3527
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3001949420466634,
+      "learning_rate": 1.8851968348858452e-05,
+      "loss": 0.2381,
+      "step": 3528
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0896299953265736,
+      "learning_rate": 1.8851202031523357e-05,
+      "loss": 0.1966,
+      "step": 3529
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.149107350644368,
+      "learning_rate": 1.8850435474098903e-05,
+      "loss": 0.2114,
+      "step": 3530
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1056201887813963,
+      "learning_rate": 1.8849668676605892e-05,
+      "loss": 0.2094,
+      "step": 3531
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0996511570022964,
+      "learning_rate": 1.884890163906512e-05,
+      "loss": 0.2075,
+      "step": 3532
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2245004022629145,
+      "learning_rate": 1.8848134361497385e-05,
+      "loss": 0.2215,
+      "step": 3533
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9834659334969303,
+      "learning_rate": 1.8847366843923512e-05,
+      "loss": 0.2498,
+      "step": 3534
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.034031100069306,
+      "learning_rate": 1.8846599086364307e-05,
+      "loss": 0.2281,
+      "step": 3535
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1124308025623266,
+      "learning_rate": 1.8845831088840607e-05,
+      "loss": 0.2255,
+      "step": 3536
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9862681621323288,
+      "learning_rate": 1.884506285137324e-05,
+      "loss": 0.205,
+      "step": 3537
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9765289155940603,
+      "learning_rate": 1.884429437398304e-05,
+      "loss": 0.2293,
+      "step": 3538
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1166529919622326,
+      "learning_rate": 1.8843525656690856e-05,
+      "loss": 0.2419,
+      "step": 3539
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.310916406449081,
+      "learning_rate": 1.884275669951754e-05,
+      "loss": 0.2174,
+      "step": 3540
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1772746264552305,
+      "learning_rate": 1.8841987502483947e-05,
+      "loss": 0.2124,
+      "step": 3541
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.001411267184766,
+      "learning_rate": 1.8841218065610946e-05,
+      "loss": 0.2046,
+      "step": 3542
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7958941047302297,
+      "learning_rate": 1.8840448388919404e-05,
+      "loss": 0.2354,
+      "step": 3543
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3784337374121243,
+      "learning_rate": 1.8839678472430202e-05,
+      "loss": 0.2234,
+      "step": 3544
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.386781588632047,
+      "learning_rate": 1.883890831616422e-05,
+      "loss": 0.2139,
+      "step": 3545
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0081326543848246,
+      "learning_rate": 1.8838137920142353e-05,
+      "loss": 0.2172,
+      "step": 3546
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.925414447020237,
+      "learning_rate": 1.8837367284385495e-05,
+      "loss": 0.2212,
+      "step": 3547
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3165789878342848,
+      "learning_rate": 1.883659640891455e-05,
+      "loss": 0.218,
+      "step": 3548
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4404730700778872,
+      "learning_rate": 1.883582529375043e-05,
+      "loss": 0.2222,
+      "step": 3549
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5777733140194132,
+      "learning_rate": 1.883505393891405e-05,
+      "loss": 0.247,
+      "step": 3550
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.363219632836876,
+      "learning_rate": 1.883428234442633e-05,
+      "loss": 0.2386,
+      "step": 3551
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3569503082449117,
+      "learning_rate": 1.8833510510308205e-05,
+      "loss": 0.2317,
+      "step": 3552
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2587246937536893,
+      "learning_rate": 1.8832738436580606e-05,
+      "loss": 0.2188,
+      "step": 3553
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.12233768922461,
+      "learning_rate": 1.883196612326448e-05,
+      "loss": 0.2236,
+      "step": 3554
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2126151051536773,
+      "learning_rate": 1.8831193570380773e-05,
+      "loss": 0.2086,
+      "step": 3555
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2393899252829297,
+      "learning_rate": 1.883042077795044e-05,
+      "loss": 0.2391,
+      "step": 3556
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4836398610286818,
+      "learning_rate": 1.8829647745994445e-05,
+      "loss": 0.2554,
+      "step": 3557
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9733825299769355,
+      "learning_rate": 1.882887447453376e-05,
+      "loss": 0.2076,
+      "step": 3558
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.169740840648058,
+      "learning_rate": 1.8828100963589357e-05,
+      "loss": 0.1992,
+      "step": 3559
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0783743006030566,
+      "learning_rate": 1.882732721318221e-05,
+      "loss": 0.2232,
+      "step": 3560
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2929527357581567,
+      "learning_rate": 1.882655322333332e-05,
+      "loss": 0.2312,
+      "step": 3561
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8276497602849175,
+      "learning_rate": 1.8825778994063672e-05,
+      "loss": 0.1998,
+      "step": 3562
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.822612547922766,
+      "learning_rate": 1.882500452539427e-05,
+      "loss": 0.2186,
+      "step": 3563
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.185368918244924,
+      "learning_rate": 1.8824229817346124e-05,
+      "loss": 0.2269,
+      "step": 3564
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.044763669380613,
+      "learning_rate": 1.8823454869940243e-05,
+      "loss": 0.2022,
+      "step": 3565
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6220896032433259,
+      "learning_rate": 1.8822679683197654e-05,
+      "loss": 0.2309,
+      "step": 3566
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9932681947347778,
+      "learning_rate": 1.882190425713938e-05,
+      "loss": 0.2171,
+      "step": 3567
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3632488336960016,
+      "learning_rate": 1.882112859178645e-05,
+      "loss": 0.2188,
+      "step": 3568
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6684759325774463,
+      "learning_rate": 1.8820352687159912e-05,
+      "loss": 0.2151,
+      "step": 3569
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1266573469024315,
+      "learning_rate": 1.881957654328081e-05,
+      "loss": 0.2164,
+      "step": 3570
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2447540550144214,
+      "learning_rate": 1.8818800160170193e-05,
+      "loss": 0.2361,
+      "step": 3571
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3002904781282625,
+      "learning_rate": 1.8818023537849124e-05,
+      "loss": 0.2265,
+      "step": 3572
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0874912434200428,
+      "learning_rate": 1.8817246676338674e-05,
+      "loss": 0.2042,
+      "step": 3573
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2065451355021113,
+      "learning_rate": 1.8816469575659905e-05,
+      "loss": 0.1963,
+      "step": 3574
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6011427686798987,
+      "learning_rate": 1.8815692235833903e-05,
+      "loss": 0.2542,
+      "step": 3575
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0630131633964672,
+      "learning_rate": 1.881491465688175e-05,
+      "loss": 0.1862,
+      "step": 3576
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.192757320656651,
+      "learning_rate": 1.881413683882454e-05,
+      "loss": 0.2016,
+      "step": 3577
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0317469350067587,
+      "learning_rate": 1.881335878168337e-05,
+      "loss": 0.2074,
+      "step": 3578
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.239604331744985,
+      "learning_rate": 1.881258048547934e-05,
+      "loss": 0.2103,
+      "step": 3579
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3107029936706969,
+      "learning_rate": 1.8811801950233576e-05,
+      "loss": 0.2162,
+      "step": 3580
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0157843501686425,
+      "learning_rate": 1.881102317596718e-05,
+      "loss": 0.2354,
+      "step": 3581
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9623200968488543,
+      "learning_rate": 1.8810244162701282e-05,
+      "loss": 0.229,
+      "step": 3582
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1685273424380547,
+      "learning_rate": 1.8809464910457018e-05,
+      "loss": 0.244,
+      "step": 3583
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1103339740841685,
+      "learning_rate": 1.880868541925552e-05,
+      "loss": 0.2169,
+      "step": 3584
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.311960033700126,
+      "learning_rate": 1.8807905689117932e-05,
+      "loss": 0.2723,
+      "step": 3585
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1836207348653984,
+      "learning_rate": 1.8807125720065402e-05,
+      "loss": 0.2174,
+      "step": 3586
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.006572134286583,
+      "learning_rate": 1.8806345512119094e-05,
+      "loss": 0.2188,
+      "step": 3587
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9780003684667589,
+      "learning_rate": 1.880556506530016e-05,
+      "loss": 0.2171,
+      "step": 3588
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1732560568868864,
+      "learning_rate": 1.8804784379629782e-05,
+      "loss": 0.1826,
+      "step": 3589
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9094677603104601,
+      "learning_rate": 1.880400345512913e-05,
+      "loss": 0.2436,
+      "step": 3590
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5987479896408254,
+      "learning_rate": 1.8803222291819384e-05,
+      "loss": 0.2424,
+      "step": 3591
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9268098947402248,
+      "learning_rate": 1.8802440889721738e-05,
+      "loss": 0.2332,
+      "step": 3592
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1590799031555827,
+      "learning_rate": 1.8801659248857387e-05,
+      "loss": 0.2174,
+      "step": 3593
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8734401012949662,
+      "learning_rate": 1.880087736924753e-05,
+      "loss": 0.1889,
+      "step": 3594
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9763292582262146,
+      "learning_rate": 1.8800095250913378e-05,
+      "loss": 0.2301,
+      "step": 3595
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1036759320154053,
+      "learning_rate": 1.8799312893876144e-05,
+      "loss": 0.2354,
+      "step": 3596
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2068573877020075,
+      "learning_rate": 1.8798530298157053e-05,
+      "loss": 0.2257,
+      "step": 3597
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3724862448248425,
+      "learning_rate": 1.879774746377733e-05,
+      "loss": 0.2239,
+      "step": 3598
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4278671837999257,
+      "learning_rate": 1.8796964390758208e-05,
+      "loss": 0.2295,
+      "step": 3599
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0753048332681465,
+      "learning_rate": 1.879618107912093e-05,
+      "loss": 0.2333,
+      "step": 3600
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.435980345300096,
+      "learning_rate": 1.8795397528886744e-05,
+      "loss": 0.195,
+      "step": 3601
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4710044713102228,
+      "learning_rate": 1.8794613740076905e-05,
+      "loss": 0.2436,
+      "step": 3602
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1614473226565285,
+      "learning_rate": 1.8793829712712674e-05,
+      "loss": 0.2332,
+      "step": 3603
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1045520516660092,
+      "learning_rate": 1.879304544681531e-05,
+      "loss": 0.224,
+      "step": 3604
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.319759137203225,
+      "learning_rate": 1.8792260942406093e-05,
+      "loss": 0.2268,
+      "step": 3605
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1487889779994411,
+      "learning_rate": 1.87914761995063e-05,
+      "loss": 0.2067,
+      "step": 3606
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3210062460129575,
+      "learning_rate": 1.8790691218137223e-05,
+      "loss": 0.1936,
+      "step": 3607
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0186501739958207,
+      "learning_rate": 1.8789905998320148e-05,
+      "loss": 0.2092,
+      "step": 3608
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1235348787736643,
+      "learning_rate": 1.8789120540076377e-05,
+      "loss": 0.2477,
+      "step": 3609
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8505143793933634,
+      "learning_rate": 1.8788334843427213e-05,
+      "loss": 0.2175,
+      "step": 3610
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0042245499894789,
+      "learning_rate": 1.878754890839397e-05,
+      "loss": 0.2083,
+      "step": 3611
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.088541889713761,
+      "learning_rate": 1.8786762734997967e-05,
+      "loss": 0.2161,
+      "step": 3612
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7685156296862572,
+      "learning_rate": 1.878597632326053e-05,
+      "loss": 0.2166,
+      "step": 3613
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.12394810734318,
+      "learning_rate": 1.8785189673202987e-05,
+      "loss": 0.2276,
+      "step": 3614
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3965025337178076,
+      "learning_rate": 1.8784402784846683e-05,
+      "loss": 0.2035,
+      "step": 3615
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0159742971193932,
+      "learning_rate": 1.8783615658212954e-05,
+      "loss": 0.2333,
+      "step": 3616
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3262982321738739,
+      "learning_rate": 1.8782828293323148e-05,
+      "loss": 0.2175,
+      "step": 3617
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9727403288284343,
+      "learning_rate": 1.8782040690198638e-05,
+      "loss": 0.2162,
+      "step": 3618
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.147172083948449,
+      "learning_rate": 1.8781252848860774e-05,
+      "loss": 0.1996,
+      "step": 3619
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6406618327820583,
+      "learning_rate": 1.878046476933093e-05,
+      "loss": 0.2053,
+      "step": 3620
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1381954215972985,
+      "learning_rate": 1.8779676451630483e-05,
+      "loss": 0.2293,
+      "step": 3621
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2442005673537484,
+      "learning_rate": 1.877888789578082e-05,
+      "loss": 0.22,
+      "step": 3622
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.12587064239475,
+      "learning_rate": 1.877809910180332e-05,
+      "loss": 0.2325,
+      "step": 3623
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9723723171760922,
+      "learning_rate": 1.8777310069719395e-05,
+      "loss": 0.216,
+      "step": 3624
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8174785819072351,
+      "learning_rate": 1.8776520799550432e-05,
+      "loss": 0.2109,
+      "step": 3625
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9122603631781808,
+      "learning_rate": 1.8775731291317848e-05,
+      "loss": 0.2309,
+      "step": 3626
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8848068716660894,
+      "learning_rate": 1.877494154504306e-05,
+      "loss": 0.2194,
+      "step": 3627
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9278675771741361,
+      "learning_rate": 1.8774151560747483e-05,
+      "loss": 0.2183,
+      "step": 3628
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9191651194789452,
+      "learning_rate": 1.8773361338452552e-05,
+      "loss": 0.2161,
+      "step": 3629
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0679724192636588,
+      "learning_rate": 1.87725708781797e-05,
+      "loss": 0.2244,
+      "step": 3630
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0051450683135599,
+      "learning_rate": 1.8771780179950365e-05,
+      "loss": 0.236,
+      "step": 3631
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.054895456949411,
+      "learning_rate": 1.8770989243785996e-05,
+      "loss": 0.2241,
+      "step": 3632
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9056539769716427,
+      "learning_rate": 1.8770198069708053e-05,
+      "loss": 0.2322,
+      "step": 3633
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0234613563851183,
+      "learning_rate": 1.8769406657737987e-05,
+      "loss": 0.2316,
+      "step": 3634
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9956590739601004,
+      "learning_rate": 1.8768615007897274e-05,
+      "loss": 0.199,
+      "step": 3635
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8869981627351374,
+      "learning_rate": 1.8767823120207382e-05,
+      "loss": 0.2109,
+      "step": 3636
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9934235815819739,
+      "learning_rate": 1.876703099468979e-05,
+      "loss": 0.2304,
+      "step": 3637
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4633457424825305,
+      "learning_rate": 1.8766238631365993e-05,
+      "loss": 0.2083,
+      "step": 3638
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1406976495870418,
+      "learning_rate": 1.8765446030257475e-05,
+      "loss": 0.2153,
+      "step": 3639
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9879296696273988,
+      "learning_rate": 1.8764653191385737e-05,
+      "loss": 0.2085,
+      "step": 3640
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.910282344054396,
+      "learning_rate": 1.876386011477229e-05,
+      "loss": 0.2094,
+      "step": 3641
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0136285939917482,
+      "learning_rate": 1.8763066800438638e-05,
+      "loss": 0.2312,
+      "step": 3642
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9982969903185347,
+      "learning_rate": 1.8762273248406308e-05,
+      "loss": 0.2199,
+      "step": 3643
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0218645304104221,
+      "learning_rate": 1.8761479458696817e-05,
+      "loss": 0.2435,
+      "step": 3644
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.801950250637673,
+      "learning_rate": 1.87606854313317e-05,
+      "loss": 0.2279,
+      "step": 3645
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9830240715600229,
+      "learning_rate": 1.87598911663325e-05,
+      "loss": 0.202,
+      "step": 3646
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.062684827845211,
+      "learning_rate": 1.8759096663720757e-05,
+      "loss": 0.2038,
+      "step": 3647
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8424173998687214,
+      "learning_rate": 1.8758301923518022e-05,
+      "loss": 0.2075,
+      "step": 3648
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2659538205110992,
+      "learning_rate": 1.8757506945745853e-05,
+      "loss": 0.2135,
+      "step": 3649
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8779279555202277,
+      "learning_rate": 1.875671173042581e-05,
+      "loss": 0.2124,
+      "step": 3650
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0724684529868374,
+      "learning_rate": 1.875591627757947e-05,
+      "loss": 0.2271,
+      "step": 3651
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.997713641099804,
+      "learning_rate": 1.8755120587228407e-05,
+      "loss": 0.211,
+      "step": 3652
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9509704006661915,
+      "learning_rate": 1.87543246593942e-05,
+      "loss": 0.2153,
+      "step": 3653
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9876044502985013,
+      "learning_rate": 1.8753528494098448e-05,
+      "loss": 0.2446,
+      "step": 3654
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9074009768086965,
+      "learning_rate": 1.8752732091362737e-05,
+      "loss": 0.2376,
+      "step": 3655
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8466323825950363,
+      "learning_rate": 1.8751935451208672e-05,
+      "loss": 0.2198,
+      "step": 3656
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9408926489341524,
+      "learning_rate": 1.875113857365787e-05,
+      "loss": 0.2191,
+      "step": 3657
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9072872274928622,
+      "learning_rate": 1.8750341458731934e-05,
+      "loss": 0.2319,
+      "step": 3658
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9542440451622091,
+      "learning_rate": 1.8749544106452493e-05,
+      "loss": 0.1976,
+      "step": 3659
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7645091657265274,
+      "learning_rate": 1.8748746516841173e-05,
+      "loss": 0.2083,
+      "step": 3660
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.156226521004843,
+      "learning_rate": 1.8747948689919613e-05,
+      "loss": 0.2487,
+      "step": 3661
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.955620834216678,
+      "learning_rate": 1.8747150625709447e-05,
+      "loss": 0.2347,
+      "step": 3662
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9942108737792252,
+      "learning_rate": 1.874635232423233e-05,
+      "loss": 0.2062,
+      "step": 3663
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.007020797262451,
+      "learning_rate": 1.874555378550991e-05,
+      "loss": 0.2038,
+      "step": 3664
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9402919883523149,
+      "learning_rate": 1.874475500956385e-05,
+      "loss": 0.2045,
+      "step": 3665
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3820371958137763,
+      "learning_rate": 1.8743955996415816e-05,
+      "loss": 0.2158,
+      "step": 3666
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3681512159679738,
+      "learning_rate": 1.8743156746087484e-05,
+      "loss": 0.2455,
+      "step": 3667
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8308239350203164,
+      "learning_rate": 1.874235725860053e-05,
+      "loss": 0.1835,
+      "step": 3668
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9181442338099888,
+      "learning_rate": 1.874155753397664e-05,
+      "loss": 0.2669,
+      "step": 3669
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5206039594663598,
+      "learning_rate": 1.874075757223751e-05,
+      "loss": 0.2338,
+      "step": 3670
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1480071255535726,
+      "learning_rate": 1.8739957373404835e-05,
+      "loss": 0.2121,
+      "step": 3671
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.389739264223165,
+      "learning_rate": 1.8739156937500323e-05,
+      "loss": 0.2205,
+      "step": 3672
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.238895768277098,
+      "learning_rate": 1.8738356264545685e-05,
+      "loss": 0.218,
+      "step": 3673
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9526703867083856,
+      "learning_rate": 1.8737555354562644e-05,
+      "loss": 0.2024,
+      "step": 3674
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8969813658287388,
+      "learning_rate": 1.873675420757292e-05,
+      "loss": 0.2295,
+      "step": 3675
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2168913052504897,
+      "learning_rate": 1.873595282359824e-05,
+      "loss": 0.2147,
+      "step": 3676
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.021148886487509,
+      "learning_rate": 1.873515120266035e-05,
+      "loss": 0.2271,
+      "step": 3677
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.089162382736061,
+      "learning_rate": 1.8734349344780985e-05,
+      "loss": 0.2538,
+      "step": 3678
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.087846460938777,
+      "learning_rate": 1.8733547249981904e-05,
+      "loss": 0.2473,
+      "step": 3679
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6091696891202787,
+      "learning_rate": 1.8732744918284866e-05,
+      "loss": 0.2228,
+      "step": 3680
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9178567817841584,
+      "learning_rate": 1.873194234971162e-05,
+      "loss": 0.1982,
+      "step": 3681
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9954008637114219,
+      "learning_rate": 1.8731139544283952e-05,
+      "loss": 0.1894,
+      "step": 3682
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8191302667217233,
+      "learning_rate": 1.873033650202363e-05,
+      "loss": 0.2321,
+      "step": 3683
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7547033099045546,
+      "learning_rate": 1.872953322295243e-05,
+      "loss": 0.2222,
+      "step": 3684
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0682176638270313,
+      "learning_rate": 1.8728729707092156e-05,
+      "loss": 0.2266,
+      "step": 3685
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0770283192110661,
+      "learning_rate": 1.8727925954464588e-05,
+      "loss": 0.209,
+      "step": 3686
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0130660718947984,
+      "learning_rate": 1.8727121965091542e-05,
+      "loss": 0.2055,
+      "step": 3687
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1315284288169722,
+      "learning_rate": 1.8726317738994817e-05,
+      "loss": 0.2378,
+      "step": 3688
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0757318688843531,
+      "learning_rate": 1.8725513276196232e-05,
+      "loss": 0.214,
+      "step": 3689
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8001330749908178,
+      "learning_rate": 1.8724708576717607e-05,
+      "loss": 0.2101,
+      "step": 3690
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2760143890057158,
+      "learning_rate": 1.872390364058077e-05,
+      "loss": 0.2274,
+      "step": 3691
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.198424221315609,
+      "learning_rate": 1.872309846780755e-05,
+      "loss": 0.2008,
+      "step": 3692
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.085392171878298,
+      "learning_rate": 1.8722293058419794e-05,
+      "loss": 0.2215,
+      "step": 3693
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0167219653500086,
+      "learning_rate": 1.8721487412439344e-05,
+      "loss": 0.2413,
+      "step": 3694
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6452475011732255,
+      "learning_rate": 1.8720681529888057e-05,
+      "loss": 0.2311,
+      "step": 3695
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.148723352911266,
+      "learning_rate": 1.8719875410787793e-05,
+      "loss": 0.2149,
+      "step": 3696
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9979650627009616,
+      "learning_rate": 1.8719069055160415e-05,
+      "loss": 0.2114,
+      "step": 3697
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.878102123527887,
+      "learning_rate": 1.8718262463027795e-05,
+      "loss": 0.2167,
+      "step": 3698
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8587460522038438,
+      "learning_rate": 1.8717455634411813e-05,
+      "loss": 0.2157,
+      "step": 3699
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9939679426048222,
+      "learning_rate": 1.8716648569334355e-05,
+      "loss": 0.2142,
+      "step": 3700
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0933099196445442,
+      "learning_rate": 1.8715841267817313e-05,
+      "loss": 0.24,
+      "step": 3701
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9072361235671731,
+      "learning_rate": 1.8715033729882585e-05,
+      "loss": 0.2267,
+      "step": 3702
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5972984097976861,
+      "learning_rate": 1.8714225955552074e-05,
+      "loss": 0.2271,
+      "step": 3703
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.195176661277982,
+      "learning_rate": 1.8713417944847688e-05,
+      "loss": 0.2464,
+      "step": 3704
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.964239160509171,
+      "learning_rate": 1.871260969779135e-05,
+      "loss": 0.224,
+      "step": 3705
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1009244798701132,
+      "learning_rate": 1.8711801214404986e-05,
+      "loss": 0.2277,
+      "step": 3706
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4679413308921536,
+      "learning_rate": 1.871099249471052e-05,
+      "loss": 0.2207,
+      "step": 3707
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4215098692749524,
+      "learning_rate": 1.871018353872989e-05,
+      "loss": 0.2172,
+      "step": 3708
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0315434809422932,
+      "learning_rate": 1.870937434648504e-05,
+      "loss": 0.221,
+      "step": 3709
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8370794377256259,
+      "learning_rate": 1.8708564917997917e-05,
+      "loss": 0.2127,
+      "step": 3710
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3562200726557778,
+      "learning_rate": 1.870775525329048e-05,
+      "loss": 0.2168,
+      "step": 3711
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0137792718910874,
+      "learning_rate": 1.870694535238469e-05,
+      "loss": 0.2214,
+      "step": 3712
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9076632568937337,
+      "learning_rate": 1.8706135215302517e-05,
+      "loss": 0.2201,
+      "step": 3713
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.878009141999452,
+      "learning_rate": 1.8705324842065933e-05,
+      "loss": 0.2232,
+      "step": 3714
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0509396224366285,
+      "learning_rate": 1.870451423269692e-05,
+      "loss": 0.212,
+      "step": 3715
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.043258822287421,
+      "learning_rate": 1.870370338721747e-05,
+      "loss": 0.212,
+      "step": 3716
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0121568442288489,
+      "learning_rate": 1.870289230564957e-05,
+      "loss": 0.2585,
+      "step": 3717
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0953299936197005,
+      "learning_rate": 1.870208098801523e-05,
+      "loss": 0.2228,
+      "step": 3718
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0607744241849235,
+      "learning_rate": 1.8701269434336447e-05,
+      "loss": 0.2489,
+      "step": 3719
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8472590526760381,
+      "learning_rate": 1.870045764463524e-05,
+      "loss": 0.2173,
+      "step": 3720
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0576878592684522,
+      "learning_rate": 1.8699645618933628e-05,
+      "loss": 0.2177,
+      "step": 3721
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0360486988527098,
+      "learning_rate": 1.869883335725364e-05,
+      "loss": 0.2238,
+      "step": 3722
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1969945924023768,
+      "learning_rate": 1.8698020859617303e-05,
+      "loss": 0.229,
+      "step": 3723
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1733345648184206,
+      "learning_rate": 1.869720812604666e-05,
+      "loss": 0.2289,
+      "step": 3724
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1734416691118115,
+      "learning_rate": 1.8696395156563753e-05,
+      "loss": 0.2121,
+      "step": 3725
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7741114933790696,
+      "learning_rate": 1.8695581951190636e-05,
+      "loss": 0.2237,
+      "step": 3726
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8425428534413996,
+      "learning_rate": 1.8694768509949374e-05,
+      "loss": 0.2082,
+      "step": 3727
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.908533200343883,
+      "learning_rate": 1.8693954832862017e-05,
+      "loss": 0.2088,
+      "step": 3728
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.084859595161081,
+      "learning_rate": 1.869314091995065e-05,
+      "loss": 0.191,
+      "step": 3729
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.840271807447316,
+      "learning_rate": 1.8692326771237344e-05,
+      "loss": 0.258,
+      "step": 3730
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9170431065446714,
+      "learning_rate": 1.8691512386744183e-05,
+      "loss": 0.225,
+      "step": 3731
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0010752864096515,
+      "learning_rate": 1.8690697766493252e-05,
+      "loss": 0.2519,
+      "step": 3732
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8686061614515938,
+      "learning_rate": 1.8689882910506658e-05,
+      "loss": 0.2076,
+      "step": 3733
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0749405681204596,
+      "learning_rate": 1.8689067818806503e-05,
+      "loss": 0.2127,
+      "step": 3734
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0258137289693825,
+      "learning_rate": 1.8688252491414886e-05,
+      "loss": 0.2108,
+      "step": 3735
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0969570804965412,
+      "learning_rate": 1.868743692835393e-05,
+      "loss": 0.1829,
+      "step": 3736
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.259486363094936,
+      "learning_rate": 1.868662112964576e-05,
+      "loss": 0.2184,
+      "step": 3737
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0370119946877436,
+      "learning_rate": 1.8685805095312498e-05,
+      "loss": 0.2101,
+      "step": 3738
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0349854099522016,
+      "learning_rate": 1.8684988825376285e-05,
+      "loss": 0.21,
+      "step": 3739
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2262944485867913,
+      "learning_rate": 1.8684172319859258e-05,
+      "loss": 0.2563,
+      "step": 3740
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1655799061302887,
+      "learning_rate": 1.8683355578783567e-05,
+      "loss": 0.2179,
+      "step": 3741
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4736756001428297,
+      "learning_rate": 1.868253860217137e-05,
+      "loss": 0.2407,
+      "step": 3742
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9410716912521949,
+      "learning_rate": 1.8681721390044817e-05,
+      "loss": 0.2069,
+      "step": 3743
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4025378391460672,
+      "learning_rate": 1.868090394242608e-05,
+      "loss": 0.2414,
+      "step": 3744
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.313299060245258,
+      "learning_rate": 1.8680086259337337e-05,
+      "loss": 0.2157,
+      "step": 3745
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.920574490398171,
+      "learning_rate": 1.8679268340800764e-05,
+      "loss": 0.2423,
+      "step": 3746
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0078079186913158,
+      "learning_rate": 1.8678450186838545e-05,
+      "loss": 0.2178,
+      "step": 3747
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8418987504966208,
+      "learning_rate": 1.8677631797472874e-05,
+      "loss": 0.2322,
+      "step": 3748
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2665665598701452,
+      "learning_rate": 1.867681317272595e-05,
+      "loss": 0.2214,
+      "step": 3749
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9829511796364977,
+      "learning_rate": 1.867599431261998e-05,
+      "loss": 0.2184,
+      "step": 3750
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3460645041914767,
+      "learning_rate": 1.8675175217177176e-05,
+      "loss": 0.2331,
+      "step": 3751
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9805157466720837,
+      "learning_rate": 1.867435588641975e-05,
+      "loss": 0.2293,
+      "step": 3752
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.234993024768991,
+      "learning_rate": 1.8673536320369936e-05,
+      "loss": 0.2311,
+      "step": 3753
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.373040649077192,
+      "learning_rate": 1.8672716519049957e-05,
+      "loss": 0.2258,
+      "step": 3754
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0665080193056045,
+      "learning_rate": 1.867189648248205e-05,
+      "loss": 0.2163,
+      "step": 3755
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.42465957114889,
+      "learning_rate": 1.8671076210688467e-05,
+      "loss": 0.2035,
+      "step": 3756
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2238898901999282,
+      "learning_rate": 1.867025570369145e-05,
+      "loss": 0.2401,
+      "step": 3757
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.176117512236095,
+      "learning_rate": 1.8669434961513256e-05,
+      "loss": 0.2117,
+      "step": 3758
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1875416816543796,
+      "learning_rate": 1.866861398417615e-05,
+      "loss": 0.2207,
+      "step": 3759
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2449151104729497,
+      "learning_rate": 1.8667792771702397e-05,
+      "loss": 0.2156,
+      "step": 3760
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8601865869175008,
+      "learning_rate": 1.8666971324114277e-05,
+      "loss": 0.1938,
+      "step": 3761
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0584334643359323,
+      "learning_rate": 1.866614964143407e-05,
+      "loss": 0.2206,
+      "step": 3762
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9599952315995963,
+      "learning_rate": 1.8665327723684065e-05,
+      "loss": 0.2334,
+      "step": 3763
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.068548285779253,
+      "learning_rate": 1.8664505570886557e-05,
+      "loss": 0.2138,
+      "step": 3764
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9137698886234791,
+      "learning_rate": 1.8663683183063846e-05,
+      "loss": 0.2418,
+      "step": 3765
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8822421507354169,
+      "learning_rate": 1.8662860560238238e-05,
+      "loss": 0.2357,
+      "step": 3766
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9839602464096109,
+      "learning_rate": 1.866203770243205e-05,
+      "loss": 0.219,
+      "step": 3767
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.017921041143888,
+      "learning_rate": 1.86612146096676e-05,
+      "loss": 0.1929,
+      "step": 3768
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9666760157968136,
+      "learning_rate": 1.8660391281967213e-05,
+      "loss": 0.2327,
+      "step": 3769
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2427749818036062,
+      "learning_rate": 1.8659567719353223e-05,
+      "loss": 0.2206,
+      "step": 3770
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0818309717884633,
+      "learning_rate": 1.865874392184797e-05,
+      "loss": 0.2248,
+      "step": 3771
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5483282083497667,
+      "learning_rate": 1.86579198894738e-05,
+      "loss": 0.2569,
+      "step": 3772
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0491292439109805,
+      "learning_rate": 1.8657095622253064e-05,
+      "loss": 0.2199,
+      "step": 3773
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.98191139928569,
+      "learning_rate": 1.8656271120208118e-05,
+      "loss": 0.2415,
+      "step": 3774
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8249343376451618,
+      "learning_rate": 1.8655446383361332e-05,
+      "loss": 0.2419,
+      "step": 3775
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0439989508003285,
+      "learning_rate": 1.8654621411735072e-05,
+      "loss": 0.2125,
+      "step": 3776
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9956541278544652,
+      "learning_rate": 1.8653796205351717e-05,
+      "loss": 0.2423,
+      "step": 3777
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7951566731952017,
+      "learning_rate": 1.865297076423365e-05,
+      "loss": 0.2042,
+      "step": 3778
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0345974375671212,
+      "learning_rate": 1.8652145088403267e-05,
+      "loss": 0.2209,
+      "step": 3779
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9969738180774795,
+      "learning_rate": 1.8651319177882957e-05,
+      "loss": 0.2287,
+      "step": 3780
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4389658031539585,
+      "learning_rate": 1.8650493032695124e-05,
+      "loss": 0.2236,
+      "step": 3781
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.134999809634415,
+      "learning_rate": 1.8649666652862183e-05,
+      "loss": 0.2247,
+      "step": 3782
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9820850253804223,
+      "learning_rate": 1.864884003840654e-05,
+      "loss": 0.2137,
+      "step": 3783
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8792425222701588,
+      "learning_rate": 1.8648013189350628e-05,
+      "loss": 0.2228,
+      "step": 3784
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2159676018930894,
+      "learning_rate": 1.864718610571687e-05,
+      "loss": 0.2443,
+      "step": 3785
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9169076017661532,
+      "learning_rate": 1.8646358787527697e-05,
+      "loss": 0.197,
+      "step": 3786
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1325155925959844,
+      "learning_rate": 1.8645531234805554e-05,
+      "loss": 0.2191,
+      "step": 3787
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2026712867680391,
+      "learning_rate": 1.864470344757289e-05,
+      "loss": 0.2246,
+      "step": 3788
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1107125406261882,
+      "learning_rate": 1.8643875425852155e-05,
+      "loss": 0.2394,
+      "step": 3789
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4123550933251017,
+      "learning_rate": 1.864304716966581e-05,
+      "loss": 0.2606,
+      "step": 3790
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.325865887213283,
+      "learning_rate": 1.8642218679036324e-05,
+      "loss": 0.2172,
+      "step": 3791
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0620807363636982,
+      "learning_rate": 1.8641389953986165e-05,
+      "loss": 0.2409,
+      "step": 3792
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.071205185857897,
+      "learning_rate": 1.864056099453782e-05,
+      "loss": 0.2386,
+      "step": 3793
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8456964313282443,
+      "learning_rate": 1.8639731800713766e-05,
+      "loss": 0.1992,
+      "step": 3794
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1931661849786743,
+      "learning_rate": 1.8638902372536502e-05,
+      "loss": 0.2136,
+      "step": 3795
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.120405301713219,
+      "learning_rate": 1.8638072710028523e-05,
+      "loss": 0.2087,
+      "step": 3796
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9879170452045153,
+      "learning_rate": 1.8637242813212334e-05,
+      "loss": 0.2056,
+      "step": 3797
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2824930991489314,
+      "learning_rate": 1.8636412682110445e-05,
+      "loss": 0.2151,
+      "step": 3798
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.293469054776063,
+      "learning_rate": 1.8635582316745374e-05,
+      "loss": 0.2052,
+      "step": 3799
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.025378532115027,
+      "learning_rate": 1.8634751717139644e-05,
+      "loss": 0.2118,
+      "step": 3800
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7830661665580738,
+      "learning_rate": 1.863392088331579e-05,
+      "loss": 0.2128,
+      "step": 3801
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3021193311965715,
+      "learning_rate": 1.863308981529634e-05,
+      "loss": 0.2381,
+      "step": 3802
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.762346379086396,
+      "learning_rate": 1.863225851310384e-05,
+      "loss": 0.2063,
+      "step": 3803
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4515520631191818,
+      "learning_rate": 1.8631426976760844e-05,
+      "loss": 0.2143,
+      "step": 3804
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.171511742865271,
+      "learning_rate": 1.8630595206289904e-05,
+      "loss": 0.227,
+      "step": 3805
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8935558090507881,
+      "learning_rate": 1.862976320171358e-05,
+      "loss": 0.2127,
+      "step": 3806
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9868740593439831,
+      "learning_rate": 1.8628930963054444e-05,
+      "loss": 0.2191,
+      "step": 3807
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9324619016671554,
+      "learning_rate": 1.8628098490335064e-05,
+      "loss": 0.2184,
+      "step": 3808
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.002216078090249,
+      "learning_rate": 1.8627265783578028e-05,
+      "loss": 0.2105,
+      "step": 3809
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1699929258731216,
+      "learning_rate": 1.862643284280592e-05,
+      "loss": 0.2267,
+      "step": 3810
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.837422894813569,
+      "learning_rate": 1.8625599668041334e-05,
+      "loss": 0.1931,
+      "step": 3811
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9597112271853646,
+      "learning_rate": 1.862476625930687e-05,
+      "loss": 0.222,
+      "step": 3812
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9219677680827443,
+      "learning_rate": 1.8623932616625133e-05,
+      "loss": 0.2037,
+      "step": 3813
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9555404866784017,
+      "learning_rate": 1.8623098740018736e-05,
+      "loss": 0.244,
+      "step": 3814
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8565637533356807,
+      "learning_rate": 1.86222646295103e-05,
+      "loss": 0.2166,
+      "step": 3815
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0481091626759864,
+      "learning_rate": 1.8621430285122447e-05,
+      "loss": 0.231,
+      "step": 3816
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8774766428514378,
+      "learning_rate": 1.8620595706877816e-05,
+      "loss": 0.214,
+      "step": 3817
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9865676986752081,
+      "learning_rate": 1.8619760894799034e-05,
+      "loss": 0.2236,
+      "step": 3818
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2433378952932157,
+      "learning_rate": 1.8618925848908757e-05,
+      "loss": 0.2178,
+      "step": 3819
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8438590361016094,
+      "learning_rate": 1.861809056922962e-05,
+      "loss": 0.2188,
+      "step": 3820
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9083392931246398,
+      "learning_rate": 1.8617255055784298e-05,
+      "loss": 0.2152,
+      "step": 3821
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9041176626279603,
+      "learning_rate": 1.8616419308595443e-05,
+      "loss": 0.2471,
+      "step": 3822
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8617167966436966,
+      "learning_rate": 1.8615583327685726e-05,
+      "loss": 0.2185,
+      "step": 3823
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.831294907600138,
+      "learning_rate": 1.8614747113077826e-05,
+      "loss": 0.238,
+      "step": 3824
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1312425674849975,
+      "learning_rate": 1.8613910664794424e-05,
+      "loss": 0.2646,
+      "step": 3825
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9087120849416299,
+      "learning_rate": 1.861307398285821e-05,
+      "loss": 0.2043,
+      "step": 3826
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0711854562301435,
+      "learning_rate": 1.8612237067291878e-05,
+      "loss": 0.2137,
+      "step": 3827
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1537888299444026,
+      "learning_rate": 1.8611399918118124e-05,
+      "loss": 0.2275,
+      "step": 3828
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.053077637022216,
+      "learning_rate": 1.8610562535359667e-05,
+      "loss": 0.2573,
+      "step": 3829
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8794836263083736,
+      "learning_rate": 1.8609724919039213e-05,
+      "loss": 0.2217,
+      "step": 3830
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1927003302097705,
+      "learning_rate": 1.8608887069179483e-05,
+      "loss": 0.2239,
+      "step": 3831
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9961929880291288,
+      "learning_rate": 1.8608048985803205e-05,
+      "loss": 0.2101,
+      "step": 3832
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7476771727973698,
+      "learning_rate": 1.8607210668933114e-05,
+      "loss": 0.2028,
+      "step": 3833
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1708729140370346,
+      "learning_rate": 1.8606372118591943e-05,
+      "loss": 0.2418,
+      "step": 3834
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.95154019222484,
+      "learning_rate": 1.8605533334802448e-05,
+      "loss": 0.22,
+      "step": 3835
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9420981207262242,
+      "learning_rate": 1.8604694317587372e-05,
+      "loss": 0.1962,
+      "step": 3836
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7574527137216649,
+      "learning_rate": 1.8603855066969478e-05,
+      "loss": 0.2019,
+      "step": 3837
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1338212264538146,
+      "learning_rate": 1.860301558297153e-05,
+      "loss": 0.234,
+      "step": 3838
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9869099042527179,
+      "learning_rate": 1.8602175865616296e-05,
+      "loss": 0.233,
+      "step": 3839
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9117557983918922,
+      "learning_rate": 1.8601335914926558e-05,
+      "loss": 0.2072,
+      "step": 3840
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0213142783338018,
+      "learning_rate": 1.8600495730925095e-05,
+      "loss": 0.2166,
+      "step": 3841
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0960729307387658,
+      "learning_rate": 1.8599655313634702e-05,
+      "loss": 0.1988,
+      "step": 3842
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2800084319729785,
+      "learning_rate": 1.859881466307817e-05,
+      "loss": 0.2403,
+      "step": 3843
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.643017787270073,
+      "learning_rate": 1.8597973779278307e-05,
+      "loss": 0.2231,
+      "step": 3844
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0880616948226376,
+      "learning_rate": 1.859713266225792e-05,
+      "loss": 0.2414,
+      "step": 3845
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9851831423429016,
+      "learning_rate": 1.859629131203982e-05,
+      "loss": 0.2098,
+      "step": 3846
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9979121868241856,
+      "learning_rate": 1.859544972864684e-05,
+      "loss": 0.223,
+      "step": 3847
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9845690582968679,
+      "learning_rate": 1.8594607912101797e-05,
+      "loss": 0.2172,
+      "step": 3848
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2156036882618944,
+      "learning_rate": 1.8593765862427526e-05,
+      "loss": 0.2268,
+      "step": 3849
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.037311775310428,
+      "learning_rate": 1.8592923579646874e-05,
+      "loss": 0.2036,
+      "step": 3850
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.254690899134614,
+      "learning_rate": 1.8592081063782685e-05,
+      "loss": 0.2185,
+      "step": 3851
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3888197734442886,
+      "learning_rate": 1.8591238314857806e-05,
+      "loss": 0.2096,
+      "step": 3852
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7310357928446478,
+      "learning_rate": 1.859039533289511e-05,
+      "loss": 0.2426,
+      "step": 3853
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8752949059665149,
+      "learning_rate": 1.858955211791745e-05,
+      "loss": 0.2248,
+      "step": 3854
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9799619031728044,
+      "learning_rate": 1.858870866994771e-05,
+      "loss": 0.2234,
+      "step": 3855
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8881141779724255,
+      "learning_rate": 1.8587864989008758e-05,
+      "loss": 0.2308,
+      "step": 3856
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.929858899344705,
+      "learning_rate": 1.8587021075123482e-05,
+      "loss": 0.1902,
+      "step": 3857
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8424469468404533,
+      "learning_rate": 1.8586176928314774e-05,
+      "loss": 0.2072,
+      "step": 3858
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7706289278768073,
+      "learning_rate": 1.858533254860553e-05,
+      "loss": 0.223,
+      "step": 3859
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2901367072560834,
+      "learning_rate": 1.8584487936018663e-05,
+      "loss": 0.2194,
+      "step": 3860
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9724146249544461,
+      "learning_rate": 1.8583643090577072e-05,
+      "loss": 0.2297,
+      "step": 3861
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9272187008477141,
+      "learning_rate": 1.8582798012303674e-05,
+      "loss": 0.2279,
+      "step": 3862
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8680660636380514,
+      "learning_rate": 1.85819527012214e-05,
+      "loss": 0.2178,
+      "step": 3863
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0452921030928661,
+      "learning_rate": 1.8581107157353175e-05,
+      "loss": 0.2579,
+      "step": 3864
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8615649767726768,
+      "learning_rate": 1.8580261380721932e-05,
+      "loss": 0.2024,
+      "step": 3865
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2325980969486496,
+      "learning_rate": 1.8579415371350613e-05,
+      "loss": 0.2322,
+      "step": 3866
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8933378457355367,
+      "learning_rate": 1.8578569129262168e-05,
+      "loss": 0.2031,
+      "step": 3867
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9453357111940053,
+      "learning_rate": 1.857772265447955e-05,
+      "loss": 0.2287,
+      "step": 3868
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9394950082674527,
+      "learning_rate": 1.8576875947025725e-05,
+      "loss": 0.2162,
+      "step": 3869
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3055556466692884,
+      "learning_rate": 1.8576029006923653e-05,
+      "loss": 0.2314,
+      "step": 3870
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9320739710953067,
+      "learning_rate": 1.8575181834196308e-05,
+      "loss": 0.2099,
+      "step": 3871
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2266250587407823,
+      "learning_rate": 1.857433442886667e-05,
+      "loss": 0.2115,
+      "step": 3872
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.034361450688478,
+      "learning_rate": 1.8573486790957732e-05,
+      "loss": 0.2377,
+      "step": 3873
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.061886386036358,
+      "learning_rate": 1.8572638920492476e-05,
+      "loss": 0.2303,
+      "step": 3874
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9135745761327706,
+      "learning_rate": 1.857179081749391e-05,
+      "loss": 0.227,
+      "step": 3875
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.692975173577015,
+      "learning_rate": 1.8570942481985027e-05,
+      "loss": 0.2338,
+      "step": 3876
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.114073471726701,
+      "learning_rate": 1.857009391398885e-05,
+      "loss": 0.2414,
+      "step": 3877
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8604903674257696,
+      "learning_rate": 1.856924511352839e-05,
+      "loss": 0.2341,
+      "step": 3878
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0061238541965627,
+      "learning_rate": 1.8568396080626673e-05,
+      "loss": 0.2605,
+      "step": 3879
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8186410015687037,
+      "learning_rate": 1.8567546815306726e-05,
+      "loss": 0.2205,
+      "step": 3880
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9094401156407541,
+      "learning_rate": 1.856669731759159e-05,
+      "loss": 0.2217,
+      "step": 3881
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8157865383260736,
+      "learning_rate": 1.8565847587504305e-05,
+      "loss": 0.2128,
+      "step": 3882
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0473730645092414,
+      "learning_rate": 1.856499762506792e-05,
+      "loss": 0.213,
+      "step": 3883
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9802341644805417,
+      "learning_rate": 1.8564147430305493e-05,
+      "loss": 0.272,
+      "step": 3884
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9530522304023589,
+      "learning_rate": 1.8563297003240078e-05,
+      "loss": 0.225,
+      "step": 3885
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9902349476942686,
+      "learning_rate": 1.8562446343894753e-05,
+      "loss": 0.214,
+      "step": 3886
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9234130370394749,
+      "learning_rate": 1.8561595452292587e-05,
+      "loss": 0.2148,
+      "step": 3887
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.948780172751885,
+      "learning_rate": 1.8560744328456657e-05,
+      "loss": 0.2093,
+      "step": 3888
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1596350769709731,
+      "learning_rate": 1.855989297241006e-05,
+      "loss": 0.2314,
+      "step": 3889
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8493065367210507,
+      "learning_rate": 1.8559041384175874e-05,
+      "loss": 0.2287,
+      "step": 3890
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0826434951585162,
+      "learning_rate": 1.8558189563777213e-05,
+      "loss": 0.2296,
+      "step": 3891
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.233129074251226,
+      "learning_rate": 1.8557337511237178e-05,
+      "loss": 0.2124,
+      "step": 3892
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.275827478065935,
+      "learning_rate": 1.8556485226578877e-05,
+      "loss": 0.2059,
+      "step": 3893
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0039858133301296,
+      "learning_rate": 1.8555632709825427e-05,
+      "loss": 0.2189,
+      "step": 3894
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.298069959049279,
+      "learning_rate": 1.8554779960999963e-05,
+      "loss": 0.2339,
+      "step": 3895
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9895375285393244,
+      "learning_rate": 1.8553926980125608e-05,
+      "loss": 0.2092,
+      "step": 3896
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2551382881470876,
+      "learning_rate": 1.85530737672255e-05,
+      "loss": 0.2152,
+      "step": 3897
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0977374431389633,
+      "learning_rate": 1.8552220322322782e-05,
+      "loss": 0.235,
+      "step": 3898
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1044374060859814,
+      "learning_rate": 1.8551366645440604e-05,
+      "loss": 0.2008,
+      "step": 3899
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2014769038354687,
+      "learning_rate": 1.8550512736602125e-05,
+      "loss": 0.2217,
+      "step": 3900
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9532092971485372,
+      "learning_rate": 1.8549658595830505e-05,
+      "loss": 0.2228,
+      "step": 3901
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.841470746178915,
+      "learning_rate": 1.8548804223148914e-05,
+      "loss": 0.2394,
+      "step": 3902
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1083325341107813,
+      "learning_rate": 1.854794961858052e-05,
+      "loss": 0.2163,
+      "step": 3903
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9741102853470924,
+      "learning_rate": 1.8547094782148513e-05,
+      "loss": 0.2372,
+      "step": 3904
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9752661376202427,
+      "learning_rate": 1.8546239713876077e-05,
+      "loss": 0.2322,
+      "step": 3905
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.077642457000483,
+      "learning_rate": 1.8545384413786406e-05,
+      "loss": 0.2166,
+      "step": 3906
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.112417457963183,
+      "learning_rate": 1.85445288819027e-05,
+      "loss": 0.2521,
+      "step": 3907
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.133826703746313,
+      "learning_rate": 1.8543673118248167e-05,
+      "loss": 0.207,
+      "step": 3908
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9714603769415149,
+      "learning_rate": 1.854281712284602e-05,
+      "loss": 0.2076,
+      "step": 3909
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0225184907059015,
+      "learning_rate": 1.8541960895719473e-05,
+      "loss": 0.228,
+      "step": 3910
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.420720984524292,
+      "learning_rate": 1.8541104436891756e-05,
+      "loss": 0.2016,
+      "step": 3911
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9235109665556697,
+      "learning_rate": 1.8540247746386095e-05,
+      "loss": 0.2355,
+      "step": 3912
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0854904386542399,
+      "learning_rate": 1.8539390824225735e-05,
+      "loss": 0.2196,
+      "step": 3913
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9777382832585857,
+      "learning_rate": 1.8538533670433912e-05,
+      "loss": 0.2034,
+      "step": 3914
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0064862201866491,
+      "learning_rate": 1.8537676285033886e-05,
+      "loss": 0.1908,
+      "step": 3915
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9964750671495186,
+      "learning_rate": 1.8536818668048906e-05,
+      "loss": 0.2261,
+      "step": 3916
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0298903324507094,
+      "learning_rate": 1.8535960819502243e-05,
+      "loss": 0.2136,
+      "step": 3917
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.023825717413679,
+      "learning_rate": 1.853510273941715e-05,
+      "loss": 0.2548,
+      "step": 3918
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0799038437526578,
+      "learning_rate": 1.8534244427816924e-05,
+      "loss": 0.2313,
+      "step": 3919
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.683700267828652,
+      "learning_rate": 1.853338588472483e-05,
+      "loss": 0.2098,
+      "step": 3920
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.002263888788098,
+      "learning_rate": 1.8532527110164163e-05,
+      "loss": 0.2272,
+      "step": 3921
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8523969928905435,
+      "learning_rate": 1.8531668104158215e-05,
+      "loss": 0.1953,
+      "step": 3922
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9766389662463112,
+      "learning_rate": 1.853080886673029e-05,
+      "loss": 0.2032,
+      "step": 3923
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9793413109004094,
+      "learning_rate": 1.8529949397903692e-05,
+      "loss": 0.2229,
+      "step": 3924
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0853519331853911,
+      "learning_rate": 1.8529089697701735e-05,
+      "loss": 0.2243,
+      "step": 3925
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0091941317388682,
+      "learning_rate": 1.8528229766147737e-05,
+      "loss": 0.2156,
+      "step": 3926
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.428407186069211,
+      "learning_rate": 1.8527369603265027e-05,
+      "loss": 0.2173,
+      "step": 3927
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9884423582417243,
+      "learning_rate": 1.852650920907693e-05,
+      "loss": 0.2074,
+      "step": 3928
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0081667501207707,
+      "learning_rate": 1.8525648583606797e-05,
+      "loss": 0.2362,
+      "step": 3929
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9400868797070588,
+      "learning_rate": 1.8524787726877956e-05,
+      "loss": 0.2298,
+      "step": 3930
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0042055292298442,
+      "learning_rate": 1.8523926638913772e-05,
+      "loss": 0.207,
+      "step": 3931
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0274160199055828,
+      "learning_rate": 1.8523065319737597e-05,
+      "loss": 0.2311,
+      "step": 3932
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8839768613869868,
+      "learning_rate": 1.8522203769372794e-05,
+      "loss": 0.2022,
+      "step": 3933
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9604317794149866,
+      "learning_rate": 1.8521341987842726e-05,
+      "loss": 0.2192,
+      "step": 3934
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1657719535264452,
+      "learning_rate": 1.8520479975170783e-05,
+      "loss": 0.228,
+      "step": 3935
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0916660032667729,
+      "learning_rate": 1.8519617731380334e-05,
+      "loss": 0.2314,
+      "step": 3936
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9529536523373129,
+      "learning_rate": 1.8518755256494776e-05,
+      "loss": 0.2706,
+      "step": 3937
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9799853626728516,
+      "learning_rate": 1.85178925505375e-05,
+      "loss": 0.2381,
+      "step": 3938
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.987670708955151,
+      "learning_rate": 1.851702961353191e-05,
+      "loss": 0.222,
+      "step": 3939
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0361197104887185,
+      "learning_rate": 1.8516166445501405e-05,
+      "loss": 0.2459,
+      "step": 3940
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1496121131841177,
+      "learning_rate": 1.8515303046469407e-05,
+      "loss": 0.2305,
+      "step": 3941
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.175067670408769,
+      "learning_rate": 1.851443941645933e-05,
+      "loss": 0.2046,
+      "step": 3942
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0394576988286925,
+      "learning_rate": 1.851357555549461e-05,
+      "loss": 0.2249,
+      "step": 3943
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1378105288243487,
+      "learning_rate": 1.8512711463598666e-05,
+      "loss": 0.2161,
+      "step": 3944
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9691228729639899,
+      "learning_rate": 1.8511847140794944e-05,
+      "loss": 0.2451,
+      "step": 3945
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9107941557336705,
+      "learning_rate": 1.8510982587106888e-05,
+      "loss": 0.2163,
+      "step": 3946
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8438094197302133,
+      "learning_rate": 1.8510117802557948e-05,
+      "loss": 0.2172,
+      "step": 3947
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.122561032622386,
+      "learning_rate": 1.8509252787171585e-05,
+      "loss": 0.2091,
+      "step": 3948
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1670322241173445,
+      "learning_rate": 1.8508387540971258e-05,
+      "loss": 0.2265,
+      "step": 3949
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8737935359329443,
+      "learning_rate": 1.8507522063980436e-05,
+      "loss": 0.2014,
+      "step": 3950
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.957459298710243,
+      "learning_rate": 1.85066563562226e-05,
+      "loss": 0.2232,
+      "step": 3951
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9140556809479499,
+      "learning_rate": 1.850579041772123e-05,
+      "loss": 0.2213,
+      "step": 3952
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9341055775625612,
+      "learning_rate": 1.850492424849981e-05,
+      "loss": 0.2475,
+      "step": 3953
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8285233367924434,
+      "learning_rate": 1.8504057848581846e-05,
+      "loss": 0.2267,
+      "step": 3954
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2274016809994188,
+      "learning_rate": 1.850319121799083e-05,
+      "loss": 0.2167,
+      "step": 3955
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9569895683683748,
+      "learning_rate": 1.850232435675027e-05,
+      "loss": 0.2139,
+      "step": 3956
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9707813355784036,
+      "learning_rate": 1.8501457264883684e-05,
+      "loss": 0.2247,
+      "step": 3957
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6606571832523476,
+      "learning_rate": 1.850058994241459e-05,
+      "loss": 0.207,
+      "step": 3958
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9890119440101421,
+      "learning_rate": 1.8499722389366513e-05,
+      "loss": 0.2126,
+      "step": 3959
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8781210948864995,
+      "learning_rate": 1.8498854605762982e-05,
+      "loss": 0.2236,
+      "step": 3960
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2677885365794521,
+      "learning_rate": 1.8497986591627546e-05,
+      "loss": 0.2384,
+      "step": 3961
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8925478669144093,
+      "learning_rate": 1.849711834698374e-05,
+      "loss": 0.192,
+      "step": 3962
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.096343147470085,
+      "learning_rate": 1.849624987185512e-05,
+      "loss": 0.1939,
+      "step": 3963
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9426657053588514,
+      "learning_rate": 1.8495381166265243e-05,
+      "loss": 0.2222,
+      "step": 3964
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0983582096020919,
+      "learning_rate": 1.849451223023767e-05,
+      "loss": 0.2271,
+      "step": 3965
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1097098319224445,
+      "learning_rate": 1.8493643063795973e-05,
+      "loss": 0.21,
+      "step": 3966
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9815198326231998,
+      "learning_rate": 1.8492773666963734e-05,
+      "loss": 0.197,
+      "step": 3967
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1798340581543854,
+      "learning_rate": 1.8491904039764523e-05,
+      "loss": 0.2297,
+      "step": 3968
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0813074125313995,
+      "learning_rate": 1.849103418222194e-05,
+      "loss": 0.24,
+      "step": 3969
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0502962640861704,
+      "learning_rate": 1.849016409435957e-05,
+      "loss": 0.2127,
+      "step": 3970
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.005448488445372,
+      "learning_rate": 1.848929377620102e-05,
+      "loss": 0.2212,
+      "step": 3971
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3691830425652256,
+      "learning_rate": 1.84884232277699e-05,
+      "loss": 0.2317,
+      "step": 3972
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4197298065121104,
+      "learning_rate": 1.848755244908982e-05,
+      "loss": 0.2427,
+      "step": 3973
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2433237283809655,
+      "learning_rate": 1.84866814401844e-05,
+      "loss": 0.2357,
+      "step": 3974
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5182774149915292,
+      "learning_rate": 1.8485810201077266e-05,
+      "loss": 0.2133,
+      "step": 3975
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0417444382667709,
+      "learning_rate": 1.8484938731792052e-05,
+      "loss": 0.247,
+      "step": 3976
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1087194985346591,
+      "learning_rate": 1.8484067032352394e-05,
+      "loss": 0.2289,
+      "step": 3977
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.227029058062171,
+      "learning_rate": 1.848319510278194e-05,
+      "loss": 0.2142,
+      "step": 3978
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0410937941188434,
+      "learning_rate": 1.848232294310434e-05,
+      "loss": 0.207,
+      "step": 3979
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.013795250561338,
+      "learning_rate": 1.848145055334325e-05,
+      "loss": 0.2171,
+      "step": 3980
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9561721192043167,
+      "learning_rate": 1.8480577933522337e-05,
+      "loss": 0.2293,
+      "step": 3981
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9544153277598909,
+      "learning_rate": 1.8479705083665265e-05,
+      "loss": 0.1992,
+      "step": 3982
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4377696993777693,
+      "learning_rate": 1.8478832003795718e-05,
+      "loss": 0.2259,
+      "step": 3983
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0252596856842724,
+      "learning_rate": 1.847795869393737e-05,
+      "loss": 0.2228,
+      "step": 3984
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1646243348780456,
+      "learning_rate": 1.8477085154113913e-05,
+      "loss": 0.2211,
+      "step": 3985
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.435138324835618,
+      "learning_rate": 1.8476211384349048e-05,
+      "loss": 0.2141,
+      "step": 3986
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1034704971175715,
+      "learning_rate": 1.8475337384666464e-05,
+      "loss": 0.2461,
+      "step": 3987
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.040306523846338,
+      "learning_rate": 1.8474463155089882e-05,
+      "loss": 0.2177,
+      "step": 3988
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9668078218640307,
+      "learning_rate": 1.8473588695643002e-05,
+      "loss": 0.2249,
+      "step": 3989
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8420323004225635,
+      "learning_rate": 1.8472714006349554e-05,
+      "loss": 0.2322,
+      "step": 3990
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0936358241608921,
+      "learning_rate": 1.8471839087233255e-05,
+      "loss": 0.241,
+      "step": 3991
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.139357797659001,
+      "learning_rate": 1.8470963938317846e-05,
+      "loss": 0.2545,
+      "step": 3992
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9151370967374843,
+      "learning_rate": 1.847008855962706e-05,
+      "loss": 0.2387,
+      "step": 3993
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9739624100709218,
+      "learning_rate": 1.8469212951184646e-05,
+      "loss": 0.2377,
+      "step": 3994
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9524532724244088,
+      "learning_rate": 1.846833711301435e-05,
+      "loss": 0.2261,
+      "step": 3995
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8491076395346986,
+      "learning_rate": 1.8467461045139933e-05,
+      "loss": 0.2342,
+      "step": 3996
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.129923799376718,
+      "learning_rate": 1.8466584747585153e-05,
+      "loss": 0.2223,
+      "step": 3997
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7812650811136905,
+      "learning_rate": 1.846570822037379e-05,
+      "loss": 0.2453,
+      "step": 3998
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9166033090870708,
+      "learning_rate": 1.846483146352961e-05,
+      "loss": 0.2068,
+      "step": 3999
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3533118562066813,
+      "learning_rate": 1.84639544770764e-05,
+      "loss": 0.2234,
+      "step": 4000
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0150576918701262,
+      "learning_rate": 1.8463077261037946e-05,
+      "loss": 0.2315,
+      "step": 4001
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7586040200761496,
+      "learning_rate": 1.846219981543804e-05,
+      "loss": 0.1942,
+      "step": 4002
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.086690733510189,
+      "learning_rate": 1.8461322140300487e-05,
+      "loss": 0.2162,
+      "step": 4003
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0870007848264334,
+      "learning_rate": 1.8460444235649097e-05,
+      "loss": 0.2274,
+      "step": 4004
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0814676375496997,
+      "learning_rate": 1.8459566101507675e-05,
+      "loss": 0.2228,
+      "step": 4005
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2847826029249587,
+      "learning_rate": 1.845868773790005e-05,
+      "loss": 0.2466,
+      "step": 4006
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0485152699205598,
+      "learning_rate": 1.845780914485004e-05,
+      "loss": 0.2193,
+      "step": 4007
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3231739436000518,
+      "learning_rate": 1.8456930322381476e-05,
+      "loss": 0.1984,
+      "step": 4008
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1393536976170808,
+      "learning_rate": 1.8456051270518204e-05,
+      "loss": 0.1964,
+      "step": 4009
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9858193805660505,
+      "learning_rate": 1.845517198928406e-05,
+      "loss": 0.2093,
+      "step": 4010
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.102746359893761,
+      "learning_rate": 1.8454292478702898e-05,
+      "loss": 0.2292,
+      "step": 4011
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2060706071043537,
+      "learning_rate": 1.8453412738798577e-05,
+      "loss": 0.2282,
+      "step": 4012
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9578800418566528,
+      "learning_rate": 1.8452532769594956e-05,
+      "loss": 0.1901,
+      "step": 4013
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0015965893238714,
+      "learning_rate": 1.845165257111591e-05,
+      "loss": 0.2223,
+      "step": 4014
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.23810803358979,
+      "learning_rate": 1.845077214338531e-05,
+      "loss": 0.1973,
+      "step": 4015
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8611856599209402,
+      "learning_rate": 1.8449891486427037e-05,
+      "loss": 0.2108,
+      "step": 4016
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0767077470178619,
+      "learning_rate": 1.844901060026498e-05,
+      "loss": 0.2134,
+      "step": 4017
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9266494162460877,
+      "learning_rate": 1.844812948492303e-05,
+      "loss": 0.1851,
+      "step": 4018
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0603717922287728,
+      "learning_rate": 1.8447248140425093e-05,
+      "loss": 0.2049,
+      "step": 4019
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3324952507743006,
+      "learning_rate": 1.8446366566795075e-05,
+      "loss": 0.2196,
+      "step": 4020
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9255293165654925,
+      "learning_rate": 1.8445484764056886e-05,
+      "loss": 0.2316,
+      "step": 4021
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.216017989027153,
+      "learning_rate": 1.8444602732234448e-05,
+      "loss": 0.2204,
+      "step": 4022
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.389842141852318,
+      "learning_rate": 1.844372047135168e-05,
+      "loss": 0.2344,
+      "step": 4023
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4110691252305398,
+      "learning_rate": 1.844283798143252e-05,
+      "loss": 0.1954,
+      "step": 4024
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.005032913346493,
+      "learning_rate": 1.84419552625009e-05,
+      "loss": 0.2107,
+      "step": 4025
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9342081803585582,
+      "learning_rate": 1.8441072314580768e-05,
+      "loss": 0.2152,
+      "step": 4026
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.53410098202529,
+      "learning_rate": 1.8440189137696075e-05,
+      "loss": 0.2326,
+      "step": 4027
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.281389367713628,
+      "learning_rate": 1.843930573187077e-05,
+      "loss": 0.2088,
+      "step": 4028
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9824915018152038,
+      "learning_rate": 1.8438422097128824e-05,
+      "loss": 0.221,
+      "step": 4029
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9399054436239845,
+      "learning_rate": 1.8437538233494205e-05,
+      "loss": 0.2168,
+      "step": 4030
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0611693014693109,
+      "learning_rate": 1.8436654140990877e-05,
+      "loss": 0.2201,
+      "step": 4031
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6926024066852987,
+      "learning_rate": 1.8435769819642835e-05,
+      "loss": 0.2168,
+      "step": 4032
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.425124544776791,
+      "learning_rate": 1.843488526947406e-05,
+      "loss": 0.2332,
+      "step": 4033
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2986176817240906,
+      "learning_rate": 1.843400049050854e-05,
+      "loss": 0.2054,
+      "step": 4034
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4592279022760886,
+      "learning_rate": 1.8433115482770286e-05,
+      "loss": 0.1938,
+      "step": 4035
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1344940384527544,
+      "learning_rate": 1.8432230246283298e-05,
+      "loss": 0.2134,
+      "step": 4036
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4920957263868553,
+      "learning_rate": 1.8431344781071586e-05,
+      "loss": 0.2299,
+      "step": 4037
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6332568235688114,
+      "learning_rate": 1.8430459087159172e-05,
+      "loss": 0.2037,
+      "step": 4038
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0305365277043756,
+      "learning_rate": 1.842957316457008e-05,
+      "loss": 0.2396,
+      "step": 4039
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2429204314009332,
+      "learning_rate": 1.8428687013328338e-05,
+      "loss": 0.2146,
+      "step": 4040
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.46064835422682,
+      "learning_rate": 1.8427800633457984e-05,
+      "loss": 0.213,
+      "step": 4041
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.315399648856405,
+      "learning_rate": 1.8426914024983064e-05,
+      "loss": 0.2228,
+      "step": 4042
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.148265946387433,
+      "learning_rate": 1.8426027187927622e-05,
+      "loss": 0.2231,
+      "step": 4043
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2784427788505968,
+      "learning_rate": 1.842514012231572e-05,
+      "loss": 0.214,
+      "step": 4044
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3232033473579259,
+      "learning_rate": 1.8424252828171415e-05,
+      "loss": 0.2424,
+      "step": 4045
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8993518609330069,
+      "learning_rate": 1.842336530551878e-05,
+      "loss": 0.2106,
+      "step": 4046
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9717114990888179,
+      "learning_rate": 1.8422477554381877e-05,
+      "loss": 0.2137,
+      "step": 4047
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1005113781951843,
+      "learning_rate": 1.84215895747848e-05,
+      "loss": 0.2485,
+      "step": 4048
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.78659538266863,
+      "learning_rate": 1.842070136675163e-05,
+      "loss": 0.2314,
+      "step": 4049
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0992894999940652,
+      "learning_rate": 1.8419812930306456e-05,
+      "loss": 0.2355,
+      "step": 4050
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9472424028153374,
+      "learning_rate": 1.8418924265473387e-05,
+      "loss": 0.2097,
+      "step": 4051
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2428118737564922,
+      "learning_rate": 1.8418035372276518e-05,
+      "loss": 0.2048,
+      "step": 4052
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.115620427326506,
+      "learning_rate": 1.8417146250739963e-05,
+      "loss": 0.245,
+      "step": 4053
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1725585511883763,
+      "learning_rate": 1.8416256900887846e-05,
+      "loss": 0.2305,
+      "step": 4054
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.164010195877797,
+      "learning_rate": 1.841536732274428e-05,
+      "loss": 0.2317,
+      "step": 4055
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9901967708490926,
+      "learning_rate": 1.8414477516333404e-05,
+      "loss": 0.2073,
+      "step": 4056
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.203022719344018,
+      "learning_rate": 1.8413587481679343e-05,
+      "loss": 0.221,
+      "step": 4057
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1685669183493879,
+      "learning_rate": 1.841269721880625e-05,
+      "loss": 0.2065,
+      "step": 4058
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1292187794920885,
+      "learning_rate": 1.8411806727738276e-05,
+      "loss": 0.2128,
+      "step": 4059
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0460899988252699,
+      "learning_rate": 1.8410916008499563e-05,
+      "loss": 0.2065,
+      "step": 4060
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9625875477372454,
+      "learning_rate": 1.8410025061114278e-05,
+      "loss": 0.1914,
+      "step": 4061
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.013377246352948,
+      "learning_rate": 1.840913388560659e-05,
+      "loss": 0.2486,
+      "step": 4062
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.054808112700336,
+      "learning_rate": 1.840824248200067e-05,
+      "loss": 0.2036,
+      "step": 4063
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.915720731760347,
+      "learning_rate": 1.8407350850320698e-05,
+      "loss": 0.215,
+      "step": 4064
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9131373890768805,
+      "learning_rate": 1.840645899059086e-05,
+      "loss": 0.2176,
+      "step": 4065
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4499765627745933,
+      "learning_rate": 1.8405566902835345e-05,
+      "loss": 0.2072,
+      "step": 4066
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0072741469965514,
+      "learning_rate": 1.8404674587078355e-05,
+      "loss": 0.216,
+      "step": 4067
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.116054906081107,
+      "learning_rate": 1.840378204334409e-05,
+      "loss": 0.2289,
+      "step": 4068
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.393753900145944,
+      "learning_rate": 1.8402889271656764e-05,
+      "loss": 0.2156,
+      "step": 4069
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7990680573099183,
+      "learning_rate": 1.8401996272040593e-05,
+      "loss": 0.2357,
+      "step": 4070
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9247437119879226,
+      "learning_rate": 1.8401103044519798e-05,
+      "loss": 0.2104,
+      "step": 4071
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0478663800704129,
+      "learning_rate": 1.8400209589118606e-05,
+      "loss": 0.2605,
+      "step": 4072
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.606507652653388,
+      "learning_rate": 1.839931590586126e-05,
+      "loss": 0.2146,
+      "step": 4073
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8734486664573401,
+      "learning_rate": 1.8398421994771987e-05,
+      "loss": 0.2386,
+      "step": 4074
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8981227604859134,
+      "learning_rate": 1.839752785587505e-05,
+      "loss": 0.248,
+      "step": 4075
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8195126193179083,
+      "learning_rate": 1.8396633489194693e-05,
+      "loss": 0.2247,
+      "step": 4076
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9562878691843798,
+      "learning_rate": 1.839573889475518e-05,
+      "loss": 0.2231,
+      "step": 4077
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0949687933363768,
+      "learning_rate": 1.8394844072580772e-05,
+      "loss": 0.2342,
+      "step": 4078
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8892593014012868,
+      "learning_rate": 1.8393949022695747e-05,
+      "loss": 0.1962,
+      "step": 4079
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1401884052275038,
+      "learning_rate": 1.839305374512438e-05,
+      "loss": 0.2407,
+      "step": 4080
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9840977904794602,
+      "learning_rate": 1.8392158239890957e-05,
+      "loss": 0.2285,
+      "step": 4081
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0557285677899844,
+      "learning_rate": 1.839126250701977e-05,
+      "loss": 0.1971,
+      "step": 4082
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0082003899112544,
+      "learning_rate": 1.8390366546535107e-05,
+      "loss": 0.2373,
+      "step": 4083
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.482950351936631,
+      "learning_rate": 1.8389470358461282e-05,
+      "loss": 0.217,
+      "step": 4084
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6604273182636915,
+      "learning_rate": 1.8388573942822597e-05,
+      "loss": 0.1997,
+      "step": 4085
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.061341683760589,
+      "learning_rate": 1.8387677299643374e-05,
+      "loss": 0.2058,
+      "step": 4086
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8804008446399976,
+      "learning_rate": 1.838678042894793e-05,
+      "loss": 0.2171,
+      "step": 4087
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.159037826236392,
+      "learning_rate": 1.838588333076059e-05,
+      "loss": 0.2258,
+      "step": 4088
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3007209396237813,
+      "learning_rate": 1.8384986005105694e-05,
+      "loss": 0.197,
+      "step": 4089
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1739064368395806,
+      "learning_rate": 1.838408845200758e-05,
+      "loss": 0.2115,
+      "step": 4090
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3309068976708802,
+      "learning_rate": 1.8383190671490586e-05,
+      "loss": 0.2085,
+      "step": 4091
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0072035766635603,
+      "learning_rate": 1.838229266357908e-05,
+      "loss": 0.2207,
+      "step": 4092
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0280329947360802,
+      "learning_rate": 1.8381394428297406e-05,
+      "loss": 0.2161,
+      "step": 4093
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9584474084596254,
+      "learning_rate": 1.8380495965669938e-05,
+      "loss": 0.2323,
+      "step": 4094
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.192651766726789,
+      "learning_rate": 1.8379597275721043e-05,
+      "loss": 0.2065,
+      "step": 4095
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.279939931690827,
+      "learning_rate": 1.83786983584751e-05,
+      "loss": 0.2361,
+      "step": 4096
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6723687874802045,
+      "learning_rate": 1.8377799213956487e-05,
+      "loss": 0.2095,
+      "step": 4097
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0774746273066567,
+      "learning_rate": 1.83768998421896e-05,
+      "loss": 0.2343,
+      "step": 4098
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8065607882690657,
+      "learning_rate": 1.837600024319883e-05,
+      "loss": 0.2407,
+      "step": 4099
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9560382672546652,
+      "learning_rate": 1.8375100417008584e-05,
+      "loss": 0.2235,
+      "step": 4100
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9020948472107335,
+      "learning_rate": 1.8374200363643263e-05,
+      "loss": 0.2353,
+      "step": 4101
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9690265583262132,
+      "learning_rate": 1.8373300083127284e-05,
+      "loss": 0.2275,
+      "step": 4102
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3507623603560543,
+      "learning_rate": 1.8372399575485068e-05,
+      "loss": 0.2636,
+      "step": 4103
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4806389422531303,
+      "learning_rate": 1.837149884074104e-05,
+      "loss": 0.1946,
+      "step": 4104
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.002612736176136,
+      "learning_rate": 1.8370597878919633e-05,
+      "loss": 0.2273,
+      "step": 4105
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0672473118919987,
+      "learning_rate": 1.8369696690045288e-05,
+      "loss": 0.2122,
+      "step": 4106
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.029358419541676,
+      "learning_rate": 1.8368795274142446e-05,
+      "loss": 0.2098,
+      "step": 4107
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8667773516314685,
+      "learning_rate": 1.8367893631235558e-05,
+      "loss": 0.2162,
+      "step": 4108
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8290441652256468,
+      "learning_rate": 1.8366991761349084e-05,
+      "loss": 0.2286,
+      "step": 4109
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6958139792602507,
+      "learning_rate": 1.8366089664507488e-05,
+      "loss": 0.219,
+      "step": 4110
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.04419481834506,
+      "learning_rate": 1.836518734073523e-05,
+      "loss": 0.2209,
+      "step": 4111
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9153247072007938,
+      "learning_rate": 1.8364284790056804e-05,
+      "loss": 0.246,
+      "step": 4112
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0906657376798792,
+      "learning_rate": 1.8363382012496672e-05,
+      "loss": 0.1906,
+      "step": 4113
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0728346171084582,
+      "learning_rate": 1.8362479008079334e-05,
+      "loss": 0.2121,
+      "step": 4114
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0315598757042255,
+      "learning_rate": 1.836157577682928e-05,
+      "loss": 0.2437,
+      "step": 4115
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9828449756288894,
+      "learning_rate": 1.8360672318771013e-05,
+      "loss": 0.2268,
+      "step": 4116
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1873815853782363,
+      "learning_rate": 1.8359768633929035e-05,
+      "loss": 0.217,
+      "step": 4117
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8210484261375784,
+      "learning_rate": 1.835886472232786e-05,
+      "loss": 0.1881,
+      "step": 4118
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2900341454161641,
+      "learning_rate": 1.835796058399201e-05,
+      "loss": 0.2366,
+      "step": 4119
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0392124418066642,
+      "learning_rate": 1.8357056218946003e-05,
+      "loss": 0.2439,
+      "step": 4120
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8318283207469718,
+      "learning_rate": 1.8356151627214377e-05,
+      "loss": 0.206,
+      "step": 4121
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.827521381720995,
+      "learning_rate": 1.8355246808821664e-05,
+      "loss": 0.2123,
+      "step": 4122
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0971340042134967,
+      "learning_rate": 1.835434176379241e-05,
+      "loss": 0.208,
+      "step": 4123
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0739765981940956,
+      "learning_rate": 1.8353436492151165e-05,
+      "loss": 0.2095,
+      "step": 4124
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6935205221026808,
+      "learning_rate": 1.8352530993922483e-05,
+      "loss": 0.1864,
+      "step": 4125
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.247650082249108,
+      "learning_rate": 1.8351625269130927e-05,
+      "loss": 0.2093,
+      "step": 4126
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9056832888194429,
+      "learning_rate": 1.8350719317801062e-05,
+      "loss": 0.214,
+      "step": 4127
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3095563898978886,
+      "learning_rate": 1.8349813139957464e-05,
+      "loss": 0.2077,
+      "step": 4128
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9821047572278827,
+      "learning_rate": 1.8348906735624716e-05,
+      "loss": 0.2112,
+      "step": 4129
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0638206523680762,
+      "learning_rate": 1.8348000104827396e-05,
+      "loss": 0.2321,
+      "step": 4130
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7993847668276511,
+      "learning_rate": 1.8347093247590106e-05,
+      "loss": 0.199,
+      "step": 4131
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9723068373547615,
+      "learning_rate": 1.834618616393744e-05,
+      "loss": 0.1977,
+      "step": 4132
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2098802854352386,
+      "learning_rate": 1.8345278853894e-05,
+      "loss": 0.2031,
+      "step": 4133
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9292414406170786,
+      "learning_rate": 1.8344371317484402e-05,
+      "loss": 0.2152,
+      "step": 4134
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0821646783609271,
+      "learning_rate": 1.834346355473326e-05,
+      "loss": 0.2294,
+      "step": 4135
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9828080305508093,
+      "learning_rate": 1.83425555656652e-05,
+      "loss": 0.2166,
+      "step": 4136
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0148458077937668,
+      "learning_rate": 1.8341647350304844e-05,
+      "loss": 0.2078,
+      "step": 4137
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8506156892427307,
+      "learning_rate": 1.8340738908676837e-05,
+      "loss": 0.2113,
+      "step": 4138
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9936066991084263,
+      "learning_rate": 1.8339830240805814e-05,
+      "loss": 0.2344,
+      "step": 4139
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1197923087940107,
+      "learning_rate": 1.8338921346716426e-05,
+      "loss": 0.2049,
+      "step": 4140
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8916498553865597,
+      "learning_rate": 1.8338012226433322e-05,
+      "loss": 0.2569,
+      "step": 4141
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9060713658218383,
+      "learning_rate": 1.833710287998117e-05,
+      "loss": 0.2112,
+      "step": 4142
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4070229822572158,
+      "learning_rate": 1.833619330738463e-05,
+      "loss": 0.2375,
+      "step": 4143
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9140861151714124,
+      "learning_rate": 1.8335283508668375e-05,
+      "loss": 0.2301,
+      "step": 4144
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1048618491381426,
+      "learning_rate": 1.8334373483857083e-05,
+      "loss": 0.2266,
+      "step": 4145
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9368278246902417,
+      "learning_rate": 1.8333463232975442e-05,
+      "loss": 0.2336,
+      "step": 4146
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9103085291862169,
+      "learning_rate": 1.833255275604814e-05,
+      "loss": 0.2229,
+      "step": 4147
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8832449772404098,
+      "learning_rate": 1.8331642053099873e-05,
+      "loss": 0.2189,
+      "step": 4148
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8529873970875689,
+      "learning_rate": 1.8330731124155347e-05,
+      "loss": 0.231,
+      "step": 4149
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9615479076879223,
+      "learning_rate": 1.8329819969239263e-05,
+      "loss": 0.2316,
+      "step": 4150
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8526259123087873,
+      "learning_rate": 1.8328908588376345e-05,
+      "loss": 0.2168,
+      "step": 4151
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0171620974352265,
+      "learning_rate": 1.832799698159131e-05,
+      "loss": 0.2518,
+      "step": 4152
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.078204216193984,
+      "learning_rate": 1.832708514890889e-05,
+      "loss": 0.2242,
+      "step": 4153
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9815351765693545,
+      "learning_rate": 1.8326173090353815e-05,
+      "loss": 0.2537,
+      "step": 4154
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7385114890983147,
+      "learning_rate": 1.832526080595082e-05,
+      "loss": 0.2249,
+      "step": 4155
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0942431223529165,
+      "learning_rate": 1.8324348295724658e-05,
+      "loss": 0.2312,
+      "step": 4156
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.905791708407855,
+      "learning_rate": 1.8323435559700077e-05,
+      "loss": 0.2104,
+      "step": 4157
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0635355993872433,
+      "learning_rate": 1.832252259790184e-05,
+      "loss": 0.2407,
+      "step": 4158
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9400388503599094,
+      "learning_rate": 1.8321609410354705e-05,
+      "loss": 0.2014,
+      "step": 4159
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1323684487566767,
+      "learning_rate": 1.8320695997083443e-05,
+      "loss": 0.2531,
+      "step": 4160
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9332268296880125,
+      "learning_rate": 1.8319782358112836e-05,
+      "loss": 0.206,
+      "step": 4161
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9945749482242312,
+      "learning_rate": 1.8318868493467657e-05,
+      "loss": 0.2189,
+      "step": 4162
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1305659185344676,
+      "learning_rate": 1.8317954403172708e-05,
+      "loss": 0.2187,
+      "step": 4163
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1061371469329477,
+      "learning_rate": 1.831704008725277e-05,
+      "loss": 0.2221,
+      "step": 4164
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8551129861299791,
+      "learning_rate": 1.8316125545732653e-05,
+      "loss": 0.1808,
+      "step": 4165
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9663446712855458,
+      "learning_rate": 1.831521077863716e-05,
+      "loss": 0.2057,
+      "step": 4166
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9287018223950793,
+      "learning_rate": 1.8314295785991103e-05,
+      "loss": 0.2417,
+      "step": 4167
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.200372329075485,
+      "learning_rate": 1.8313380567819306e-05,
+      "loss": 0.2132,
+      "step": 4168
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.015871584887287,
+      "learning_rate": 1.831246512414659e-05,
+      "loss": 0.2127,
+      "step": 4169
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0873608708788338,
+      "learning_rate": 1.8311549454997788e-05,
+      "loss": 0.2173,
+      "step": 4170
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8060995006211655,
+      "learning_rate": 1.8310633560397737e-05,
+      "loss": 0.2203,
+      "step": 4171
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9484747795626762,
+      "learning_rate": 1.8309717440371282e-05,
+      "loss": 0.2061,
+      "step": 4172
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.011706199342892,
+      "learning_rate": 1.8308801094943275e-05,
+      "loss": 0.2099,
+      "step": 4173
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0158471698512874,
+      "learning_rate": 1.8307884524138564e-05,
+      "loss": 0.2141,
+      "step": 4174
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0120488489692896,
+      "learning_rate": 1.830696772798202e-05,
+      "loss": 0.2109,
+      "step": 4175
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1944801073571805,
+      "learning_rate": 1.8306050706498503e-05,
+      "loss": 0.2105,
+      "step": 4176
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1529388807643852,
+      "learning_rate": 1.8305133459712897e-05,
+      "loss": 0.2308,
+      "step": 4177
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.95321287948936,
+      "learning_rate": 1.8304215987650074e-05,
+      "loss": 0.2175,
+      "step": 4178
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0593397775678555,
+      "learning_rate": 1.830329829033492e-05,
+      "loss": 0.2177,
+      "step": 4179
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0005744492424244,
+      "learning_rate": 1.8302380367792336e-05,
+      "loss": 0.2282,
+      "step": 4180
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9528885425476983,
+      "learning_rate": 1.8301462220047213e-05,
+      "loss": 0.2404,
+      "step": 4181
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4416274627260237,
+      "learning_rate": 1.830054384712446e-05,
+      "loss": 0.2366,
+      "step": 4182
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8464081363401605,
+      "learning_rate": 1.8299625249048986e-05,
+      "loss": 0.2069,
+      "step": 4183
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9436797273759833,
+      "learning_rate": 1.8298706425845707e-05,
+      "loss": 0.2426,
+      "step": 4184
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7711041358692556,
+      "learning_rate": 1.829778737753955e-05,
+      "loss": 0.2209,
+      "step": 4185
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9606908868034619,
+      "learning_rate": 1.8296868104155437e-05,
+      "loss": 0.1989,
+      "step": 4186
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8278759406694916,
+      "learning_rate": 1.8295948605718316e-05,
+      "loss": 0.2192,
+      "step": 4187
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.921098878397764,
+      "learning_rate": 1.8295028882253113e-05,
+      "loss": 0.2132,
+      "step": 4188
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8795989807205208,
+      "learning_rate": 1.8294108933784788e-05,
+      "loss": 0.2211,
+      "step": 4189
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2470788534129764,
+      "learning_rate": 1.8293188760338285e-05,
+      "loss": 0.188,
+      "step": 4190
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9763265103504567,
+      "learning_rate": 1.829226836193857e-05,
+      "loss": 0.2235,
+      "step": 4191
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1680172459987113,
+      "learning_rate": 1.829134773861061e-05,
+      "loss": 0.2086,
+      "step": 4192
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9284572674758612,
+      "learning_rate": 1.8290426890379372e-05,
+      "loss": 0.2337,
+      "step": 4193
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8362341065194638,
+      "learning_rate": 1.8289505817269838e-05,
+      "loss": 0.199,
+      "step": 4194
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.884264449424886,
+      "learning_rate": 1.8288584519306985e-05,
+      "loss": 0.2095,
+      "step": 4195
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1174268482330303,
+      "learning_rate": 1.8287662996515815e-05,
+      "loss": 0.239,
+      "step": 4196
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9739645939599759,
+      "learning_rate": 1.8286741248921317e-05,
+      "loss": 0.2161,
+      "step": 4197
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.96075065186622,
+      "learning_rate": 1.8285819276548494e-05,
+      "loss": 0.2154,
+      "step": 4198
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7959212458472489,
+      "learning_rate": 1.8284897079422356e-05,
+      "loss": 0.2084,
+      "step": 4199
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.171455556259447,
+      "learning_rate": 1.8283974657567915e-05,
+      "loss": 0.2254,
+      "step": 4200
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.943038970137586,
+      "learning_rate": 1.8283052011010195e-05,
+      "loss": 0.1983,
+      "step": 4201
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8424994263691186,
+      "learning_rate": 1.828212913977422e-05,
+      "loss": 0.2031,
+      "step": 4202
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9446157746444535,
+      "learning_rate": 1.8281206043885024e-05,
+      "loss": 0.2401,
+      "step": 4203
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1912114330547572,
+      "learning_rate": 1.8280282723367647e-05,
+      "loss": 0.2461,
+      "step": 4204
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9212399950951802,
+      "learning_rate": 1.8279359178247134e-05,
+      "loss": 0.2282,
+      "step": 4205
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9175859980076985,
+      "learning_rate": 1.8278435408548538e-05,
+      "loss": 0.2271,
+      "step": 4206
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8472401194222476,
+      "learning_rate": 1.827751141429691e-05,
+      "loss": 0.2393,
+      "step": 4207
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.866124853602909,
+      "learning_rate": 1.827658719551732e-05,
+      "loss": 0.2121,
+      "step": 4208
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8963703373904084,
+      "learning_rate": 1.8275662752234834e-05,
+      "loss": 0.2013,
+      "step": 4209
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0791283238506562,
+      "learning_rate": 1.8274738084474525e-05,
+      "loss": 0.2207,
+      "step": 4210
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8172063107163877,
+      "learning_rate": 1.8273813192261483e-05,
+      "loss": 0.2197,
+      "step": 4211
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1167127205533103,
+      "learning_rate": 1.827288807562079e-05,
+      "loss": 0.2292,
+      "step": 4212
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.765006058341567,
+      "learning_rate": 1.8271962734577536e-05,
+      "loss": 0.2197,
+      "step": 4213
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7670932099972653,
+      "learning_rate": 1.827103716915683e-05,
+      "loss": 0.2094,
+      "step": 4214
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9312624220759745,
+      "learning_rate": 1.8270111379383773e-05,
+      "loss": 0.2108,
+      "step": 4215
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9058587404838406,
+      "learning_rate": 1.826918536528348e-05,
+      "loss": 0.2213,
+      "step": 4216
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8238733963219388,
+      "learning_rate": 1.8268259126881064e-05,
+      "loss": 0.2249,
+      "step": 4217
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.960822749130999,
+      "learning_rate": 1.8267332664201653e-05,
+      "loss": 0.2311,
+      "step": 4218
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.242662687452094,
+      "learning_rate": 1.8266405977270377e-05,
+      "loss": 0.2289,
+      "step": 4219
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9535625295217197,
+      "learning_rate": 1.826547906611237e-05,
+      "loss": 0.2116,
+      "step": 4220
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8201907851659748,
+      "learning_rate": 1.826455193075278e-05,
+      "loss": 0.1827,
+      "step": 4221
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9915348756782004,
+      "learning_rate": 1.826362457121675e-05,
+      "loss": 0.2064,
+      "step": 4222
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.783989783933111,
+      "learning_rate": 1.8262696987529434e-05,
+      "loss": 0.2036,
+      "step": 4223
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8067961548800195,
+      "learning_rate": 1.8261769179716e-05,
+      "loss": 0.2137,
+      "step": 4224
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7977405884808337,
+      "learning_rate": 1.826084114780161e-05,
+      "loss": 0.2121,
+      "step": 4225
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8556151782037259,
+      "learning_rate": 1.8259912891811433e-05,
+      "loss": 0.2025,
+      "step": 4226
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9531497815790603,
+      "learning_rate": 1.8258984411770656e-05,
+      "loss": 0.2398,
+      "step": 4227
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0105056727131227,
+      "learning_rate": 1.825805570770446e-05,
+      "loss": 0.2234,
+      "step": 4228
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7200577937137307,
+      "learning_rate": 1.825712677963804e-05,
+      "loss": 0.2025,
+      "step": 4229
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8241487265119518,
+      "learning_rate": 1.8256197627596582e-05,
+      "loss": 0.2261,
+      "step": 4230
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9253391184043739,
+      "learning_rate": 1.8255268251605303e-05,
+      "loss": 0.2122,
+      "step": 4231
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.201686032556116,
+      "learning_rate": 1.8254338651689402e-05,
+      "loss": 0.2139,
+      "step": 4232
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8768725928195028,
+      "learning_rate": 1.82534088278741e-05,
+      "loss": 0.205,
+      "step": 4233
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.059742002088647,
+      "learning_rate": 1.825247878018462e-05,
+      "loss": 0.1921,
+      "step": 4234
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.916012985496521,
+      "learning_rate": 1.8251548508646186e-05,
+      "loss": 0.2451,
+      "step": 4235
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9015541425224542,
+      "learning_rate": 1.825061801328403e-05,
+      "loss": 0.2123,
+      "step": 4236
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0308071881686403,
+      "learning_rate": 1.8249687294123396e-05,
+      "loss": 0.2194,
+      "step": 4237
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9070747617664877,
+      "learning_rate": 1.8248756351189533e-05,
+      "loss": 0.1742,
+      "step": 4238
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8358651708734051,
+      "learning_rate": 1.8247825184507683e-05,
+      "loss": 0.2048,
+      "step": 4239
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8122009336805338,
+      "learning_rate": 1.8246893794103113e-05,
+      "loss": 0.2149,
+      "step": 4240
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5371587945589475,
+      "learning_rate": 1.8245962180001075e-05,
+      "loss": 0.2427,
+      "step": 4241
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0558139292489592,
+      "learning_rate": 1.824503034222685e-05,
+      "loss": 0.219,
+      "step": 4242
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8009101214707464,
+      "learning_rate": 1.8244098280805715e-05,
+      "loss": 0.2196,
+      "step": 4243
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7883653453666257,
+      "learning_rate": 1.8243165995762947e-05,
+      "loss": 0.2169,
+      "step": 4244
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4303484593440836,
+      "learning_rate": 1.8242233487123837e-05,
+      "loss": 0.2248,
+      "step": 4245
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9900955967331986,
+      "learning_rate": 1.8241300754913677e-05,
+      "loss": 0.2525,
+      "step": 4246
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0160311186185382,
+      "learning_rate": 1.824036779915777e-05,
+      "loss": 0.1988,
+      "step": 4247
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7568400522467681,
+      "learning_rate": 1.8239434619881418e-05,
+      "loss": 0.201,
+      "step": 4248
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0200762232544731,
+      "learning_rate": 1.8238501217109938e-05,
+      "loss": 0.2425,
+      "step": 4249
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7002983112880843,
+      "learning_rate": 1.8237567590868645e-05,
+      "loss": 0.196,
+      "step": 4250
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9417901704321855,
+      "learning_rate": 1.823663374118287e-05,
+      "loss": 0.1868,
+      "step": 4251
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0660457871013291,
+      "learning_rate": 1.8235699668077938e-05,
+      "loss": 0.2183,
+      "step": 4252
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3133422232494105,
+      "learning_rate": 1.8234765371579186e-05,
+      "loss": 0.2164,
+      "step": 4253
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.766017062175045,
+      "learning_rate": 1.8233830851711957e-05,
+      "loss": 0.2402,
+      "step": 4254
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.474811583060738,
+      "learning_rate": 1.8232896108501606e-05,
+      "loss": 0.2328,
+      "step": 4255
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7342884869301584,
+      "learning_rate": 1.823196114197348e-05,
+      "loss": 0.2072,
+      "step": 4256
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1920093532059117,
+      "learning_rate": 1.823102595215294e-05,
+      "loss": 0.238,
+      "step": 4257
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9002978568597515,
+      "learning_rate": 1.823009053906536e-05,
+      "loss": 0.2045,
+      "step": 4258
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0125908254260827,
+      "learning_rate": 1.8229154902736113e-05,
+      "loss": 0.2097,
+      "step": 4259
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9305152927324657,
+      "learning_rate": 1.822821904319057e-05,
+      "loss": 0.237,
+      "step": 4260
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9620551761878576,
+      "learning_rate": 1.822728296045412e-05,
+      "loss": 0.2567,
+      "step": 4261
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9060764864986361,
+      "learning_rate": 1.8226346654552155e-05,
+      "loss": 0.2139,
+      "step": 4262
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8190218386720736,
+      "learning_rate": 1.8225410125510073e-05,
+      "loss": 0.2387,
+      "step": 4263
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7968878989634448,
+      "learning_rate": 1.8224473373353276e-05,
+      "loss": 0.189,
+      "step": 4264
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9643006351217062,
+      "learning_rate": 1.8223536398107177e-05,
+      "loss": 0.2083,
+      "step": 4265
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0532074306008665,
+      "learning_rate": 1.8222599199797188e-05,
+      "loss": 0.2094,
+      "step": 4266
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9184139480026474,
+      "learning_rate": 1.822166177844873e-05,
+      "loss": 0.1897,
+      "step": 4267
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0147986932673607,
+      "learning_rate": 1.8220724134087232e-05,
+      "loss": 0.2154,
+      "step": 4268
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9907113242466752,
+      "learning_rate": 1.8219786266738125e-05,
+      "loss": 0.1967,
+      "step": 4269
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8044965495570748,
+      "learning_rate": 1.8218848176426857e-05,
+      "loss": 0.2316,
+      "step": 4270
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9013298789933274,
+      "learning_rate": 1.8217909863178868e-05,
+      "loss": 0.2534,
+      "step": 4271
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8488274163325105,
+      "learning_rate": 1.8216971327019603e-05,
+      "loss": 0.2254,
+      "step": 4272
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8406913678522667,
+      "learning_rate": 1.821603256797453e-05,
+      "loss": 0.1963,
+      "step": 4273
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6575549638758935,
+      "learning_rate": 1.821509358606911e-05,
+      "loss": 0.2213,
+      "step": 4274
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8083485169095919,
+      "learning_rate": 1.8214154381328815e-05,
+      "loss": 0.2073,
+      "step": 4275
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9830926770574077,
+      "learning_rate": 1.8213214953779114e-05,
+      "loss": 0.2458,
+      "step": 4276
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7787523267533243,
+      "learning_rate": 1.8212275303445498e-05,
+      "loss": 0.2295,
+      "step": 4277
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8838887511848333,
+      "learning_rate": 1.8211335430353444e-05,
+      "loss": 0.2107,
+      "step": 4278
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0056438528043299,
+      "learning_rate": 1.821039533452846e-05,
+      "loss": 0.2183,
+      "step": 4279
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6772895668147175,
+      "learning_rate": 1.8209455015996034e-05,
+      "loss": 0.1904,
+      "step": 4280
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8927283512952466,
+      "learning_rate": 1.820851447478168e-05,
+      "loss": 0.2229,
+      "step": 4281
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9770840012710763,
+      "learning_rate": 1.8207573710910905e-05,
+      "loss": 0.2418,
+      "step": 4282
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8916316262230806,
+      "learning_rate": 1.820663272440923e-05,
+      "loss": 0.2156,
+      "step": 4283
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2917448104325193,
+      "learning_rate": 1.8205691515302183e-05,
+      "loss": 0.2447,
+      "step": 4284
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9632645005230199,
+      "learning_rate": 1.8204750083615283e-05,
+      "loss": 0.2363,
+      "step": 4285
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8966795248674965,
+      "learning_rate": 1.8203808429374078e-05,
+      "loss": 0.2229,
+      "step": 4286
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1035150706057364,
+      "learning_rate": 1.8202866552604104e-05,
+      "loss": 0.2508,
+      "step": 4287
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4050937171557807,
+      "learning_rate": 1.8201924453330914e-05,
+      "loss": 0.186,
+      "step": 4288
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0667736728743291,
+      "learning_rate": 1.8200982131580058e-05,
+      "loss": 0.2044,
+      "step": 4289
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8705957052526101,
+      "learning_rate": 1.82000395873771e-05,
+      "loss": 0.2234,
+      "step": 4290
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7867090078526178,
+      "learning_rate": 1.8199096820747603e-05,
+      "loss": 0.2106,
+      "step": 4291
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.904053305616991,
+      "learning_rate": 1.819815383171714e-05,
+      "loss": 0.2097,
+      "step": 4292
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9641010472452598,
+      "learning_rate": 1.8197210620311297e-05,
+      "loss": 0.2055,
+      "step": 4293
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3678290393221864,
+      "learning_rate": 1.819626718655565e-05,
+      "loss": 0.2159,
+      "step": 4294
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1708498945803798,
+      "learning_rate": 1.819532353047579e-05,
+      "loss": 0.2329,
+      "step": 4295
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1560113389228808,
+      "learning_rate": 1.819437965209732e-05,
+      "loss": 0.2324,
+      "step": 4296
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9184708309489177,
+      "learning_rate": 1.819343555144584e-05,
+      "loss": 0.2311,
+      "step": 4297
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9343546421034021,
+      "learning_rate": 1.8192491228546953e-05,
+      "loss": 0.199,
+      "step": 4298
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8418429083232742,
+      "learning_rate": 1.8191546683426283e-05,
+      "loss": 0.2182,
+      "step": 4299
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1363584516751166,
+      "learning_rate": 1.8190601916109444e-05,
+      "loss": 0.2071,
+      "step": 4300
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9556669160393405,
+      "learning_rate": 1.8189656926622066e-05,
+      "loss": 0.227,
+      "step": 4301
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0732355260283306,
+      "learning_rate": 1.8188711714989785e-05,
+      "loss": 0.2038,
+      "step": 4302
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7696215465058844,
+      "learning_rate": 1.818776628123823e-05,
+      "loss": 0.1983,
+      "step": 4303
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2331616863270052,
+      "learning_rate": 1.818682062539306e-05,
+      "loss": 0.2271,
+      "step": 4304
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0942386697318467,
+      "learning_rate": 1.8185874747479912e-05,
+      "loss": 0.2177,
+      "step": 4305
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4450879115009336,
+      "learning_rate": 1.818492864752445e-05,
+      "loss": 0.2128,
+      "step": 4306
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8904155962150285,
+      "learning_rate": 1.8183982325552338e-05,
+      "loss": 0.2184,
+      "step": 4307
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9143306634271516,
+      "learning_rate": 1.8183035781589238e-05,
+      "loss": 0.2208,
+      "step": 4308
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9853438320265109,
+      "learning_rate": 1.8182089015660836e-05,
+      "loss": 0.2013,
+      "step": 4309
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9777431632030403,
+      "learning_rate": 1.8181142027792807e-05,
+      "loss": 0.229,
+      "step": 4310
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8076202137496955,
+      "learning_rate": 1.8180194818010833e-05,
+      "loss": 0.193,
+      "step": 4311
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8673930948175105,
+      "learning_rate": 1.8179247386340617e-05,
+      "loss": 0.2152,
+      "step": 4312
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3583227306954313,
+      "learning_rate": 1.8178299732807853e-05,
+      "loss": 0.2024,
+      "step": 4313
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9205591976032004,
+      "learning_rate": 1.8177351857438248e-05,
+      "loss": 0.2488,
+      "step": 4314
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.082096516321356,
+      "learning_rate": 1.817640376025751e-05,
+      "loss": 0.2381,
+      "step": 4315
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4396388035174683,
+      "learning_rate": 1.817545544129136e-05,
+      "loss": 0.247,
+      "step": 4316
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8504382943282013,
+      "learning_rate": 1.8174506900565518e-05,
+      "loss": 0.1785,
+      "step": 4317
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9081660689283023,
+      "learning_rate": 1.8173558138105717e-05,
+      "loss": 0.2023,
+      "step": 4318
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.446106615498979,
+      "learning_rate": 1.8172609153937685e-05,
+      "loss": 0.2047,
+      "step": 4319
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.005259860103065,
+      "learning_rate": 1.8171659948087175e-05,
+      "loss": 0.2266,
+      "step": 4320
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.111929062773582,
+      "learning_rate": 1.8170710520579923e-05,
+      "loss": 0.2047,
+      "step": 4321
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9763553401215549,
+      "learning_rate": 1.8169760871441683e-05,
+      "loss": 0.199,
+      "step": 4322
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2920743833932382,
+      "learning_rate": 1.8168811000698224e-05,
+      "loss": 0.2212,
+      "step": 4323
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0361696005188632,
+      "learning_rate": 1.8167860908375307e-05,
+      "loss": 0.1859,
+      "step": 4324
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9516237535805719,
+      "learning_rate": 1.8166910594498695e-05,
+      "loss": 0.2199,
+      "step": 4325
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7244801875399887,
+      "learning_rate": 1.8165960059094174e-05,
+      "loss": 0.2319,
+      "step": 4326
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.051239459001659,
+      "learning_rate": 1.8165009302187526e-05,
+      "loss": 0.2182,
+      "step": 4327
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8694760347180661,
+      "learning_rate": 1.816405832380454e-05,
+      "loss": 0.2181,
+      "step": 4328
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9093718969648045,
+      "learning_rate": 1.8163107123971012e-05,
+      "loss": 0.199,
+      "step": 4329
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7777573023668687,
+      "learning_rate": 1.816215570271274e-05,
+      "loss": 0.1974,
+      "step": 4330
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.887007737951543,
+      "learning_rate": 1.8161204060055533e-05,
+      "loss": 0.2242,
+      "step": 4331
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0061088241270033,
+      "learning_rate": 1.8160252196025205e-05,
+      "loss": 0.2336,
+      "step": 4332
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9102452692339625,
+      "learning_rate": 1.8159300110647576e-05,
+      "loss": 0.1983,
+      "step": 4333
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9184937177752597,
+      "learning_rate": 1.815834780394847e-05,
+      "loss": 0.1995,
+      "step": 4334
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8389755034590195,
+      "learning_rate": 1.8157395275953722e-05,
+      "loss": 0.22,
+      "step": 4335
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9664717962957975,
+      "learning_rate": 1.8156442526689165e-05,
+      "loss": 0.2253,
+      "step": 4336
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.206033123512191,
+      "learning_rate": 1.8155489556180643e-05,
+      "loss": 0.2136,
+      "step": 4337
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9918052284301853,
+      "learning_rate": 1.8154536364454003e-05,
+      "loss": 0.2266,
+      "step": 4338
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8834020921272379,
+      "learning_rate": 1.8153582951535108e-05,
+      "loss": 0.2002,
+      "step": 4339
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.474921529252916,
+      "learning_rate": 1.8152629317449814e-05,
+      "loss": 0.2103,
+      "step": 4340
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7762391625825757,
+      "learning_rate": 1.815167546222399e-05,
+      "loss": 0.1926,
+      "step": 4341
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0448223957540366,
+      "learning_rate": 1.8150721385883508e-05,
+      "loss": 0.2103,
+      "step": 4342
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8368969397679634,
+      "learning_rate": 1.8149767088454248e-05,
+      "loss": 0.205,
+      "step": 4343
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2714716714784662,
+      "learning_rate": 1.8148812569962095e-05,
+      "loss": 0.1951,
+      "step": 4344
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3140737363744213,
+      "learning_rate": 1.814785783043294e-05,
+      "loss": 0.211,
+      "step": 4345
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.909422860508571,
+      "learning_rate": 1.8146902869892682e-05,
+      "loss": 0.2139,
+      "step": 4346
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9128532924458792,
+      "learning_rate": 1.8145947688367224e-05,
+      "loss": 0.2059,
+      "step": 4347
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4055777535594336,
+      "learning_rate": 1.8144992285882478e-05,
+      "loss": 0.208,
+      "step": 4348
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6393726545336686,
+      "learning_rate": 1.814403666246435e-05,
+      "loss": 0.2216,
+      "step": 4349
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.012495449839806,
+      "learning_rate": 1.814308081813877e-05,
+      "loss": 0.227,
+      "step": 4350
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8809379247618583,
+      "learning_rate": 1.8142124752931662e-05,
+      "loss": 0.2018,
+      "step": 4351
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0518298070909586,
+      "learning_rate": 1.8141168466868962e-05,
+      "loss": 0.2372,
+      "step": 4352
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.185197093186988,
+      "learning_rate": 1.8140211959976608e-05,
+      "loss": 0.221,
+      "step": 4353
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8956554633281654,
+      "learning_rate": 1.8139255232280545e-05,
+      "loss": 0.2144,
+      "step": 4354
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8595694807094763,
+      "learning_rate": 1.813829828380672e-05,
+      "loss": 0.2269,
+      "step": 4355
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8742860722630785,
+      "learning_rate": 1.81373411145811e-05,
+      "loss": 0.207,
+      "step": 4356
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8565448494079153,
+      "learning_rate": 1.8136383724629637e-05,
+      "loss": 0.203,
+      "step": 4357
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7449865604280017,
+      "learning_rate": 1.8135426113978312e-05,
+      "loss": 0.1995,
+      "step": 4358
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9507927834239618,
+      "learning_rate": 1.8134468282653092e-05,
+      "loss": 0.2466,
+      "step": 4359
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8884456657081311,
+      "learning_rate": 1.8133510230679956e-05,
+      "loss": 0.2351,
+      "step": 4360
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9767879397277819,
+      "learning_rate": 1.8132551958084902e-05,
+      "loss": 0.2417,
+      "step": 4361
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.028331879327762,
+      "learning_rate": 1.8131593464893913e-05,
+      "loss": 0.194,
+      "step": 4362
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9500463986143157,
+      "learning_rate": 1.813063475113299e-05,
+      "loss": 0.2152,
+      "step": 4363
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1014894639629662,
+      "learning_rate": 1.8129675816828144e-05,
+      "loss": 0.2095,
+      "step": 4364
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8469646740472858,
+      "learning_rate": 1.8128716662005385e-05,
+      "loss": 0.2113,
+      "step": 4365
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8171205513709684,
+      "learning_rate": 1.812775728669072e-05,
+      "loss": 0.217,
+      "step": 4366
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.052823131487029,
+      "learning_rate": 1.812679769091018e-05,
+      "loss": 0.2206,
+      "step": 4367
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0175056353860659,
+      "learning_rate": 1.81258378746898e-05,
+      "loss": 0.2229,
+      "step": 4368
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8699858373898745,
+      "learning_rate": 1.8124877838055604e-05,
+      "loss": 0.2248,
+      "step": 4369
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9018639095903852,
+      "learning_rate": 1.8123917581033642e-05,
+      "loss": 0.2302,
+      "step": 4370
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8339848039989683,
+      "learning_rate": 1.8122957103649953e-05,
+      "loss": 0.2337,
+      "step": 4371
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8561161698245993,
+      "learning_rate": 1.81219964059306e-05,
+      "loss": 0.1902,
+      "step": 4372
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9040620220822301,
+      "learning_rate": 1.8121035487901627e-05,
+      "loss": 0.2363,
+      "step": 4373
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8117035779864158,
+      "learning_rate": 1.8120074349589114e-05,
+      "loss": 0.2052,
+      "step": 4374
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7757265435166741,
+      "learning_rate": 1.8119112991019123e-05,
+      "loss": 0.1952,
+      "step": 4375
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3018928098117228,
+      "learning_rate": 1.811815141221773e-05,
+      "loss": 0.2489,
+      "step": 4376
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8514347367780373,
+      "learning_rate": 1.8117189613211023e-05,
+      "loss": 0.2428,
+      "step": 4377
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9570291743456296,
+      "learning_rate": 1.8116227594025092e-05,
+      "loss": 0.198,
+      "step": 4378
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6959724960679516,
+      "learning_rate": 1.811526535468603e-05,
+      "loss": 0.2027,
+      "step": 4379
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0305907382430886,
+      "learning_rate": 1.8114302895219936e-05,
+      "loss": 0.2047,
+      "step": 4380
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.171345519575611,
+      "learning_rate": 1.8113340215652916e-05,
+      "loss": 0.2283,
+      "step": 4381
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8155710544954411,
+      "learning_rate": 1.8112377316011086e-05,
+      "loss": 0.2179,
+      "step": 4382
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7938806830241489,
+      "learning_rate": 1.8111414196320562e-05,
+      "loss": 0.1821,
+      "step": 4383
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8821549536458148,
+      "learning_rate": 1.811045085660747e-05,
+      "loss": 0.2104,
+      "step": 4384
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9977566410273979,
+      "learning_rate": 1.8109487296897944e-05,
+      "loss": 0.2039,
+      "step": 4385
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8429772323990313,
+      "learning_rate": 1.8108523517218112e-05,
+      "loss": 0.2079,
+      "step": 4386
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.932220713953027,
+      "learning_rate": 1.8107559517594123e-05,
+      "loss": 0.2242,
+      "step": 4387
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9463348327375277,
+      "learning_rate": 1.8106595298052124e-05,
+      "loss": 0.2109,
+      "step": 4388
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.058926891012995,
+      "learning_rate": 1.810563085861827e-05,
+      "loss": 0.221,
+      "step": 4389
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7967025288772267,
+      "learning_rate": 1.810466619931872e-05,
+      "loss": 0.2069,
+      "step": 4390
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3536066297444167,
+      "learning_rate": 1.8103701320179644e-05,
+      "loss": 0.2321,
+      "step": 4391
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.943141752961332,
+      "learning_rate": 1.8102736221227212e-05,
+      "loss": 0.2142,
+      "step": 4392
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8920841817138594,
+      "learning_rate": 1.81017709024876e-05,
+      "loss": 0.2064,
+      "step": 4393
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7738233796062852,
+      "learning_rate": 1.8100805363986996e-05,
+      "loss": 0.2137,
+      "step": 4394
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1456208241941903,
+      "learning_rate": 1.809983960575159e-05,
+      "loss": 0.2198,
+      "step": 4395
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7602743082590431,
+      "learning_rate": 1.8098873627807576e-05,
+      "loss": 0.2049,
+      "step": 4396
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8663595195507451,
+      "learning_rate": 1.8097907430181157e-05,
+      "loss": 0.2049,
+      "step": 4397
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9999632680595777,
+      "learning_rate": 1.8096941012898545e-05,
+      "loss": 0.2197,
+      "step": 4398
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.899024685249853,
+      "learning_rate": 1.8095974375985945e-05,
+      "loss": 0.2233,
+      "step": 4399
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9689895282474176,
+      "learning_rate": 1.8095007519469583e-05,
+      "loss": 0.2149,
+      "step": 4400
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8861221836852705,
+      "learning_rate": 1.8094040443375692e-05,
+      "loss": 0.201,
+      "step": 4401
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.147376630225734,
+      "learning_rate": 1.8093073147730492e-05,
+      "loss": 0.2381,
+      "step": 4402
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9875533678572314,
+      "learning_rate": 1.8092105632560227e-05,
+      "loss": 0.2176,
+      "step": 4403
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8599201761775375,
+      "learning_rate": 1.8091137897891138e-05,
+      "loss": 0.2246,
+      "step": 4404
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.548059073446093,
+      "learning_rate": 1.8090169943749477e-05,
+      "loss": 0.2429,
+      "step": 4405
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4045767740844242,
+      "learning_rate": 1.80892017701615e-05,
+      "loss": 0.2402,
+      "step": 4406
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0615450153346424,
+      "learning_rate": 1.808823337715347e-05,
+      "loss": 0.2311,
+      "step": 4407
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4170629407778401,
+      "learning_rate": 1.8087264764751645e-05,
+      "loss": 0.1938,
+      "step": 4408
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9763221746153268,
+      "learning_rate": 1.8086295932982315e-05,
+      "loss": 0.2535,
+      "step": 4409
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7656690291051992,
+      "learning_rate": 1.8085326881871747e-05,
+      "loss": 0.1918,
+      "step": 4410
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.136065388243066,
+      "learning_rate": 1.8084357611446232e-05,
+      "loss": 0.2491,
+      "step": 4411
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.082286482252834,
+      "learning_rate": 1.808338812173206e-05,
+      "loss": 0.1978,
+      "step": 4412
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6479564360605128,
+      "learning_rate": 1.8082418412755527e-05,
+      "loss": 0.2176,
+      "step": 4413
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8694686452576879,
+      "learning_rate": 1.808144848454294e-05,
+      "loss": 0.2314,
+      "step": 4414
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7858990839349932,
+      "learning_rate": 1.8080478337120604e-05,
+      "loss": 0.2163,
+      "step": 4415
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.839738455990263,
+      "learning_rate": 1.807950797051484e-05,
+      "loss": 0.1958,
+      "step": 4416
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2351394659797974,
+      "learning_rate": 1.8078537384751968e-05,
+      "loss": 0.2261,
+      "step": 4417
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9296011925545019,
+      "learning_rate": 1.8077566579858306e-05,
+      "loss": 0.2095,
+      "step": 4418
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9096268661088964,
+      "learning_rate": 1.80765955558602e-05,
+      "loss": 0.2349,
+      "step": 4419
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9189740778262212,
+      "learning_rate": 1.8075624312783984e-05,
+      "loss": 0.2157,
+      "step": 4420
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8732626554297037,
+      "learning_rate": 1.8074652850656e-05,
+      "loss": 0.2223,
+      "step": 4421
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5194779373667795,
+      "learning_rate": 1.80736811695026e-05,
+      "loss": 0.237,
+      "step": 4422
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.011478056498499,
+      "learning_rate": 1.807270926935014e-05,
+      "loss": 0.2526,
+      "step": 4423
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.072836618698817,
+      "learning_rate": 1.8071737150224993e-05,
+      "loss": 0.2207,
+      "step": 4424
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.003240262280801,
+      "learning_rate": 1.8070764812153518e-05,
+      "loss": 0.2195,
+      "step": 4425
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8553296303325555,
+      "learning_rate": 1.8069792255162088e-05,
+      "loss": 0.2143,
+      "step": 4426
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9706401479122935,
+      "learning_rate": 1.8068819479277087e-05,
+      "loss": 0.2185,
+      "step": 4427
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8534705006698775,
+      "learning_rate": 1.8067846484524905e-05,
+      "loss": 0.1879,
+      "step": 4428
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8578306627076671,
+      "learning_rate": 1.806687327093193e-05,
+      "loss": 0.2213,
+      "step": 4429
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9165415623957645,
+      "learning_rate": 1.806589983852456e-05,
+      "loss": 0.2171,
+      "step": 4430
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0313737327911443,
+      "learning_rate": 1.8064926187329205e-05,
+      "loss": 0.2255,
+      "step": 4431
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9197916913770108,
+      "learning_rate": 1.8063952317372265e-05,
+      "loss": 0.1972,
+      "step": 4432
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3479992755805241,
+      "learning_rate": 1.8062978228680172e-05,
+      "loss": 0.2628,
+      "step": 4433
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8254972992938581,
+      "learning_rate": 1.8062003921279335e-05,
+      "loss": 0.2115,
+      "step": 4434
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8963891068322509,
+      "learning_rate": 1.8061029395196186e-05,
+      "loss": 0.2027,
+      "step": 4435
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9369312287584682,
+      "learning_rate": 1.8060054650457157e-05,
+      "loss": 0.2128,
+      "step": 4436
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4957509743791233,
+      "learning_rate": 1.8059079687088695e-05,
+      "loss": 0.1977,
+      "step": 4437
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8374581910764776,
+      "learning_rate": 1.805810450511724e-05,
+      "loss": 0.2171,
+      "step": 4438
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9093962161701271,
+      "learning_rate": 1.8057129104569246e-05,
+      "loss": 0.2155,
+      "step": 4439
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0819334654052752,
+      "learning_rate": 1.8056153485471167e-05,
+      "loss": 0.1863,
+      "step": 4440
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8524882699421736,
+      "learning_rate": 1.805517764784947e-05,
+      "loss": 0.2006,
+      "step": 4441
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1360487230080902,
+      "learning_rate": 1.8054201591730627e-05,
+      "loss": 0.2413,
+      "step": 4442
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.91748364528214,
+      "learning_rate": 1.8053225317141106e-05,
+      "loss": 0.1981,
+      "step": 4443
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9272715829052295,
+      "learning_rate": 1.80522488241074e-05,
+      "loss": 0.2271,
+      "step": 4444
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8682592457592926,
+      "learning_rate": 1.8051272112655983e-05,
+      "loss": 0.2152,
+      "step": 4445
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9262847244346153,
+      "learning_rate": 1.805029518281336e-05,
+      "loss": 0.2093,
+      "step": 4446
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9188337809176588,
+      "learning_rate": 1.8049318034606025e-05,
+      "loss": 0.2187,
+      "step": 4447
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2028573072889497,
+      "learning_rate": 1.8048340668060483e-05,
+      "loss": 0.2444,
+      "step": 4448
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8717286742127507,
+      "learning_rate": 1.8047363083203245e-05,
+      "loss": 0.2009,
+      "step": 4449
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8561068852096136,
+      "learning_rate": 1.8046385280060832e-05,
+      "loss": 0.2097,
+      "step": 4450
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0878027107170953,
+      "learning_rate": 1.8045407258659763e-05,
+      "loss": 0.2059,
+      "step": 4451
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0482682779855905,
+      "learning_rate": 1.8044429019026565e-05,
+      "loss": 0.2468,
+      "step": 4452
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1067602177028681,
+      "learning_rate": 1.8043450561187775e-05,
+      "loss": 0.2366,
+      "step": 4453
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9203132473066146,
+      "learning_rate": 1.8042471885169934e-05,
+      "loss": 0.2343,
+      "step": 4454
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9013155181915874,
+      "learning_rate": 1.8041492990999595e-05,
+      "loss": 0.2291,
+      "step": 4455
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.002047390111081,
+      "learning_rate": 1.8040513878703296e-05,
+      "loss": 0.2341,
+      "step": 4456
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7729887234708337,
+      "learning_rate": 1.803953454830761e-05,
+      "loss": 0.2187,
+      "step": 4457
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8295487988881244,
+      "learning_rate": 1.803855499983909e-05,
+      "loss": 0.2292,
+      "step": 4458
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9514396998810606,
+      "learning_rate": 1.803757523332432e-05,
+      "loss": 0.2199,
+      "step": 4459
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8581480115536537,
+      "learning_rate": 1.8036595248789862e-05,
+      "loss": 0.2185,
+      "step": 4460
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9062317373994906,
+      "learning_rate": 1.8035615046262305e-05,
+      "loss": 0.2406,
+      "step": 4461
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7518853352440985,
+      "learning_rate": 1.8034634625768234e-05,
+      "loss": 0.2094,
+      "step": 4462
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7640860466551985,
+      "learning_rate": 1.803365398733425e-05,
+      "loss": 0.2243,
+      "step": 4463
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0101277920294256,
+      "learning_rate": 1.803267313098694e-05,
+      "loss": 0.237,
+      "step": 4464
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1549794199372658,
+      "learning_rate": 1.8031692056752926e-05,
+      "loss": 0.207,
+      "step": 4465
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0995658668374693,
+      "learning_rate": 1.8030710764658804e-05,
+      "loss": 0.2352,
+      "step": 4466
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9735906734273259,
+      "learning_rate": 1.8029729254731204e-05,
+      "loss": 0.1905,
+      "step": 4467
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8437175055191298,
+      "learning_rate": 1.8028747526996745e-05,
+      "loss": 0.2186,
+      "step": 4468
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7296734364054673,
+      "learning_rate": 1.802776558148205e-05,
+      "loss": 0.2166,
+      "step": 4469
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.047129361115201,
+      "learning_rate": 1.8026783418213768e-05,
+      "loss": 0.2186,
+      "step": 4470
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8087153659486408,
+      "learning_rate": 1.8025801037218528e-05,
+      "loss": 0.22,
+      "step": 4471
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8948010111876558,
+      "learning_rate": 1.8024818438522984e-05,
+      "loss": 0.2275,
+      "step": 4472
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4481573277116695,
+      "learning_rate": 1.802383562215378e-05,
+      "loss": 0.2121,
+      "step": 4473
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8894301960584611,
+      "learning_rate": 1.802285258813759e-05,
+      "loss": 0.2199,
+      "step": 4474
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8852898290343183,
+      "learning_rate": 1.8021869336501066e-05,
+      "loss": 0.2371,
+      "step": 4475
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.861758973541769,
+      "learning_rate": 1.8020885867270883e-05,
+      "loss": 0.2244,
+      "step": 4476
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8802216161853176,
+      "learning_rate": 1.8019902180473717e-05,
+      "loss": 0.2146,
+      "step": 4477
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7756772290225344,
+      "learning_rate": 1.8018918276136255e-05,
+      "loss": 0.1897,
+      "step": 4478
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7696394101610075,
+      "learning_rate": 1.801793415428518e-05,
+      "loss": 0.2345,
+      "step": 4479
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9654896233324601,
+      "learning_rate": 1.8016949814947188e-05,
+      "loss": 0.2156,
+      "step": 4480
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9399179508195146,
+      "learning_rate": 1.801596525814898e-05,
+      "loss": 0.2044,
+      "step": 4481
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8600710403521159,
+      "learning_rate": 1.8014980483917263e-05,
+      "loss": 0.2086,
+      "step": 4482
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4090094998911113,
+      "learning_rate": 1.8013995492278744e-05,
+      "loss": 0.1984,
+      "step": 4483
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7783023424568493,
+      "learning_rate": 1.8013010283260146e-05,
+      "loss": 0.2126,
+      "step": 4484
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9956937185992838,
+      "learning_rate": 1.801202485688819e-05,
+      "loss": 0.1867,
+      "step": 4485
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9183789056572046,
+      "learning_rate": 1.801103921318961e-05,
+      "loss": 0.205,
+      "step": 4486
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9357084137452508,
+      "learning_rate": 1.801005335219114e-05,
+      "loss": 0.2161,
+      "step": 4487
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8767707480218008,
+      "learning_rate": 1.8009067273919518e-05,
+      "loss": 0.2093,
+      "step": 4488
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8962239859524315,
+      "learning_rate": 1.8008080978401493e-05,
+      "loss": 0.2267,
+      "step": 4489
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.719171880216484,
+      "learning_rate": 1.8007094465663823e-05,
+      "loss": 0.2018,
+      "step": 4490
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2807368660385212,
+      "learning_rate": 1.8006107735733258e-05,
+      "loss": 0.2312,
+      "step": 4491
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0189680130523846,
+      "learning_rate": 1.8005120788636574e-05,
+      "loss": 0.2144,
+      "step": 4492
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8109524372881051,
+      "learning_rate": 1.8004133624400535e-05,
+      "loss": 0.1963,
+      "step": 4493
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7260968412969371,
+      "learning_rate": 1.8003146243051916e-05,
+      "loss": 0.2186,
+      "step": 4494
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8667090314175855,
+      "learning_rate": 1.8002158644617508e-05,
+      "loss": 0.1815,
+      "step": 4495
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8716288269195112,
+      "learning_rate": 1.800117082912409e-05,
+      "loss": 0.2088,
+      "step": 4496
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9035607795569606,
+      "learning_rate": 1.8000182796598466e-05,
+      "loss": 0.2078,
+      "step": 4497
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0125622963855248,
+      "learning_rate": 1.7999194547067426e-05,
+      "loss": 0.2469,
+      "step": 4498
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.291516448269081,
+      "learning_rate": 1.7998206080557786e-05,
+      "loss": 0.2066,
+      "step": 4499
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3954274607458725,
+      "learning_rate": 1.7997217397096355e-05,
+      "loss": 0.182,
+      "step": 4500
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0939380315378482,
+      "learning_rate": 1.7996228496709948e-05,
+      "loss": 0.2223,
+      "step": 4501
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1328049643732674,
+      "learning_rate": 1.7995239379425393e-05,
+      "loss": 0.2029,
+      "step": 4502
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8570053201008193,
+      "learning_rate": 1.7994250045269518e-05,
+      "loss": 0.2048,
+      "step": 4503
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4699377915752037,
+      "learning_rate": 1.799326049426916e-05,
+      "loss": 0.2187,
+      "step": 4504
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8427132865080169,
+      "learning_rate": 1.7992270726451158e-05,
+      "loss": 0.2206,
+      "step": 4505
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0109750267372164,
+      "learning_rate": 1.7991280741842362e-05,
+      "loss": 0.2087,
+      "step": 4506
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8486429289414997,
+      "learning_rate": 1.799029054046962e-05,
+      "loss": 0.2015,
+      "step": 4507
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8440700530295214,
+      "learning_rate": 1.7989300122359802e-05,
+      "loss": 0.1891,
+      "step": 4508
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3246815330558233,
+      "learning_rate": 1.7988309487539767e-05,
+      "loss": 0.2286,
+      "step": 4509
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0044511865123302,
+      "learning_rate": 1.7987318636036382e-05,
+      "loss": 0.2147,
+      "step": 4510
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9616052245640965,
+      "learning_rate": 1.798632756787653e-05,
+      "loss": 0.2006,
+      "step": 4511
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.007859581807647,
+      "learning_rate": 1.7985336283087092e-05,
+      "loss": 0.2149,
+      "step": 4512
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9269182161077623,
+      "learning_rate": 1.7984344781694957e-05,
+      "loss": 0.2327,
+      "step": 4513
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1873968613561616,
+      "learning_rate": 1.7983353063727014e-05,
+      "loss": 0.1958,
+      "step": 4514
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7675397374303244,
+      "learning_rate": 1.7982361129210172e-05,
+      "loss": 0.2071,
+      "step": 4515
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8842111254876002,
+      "learning_rate": 1.7981368978171336e-05,
+      "loss": 0.2177,
+      "step": 4516
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9063416103174554,
+      "learning_rate": 1.798037661063741e-05,
+      "loss": 0.2146,
+      "step": 4517
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8730461639506438,
+      "learning_rate": 1.7979384026635323e-05,
+      "loss": 0.2323,
+      "step": 4518
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9719249466422303,
+      "learning_rate": 1.7978391226191993e-05,
+      "loss": 0.2191,
+      "step": 4519
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.098487152706154,
+      "learning_rate": 1.7977398209334347e-05,
+      "loss": 0.2052,
+      "step": 4520
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7818334103477671,
+      "learning_rate": 1.7976404976089327e-05,
+      "loss": 0.2216,
+      "step": 4521
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3452405241980792,
+      "learning_rate": 1.797541152648387e-05,
+      "loss": 0.2454,
+      "step": 4522
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9202422091727288,
+      "learning_rate": 1.7974417860544924e-05,
+      "loss": 0.2089,
+      "step": 4523
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0015645377017552,
+      "learning_rate": 1.797342397829944e-05,
+      "loss": 0.222,
+      "step": 4524
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.944338652787783,
+      "learning_rate": 1.797242987977439e-05,
+      "loss": 0.2476,
+      "step": 4525
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.887805161891117,
+      "learning_rate": 1.7971435564996723e-05,
+      "loss": 0.2149,
+      "step": 4526
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.463548393298575,
+      "learning_rate": 1.7970441033993416e-05,
+      "loss": 0.2617,
+      "step": 4527
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.064025199962864,
+      "learning_rate": 1.7969446286791448e-05,
+      "loss": 0.2037,
+      "step": 4528
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7289751041322153,
+      "learning_rate": 1.7968451323417798e-05,
+      "loss": 0.1879,
+      "step": 4529
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1059318897182686,
+      "learning_rate": 1.7967456143899453e-05,
+      "loss": 0.2267,
+      "step": 4530
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8711198948794134,
+      "learning_rate": 1.7966460748263412e-05,
+      "loss": 0.2093,
+      "step": 4531
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8817964338927127,
+      "learning_rate": 1.7965465136536677e-05,
+      "loss": 0.2207,
+      "step": 4532
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.087682593067581,
+      "learning_rate": 1.7964469308746246e-05,
+      "loss": 0.2616,
+      "step": 4533
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9860911239425852,
+      "learning_rate": 1.7963473264919137e-05,
+      "loss": 0.2439,
+      "step": 4534
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8686097317747554,
+      "learning_rate": 1.7962477005082367e-05,
+      "loss": 0.2096,
+      "step": 4535
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.851414230084223,
+      "learning_rate": 1.7961480529262956e-05,
+      "loss": 0.232,
+      "step": 4536
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9712965457220927,
+      "learning_rate": 1.7960483837487935e-05,
+      "loss": 0.2178,
+      "step": 4537
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7603316153751536,
+      "learning_rate": 1.7959486929784344e-05,
+      "loss": 0.2008,
+      "step": 4538
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.305590804409271,
+      "learning_rate": 1.7958489806179214e-05,
+      "loss": 0.2322,
+      "step": 4539
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5399175125976559,
+      "learning_rate": 1.7957492466699606e-05,
+      "loss": 0.2103,
+      "step": 4540
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4627194244171826,
+      "learning_rate": 1.7956494911372564e-05,
+      "loss": 0.2199,
+      "step": 4541
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8683388348736799,
+      "learning_rate": 1.7955497140225145e-05,
+      "loss": 0.204,
+      "step": 4542
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9920027394506458,
+      "learning_rate": 1.7954499153284418e-05,
+      "loss": 0.214,
+      "step": 4543
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.859917912663988,
+      "learning_rate": 1.7953500950577453e-05,
+      "loss": 0.1982,
+      "step": 4544
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0837324969574404,
+      "learning_rate": 1.7952502532131326e-05,
+      "loss": 0.2313,
+      "step": 4545
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8575927627504152,
+      "learning_rate": 1.795150389797312e-05,
+      "loss": 0.1942,
+      "step": 4546
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0625777986683949,
+      "learning_rate": 1.7950505048129917e-05,
+      "loss": 0.2335,
+      "step": 4547
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0713708397729715,
+      "learning_rate": 1.794950598262882e-05,
+      "loss": 0.2093,
+      "step": 4548
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.229848965244706,
+      "learning_rate": 1.794850670149692e-05,
+      "loss": 0.2176,
+      "step": 4549
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1091741040774479,
+      "learning_rate": 1.794750720476133e-05,
+      "loss": 0.231,
+      "step": 4550
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2156486310229675,
+      "learning_rate": 1.7946507492449158e-05,
+      "loss": 0.2472,
+      "step": 4551
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1075155329929134,
+      "learning_rate": 1.794550756458752e-05,
+      "loss": 0.2283,
+      "step": 4552
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3998141024285211,
+      "learning_rate": 1.794450742120354e-05,
+      "loss": 0.2225,
+      "step": 4553
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.011647270613291,
+      "learning_rate": 1.794350706232435e-05,
+      "loss": 0.2363,
+      "step": 4554
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9205082012421598,
+      "learning_rate": 1.794250648797708e-05,
+      "loss": 0.1958,
+      "step": 4555
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9069710312413244,
+      "learning_rate": 1.7941505698188873e-05,
+      "loss": 0.2098,
+      "step": 4556
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4467541990079829,
+      "learning_rate": 1.7940504692986876e-05,
+      "loss": 0.2196,
+      "step": 4557
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.993067278432018,
+      "learning_rate": 1.793950347239824e-05,
+      "loss": 0.2426,
+      "step": 4558
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8076871561454932,
+      "learning_rate": 1.7938502036450128e-05,
+      "loss": 0.2241,
+      "step": 4559
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2539845527913558,
+      "learning_rate": 1.7937500385169695e-05,
+      "loss": 0.2151,
+      "step": 4560
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8088486414006637,
+      "learning_rate": 1.793649851858412e-05,
+      "loss": 0.2041,
+      "step": 4561
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.186444873543506,
+      "learning_rate": 1.793549643672057e-05,
+      "loss": 0.2007,
+      "step": 4562
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.041431065850028,
+      "learning_rate": 1.793449413960623e-05,
+      "loss": 0.2139,
+      "step": 4563
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5800775392251105,
+      "learning_rate": 1.7933491627268297e-05,
+      "loss": 0.2198,
+      "step": 4564
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9856604566007479,
+      "learning_rate": 1.793248889973395e-05,
+      "loss": 0.2261,
+      "step": 4565
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2768031361915468,
+      "learning_rate": 1.793148595703039e-05,
+      "loss": 0.211,
+      "step": 4566
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2841177512879824,
+      "learning_rate": 1.793048279918483e-05,
+      "loss": 0.1976,
+      "step": 4567
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9419078723736817,
+      "learning_rate": 1.7929479426224473e-05,
+      "loss": 0.2049,
+      "step": 4568
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.17988154802931,
+      "learning_rate": 1.792847583817654e-05,
+      "loss": 0.204,
+      "step": 4569
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.230600233347415,
+      "learning_rate": 1.7927472035068252e-05,
+      "loss": 0.2062,
+      "step": 4570
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9443095266324063,
+      "learning_rate": 1.792646801692684e-05,
+      "loss": 0.2091,
+      "step": 4571
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8913426583688561,
+      "learning_rate": 1.7925463783779534e-05,
+      "loss": 0.2403,
+      "step": 4572
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1466558580416424,
+      "learning_rate": 1.7924459335653575e-05,
+      "loss": 0.1967,
+      "step": 4573
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8535354994960708,
+      "learning_rate": 1.7923454672576208e-05,
+      "loss": 0.2254,
+      "step": 4574
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9666574182720151,
+      "learning_rate": 1.7922449794574686e-05,
+      "loss": 0.2471,
+      "step": 4575
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0837227290752445,
+      "learning_rate": 1.7921444701676267e-05,
+      "loss": 0.2148,
+      "step": 4576
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0686405368447454,
+      "learning_rate": 1.7920439393908212e-05,
+      "loss": 0.2152,
+      "step": 4577
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9850592832279286,
+      "learning_rate": 1.791943387129779e-05,
+      "loss": 0.2447,
+      "step": 4578
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.155939191794011,
+      "learning_rate": 1.7918428133872278e-05,
+      "loss": 0.2238,
+      "step": 4579
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1574756641823711,
+      "learning_rate": 1.7917422181658954e-05,
+      "loss": 0.2134,
+      "step": 4580
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2070656126027186,
+      "learning_rate": 1.791641601468511e-05,
+      "loss": 0.1976,
+      "step": 4581
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.436803255730702,
+      "learning_rate": 1.791540963297803e-05,
+      "loss": 0.2095,
+      "step": 4582
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9244624436617368,
+      "learning_rate": 1.791440303656502e-05,
+      "loss": 0.226,
+      "step": 4583
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9845022785819262,
+      "learning_rate": 1.791339622547338e-05,
+      "loss": 0.2091,
+      "step": 4584
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5238633874008007,
+      "learning_rate": 1.791238919973042e-05,
+      "loss": 0.2155,
+      "step": 4585
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.213475206616372,
+      "learning_rate": 1.7911381959363456e-05,
+      "loss": 0.2218,
+      "step": 4586
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8457478367488127,
+      "learning_rate": 1.7910374504399812e-05,
+      "loss": 0.2093,
+      "step": 4587
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9261967112688656,
+      "learning_rate": 1.7909366834866805e-05,
+      "loss": 0.2165,
+      "step": 4588
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1208944218946013,
+      "learning_rate": 1.7908358950791784e-05,
+      "loss": 0.2366,
+      "step": 4589
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0587552863356118,
+      "learning_rate": 1.7907350852202078e-05,
+      "loss": 0.1997,
+      "step": 4590
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7929672703579542,
+      "learning_rate": 1.7906342539125032e-05,
+      "loss": 0.2123,
+      "step": 4591
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0807088767456519,
+      "learning_rate": 1.7905334011588e-05,
+      "loss": 0.214,
+      "step": 4592
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1805389793523693,
+      "learning_rate": 1.7904325269618335e-05,
+      "loss": 0.251,
+      "step": 4593
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.530345554012429,
+      "learning_rate": 1.7903316313243398e-05,
+      "loss": 0.2266,
+      "step": 4594
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2375683235074402,
+      "learning_rate": 1.790230714249056e-05,
+      "loss": 0.2357,
+      "step": 4595
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2876949405743958,
+      "learning_rate": 1.7901297757387198e-05,
+      "loss": 0.2271,
+      "step": 4596
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0113586380944988,
+      "learning_rate": 1.7900288157960687e-05,
+      "loss": 0.2156,
+      "step": 4597
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9330662813880115,
+      "learning_rate": 1.7899278344238414e-05,
+      "loss": 0.2374,
+      "step": 4598
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.393123277375178,
+      "learning_rate": 1.7898268316247767e-05,
+      "loss": 0.2137,
+      "step": 4599
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9077590609258267,
+      "learning_rate": 1.7897258074016152e-05,
+      "loss": 0.2313,
+      "step": 4600
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2412729292962352,
+      "learning_rate": 1.789624761757096e-05,
+      "loss": 0.2221,
+      "step": 4601
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0000254167505898,
+      "learning_rate": 1.7895236946939605e-05,
+      "loss": 0.2125,
+      "step": 4602
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1119944212749695,
+      "learning_rate": 1.7894226062149504e-05,
+      "loss": 0.1924,
+      "step": 4603
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9252582068282856,
+      "learning_rate": 1.7893214963228075e-05,
+      "loss": 0.2079,
+      "step": 4604
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6891110956771893,
+      "learning_rate": 1.7892203650202747e-05,
+      "loss": 0.2346,
+      "step": 4605
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1950376349533391,
+      "learning_rate": 1.7891192123100945e-05,
+      "loss": 0.2271,
+      "step": 4606
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.437835315691099,
+      "learning_rate": 1.7890180381950113e-05,
+      "loss": 0.2208,
+      "step": 4607
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3801696979354348,
+      "learning_rate": 1.7889168426777693e-05,
+      "loss": 0.2373,
+      "step": 4608
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.109176689844192,
+      "learning_rate": 1.7888156257611134e-05,
+      "loss": 0.1968,
+      "step": 4609
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.687963492958192,
+      "learning_rate": 1.7887143874477887e-05,
+      "loss": 0.2459,
+      "step": 4610
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7422820151768772,
+      "learning_rate": 1.788613127740542e-05,
+      "loss": 0.2015,
+      "step": 4611
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7380741907169548,
+      "learning_rate": 1.7885118466421198e-05,
+      "loss": 0.215,
+      "step": 4612
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.904065245765066,
+      "learning_rate": 1.788410544155269e-05,
+      "loss": 0.2017,
+      "step": 4613
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3133917809519462,
+      "learning_rate": 1.788309220282738e-05,
+      "loss": 0.1889,
+      "step": 4614
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5825665488429745,
+      "learning_rate": 1.788207875027274e-05,
+      "loss": 0.2168,
+      "step": 4615
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4843139736052093,
+      "learning_rate": 1.788106508391628e-05,
+      "loss": 0.2074,
+      "step": 4616
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7859163752239253,
+      "learning_rate": 1.788005120378548e-05,
+      "loss": 0.2277,
+      "step": 4617
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1062328359599314,
+      "learning_rate": 1.787903710990784e-05,
+      "loss": 0.2155,
+      "step": 4618
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.509252746197654,
+      "learning_rate": 1.7878022802310882e-05,
+      "loss": 0.1734,
+      "step": 4619
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3313165535638438,
+      "learning_rate": 1.7877008281022107e-05,
+      "loss": 0.2105,
+      "step": 4620
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8325763249283111,
+      "learning_rate": 1.7875993546069036e-05,
+      "loss": 0.1965,
+      "step": 4621
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.154417787978655,
+      "learning_rate": 1.7874978597479196e-05,
+      "loss": 0.2303,
+      "step": 4622
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.319512505731216,
+      "learning_rate": 1.7873963435280122e-05,
+      "loss": 0.2245,
+      "step": 4623
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9450188451477706,
+      "learning_rate": 1.787294805949934e-05,
+      "loss": 0.2398,
+      "step": 4624
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4580963839574002,
+      "learning_rate": 1.7871932470164396e-05,
+      "loss": 0.2106,
+      "step": 4625
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0446207440063051,
+      "learning_rate": 1.787091666730284e-05,
+      "loss": 0.2261,
+      "step": 4626
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.675123317131691,
+      "learning_rate": 1.7869900650942228e-05,
+      "loss": 0.1998,
+      "step": 4627
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4430054449111767,
+      "learning_rate": 1.7868884421110115e-05,
+      "loss": 0.2103,
+      "step": 4628
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9541529011227019,
+      "learning_rate": 1.7867867977834067e-05,
+      "loss": 0.2249,
+      "step": 4629
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3281638875589192,
+      "learning_rate": 1.7866851321141655e-05,
+      "loss": 0.2276,
+      "step": 4630
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.151641575951757,
+      "learning_rate": 1.7865834451060458e-05,
+      "loss": 0.2107,
+      "step": 4631
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2337527355105584,
+      "learning_rate": 1.7864817367618058e-05,
+      "loss": 0.2081,
+      "step": 4632
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5264640015634976,
+      "learning_rate": 1.7863800070842038e-05,
+      "loss": 0.1938,
+      "step": 4633
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1466772922197281,
+      "learning_rate": 1.7862782560760004e-05,
+      "loss": 0.2127,
+      "step": 4634
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.273989830855673,
+      "learning_rate": 1.7861764837399544e-05,
+      "loss": 0.1881,
+      "step": 4635
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3201828598781473,
+      "learning_rate": 1.786074690078827e-05,
+      "loss": 0.2431,
+      "step": 4636
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5748727581569109,
+      "learning_rate": 1.785972875095379e-05,
+      "loss": 0.1946,
+      "step": 4637
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.450493532647831,
+      "learning_rate": 1.785871038792373e-05,
+      "loss": 0.1972,
+      "step": 4638
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2015481182386216,
+      "learning_rate": 1.7857691811725702e-05,
+      "loss": 0.2096,
+      "step": 4639
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.25848049603337,
+      "learning_rate": 1.785667302238734e-05,
+      "loss": 0.2265,
+      "step": 4640
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.349525557841808,
+      "learning_rate": 1.785565401993628e-05,
+      "loss": 0.2234,
+      "step": 4641
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8130057352094755,
+      "learning_rate": 1.785463480440016e-05,
+      "loss": 0.2088,
+      "step": 4642
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8040832321063347,
+      "learning_rate": 1.7853615375806627e-05,
+      "loss": 0.2371,
+      "step": 4643
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4228604220338537,
+      "learning_rate": 1.7852595734183333e-05,
+      "loss": 0.216,
+      "step": 4644
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.310556142843705,
+      "learning_rate": 1.7851575879557937e-05,
+      "loss": 0.2051,
+      "step": 4645
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7051940112440398,
+      "learning_rate": 1.78505558119581e-05,
+      "loss": 0.1949,
+      "step": 4646
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.366779457579173,
+      "learning_rate": 1.7849535531411498e-05,
+      "loss": 0.1994,
+      "step": 4647
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.223823201496756,
+      "learning_rate": 1.7848515037945797e-05,
+      "loss": 0.2092,
+      "step": 4648
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9146218576980645,
+      "learning_rate": 1.784749433158868e-05,
+      "loss": 0.2279,
+      "step": 4649
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8574756201466331,
+      "learning_rate": 1.7846473412367845e-05,
+      "loss": 0.2066,
+      "step": 4650
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.236915808421173,
+      "learning_rate": 1.7845452280310967e-05,
+      "loss": 0.217,
+      "step": 4651
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.721888692196491,
+      "learning_rate": 1.784443093544576e-05,
+      "loss": 0.2415,
+      "step": 4652
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.487684047623923,
+      "learning_rate": 1.7843409377799914e-05,
+      "loss": 0.2167,
+      "step": 4653
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7412624169564745,
+      "learning_rate": 1.7842387607401148e-05,
+      "loss": 0.2083,
+      "step": 4654
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.40719210246589,
+      "learning_rate": 1.7841365624277176e-05,
+      "loss": 0.2482,
+      "step": 4655
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.529733771727649,
+      "learning_rate": 1.7840343428455716e-05,
+      "loss": 0.2119,
+      "step": 4656
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7667059024593066,
+      "learning_rate": 1.78393210199645e-05,
+      "loss": 0.2074,
+      "step": 4657
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1288614053658006,
+      "learning_rate": 1.7838298398831263e-05,
+      "loss": 0.2383,
+      "step": 4658
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.026408149145776,
+      "learning_rate": 1.783727556508373e-05,
+      "loss": 0.2058,
+      "step": 4659
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4706443339849393,
+      "learning_rate": 1.783625251874966e-05,
+      "loss": 0.2228,
+      "step": 4660
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0917198179045484,
+      "learning_rate": 1.78352292598568e-05,
+      "loss": 0.2419,
+      "step": 4661
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9710987685726439,
+      "learning_rate": 1.78342057884329e-05,
+      "loss": 0.2331,
+      "step": 4662
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2486942535582013,
+      "learning_rate": 1.7833182104505727e-05,
+      "loss": 0.2098,
+      "step": 4663
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4638630936861796,
+      "learning_rate": 1.7832158208103046e-05,
+      "loss": 0.2116,
+      "step": 4664
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2763911256155431,
+      "learning_rate": 1.7831134099252633e-05,
+      "loss": 0.228,
+      "step": 4665
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2850151021671732,
+      "learning_rate": 1.7830109777982264e-05,
+      "loss": 0.2158,
+      "step": 4666
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3120744860308333,
+      "learning_rate": 1.7829085244319722e-05,
+      "loss": 0.2157,
+      "step": 4667
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.495724487270258,
+      "learning_rate": 1.7828060498292807e-05,
+      "loss": 0.2049,
+      "step": 4668
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1252993166897158,
+      "learning_rate": 1.7827035539929304e-05,
+      "loss": 0.221,
+      "step": 4669
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0361332579638716,
+      "learning_rate": 1.7826010369257023e-05,
+      "loss": 0.2397,
+      "step": 4670
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9062157745950375,
+      "learning_rate": 1.7824984986303767e-05,
+      "loss": 0.2221,
+      "step": 4671
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0599935506985416,
+      "learning_rate": 1.782395939109735e-05,
+      "loss": 0.1988,
+      "step": 4672
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0761451179044146,
+      "learning_rate": 1.7822933583665595e-05,
+      "loss": 0.1832,
+      "step": 4673
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2453986166897961,
+      "learning_rate": 1.782190756403632e-05,
+      "loss": 0.2137,
+      "step": 4674
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9699109057442683,
+      "learning_rate": 1.7820881332237366e-05,
+      "loss": 0.2233,
+      "step": 4675
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0480747153682368,
+      "learning_rate": 1.7819854888296563e-05,
+      "loss": 0.2394,
+      "step": 4676
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.139973735215863,
+      "learning_rate": 1.7818828232241756e-05,
+      "loss": 0.2286,
+      "step": 4677
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6594254452523631,
+      "learning_rate": 1.781780136410079e-05,
+      "loss": 0.2264,
+      "step": 4678
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.892424483882286,
+      "learning_rate": 1.7816774283901518e-05,
+      "loss": 0.2145,
+      "step": 4679
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9634119808029931,
+      "learning_rate": 1.7815746991671804e-05,
+      "loss": 0.2322,
+      "step": 4680
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0708682845486202,
+      "learning_rate": 1.781471948743951e-05,
+      "loss": 0.2455,
+      "step": 4681
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5709346655885017,
+      "learning_rate": 1.781369177123251e-05,
+      "loss": 0.2179,
+      "step": 4682
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0173880811846758,
+      "learning_rate": 1.7812663843078677e-05,
+      "loss": 0.1937,
+      "step": 4683
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8846051241989114,
+      "learning_rate": 1.78116357030059e-05,
+      "loss": 0.1912,
+      "step": 4684
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.251193835858283,
+      "learning_rate": 1.7810607351042062e-05,
+      "loss": 0.2091,
+      "step": 4685
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1693427107436283,
+      "learning_rate": 1.780957878721506e-05,
+      "loss": 0.2114,
+      "step": 4686
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1338587726128928,
+      "learning_rate": 1.7808550011552788e-05,
+      "loss": 0.2243,
+      "step": 4687
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8121839280092648,
+      "learning_rate": 1.780752102408316e-05,
+      "loss": 0.2332,
+      "step": 4688
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9126963444453843,
+      "learning_rate": 1.780649182483408e-05,
+      "loss": 0.2059,
+      "step": 4689
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9287395868561739,
+      "learning_rate": 1.780546241383347e-05,
+      "loss": 0.2336,
+      "step": 4690
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.861224039628638,
+      "learning_rate": 1.7804432791109253e-05,
+      "loss": 0.193,
+      "step": 4691
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2011478859769236,
+      "learning_rate": 1.7803402956689353e-05,
+      "loss": 0.2309,
+      "step": 4692
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8349673231029032,
+      "learning_rate": 1.7802372910601707e-05,
+      "loss": 0.22,
+      "step": 4693
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1764232090330085,
+      "learning_rate": 1.7801342652874256e-05,
+      "loss": 0.2082,
+      "step": 4694
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.31740133504077,
+      "learning_rate": 1.7800312183534946e-05,
+      "loss": 0.1989,
+      "step": 4695
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8859932028097436,
+      "learning_rate": 1.7799281502611725e-05,
+      "loss": 0.2259,
+      "step": 4696
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.068731884988904,
+      "learning_rate": 1.7798250610132555e-05,
+      "loss": 0.2286,
+      "step": 4697
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9480913590614505,
+      "learning_rate": 1.7797219506125393e-05,
+      "loss": 0.2121,
+      "step": 4698
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1259551001945691,
+      "learning_rate": 1.7796188190618217e-05,
+      "loss": 0.2447,
+      "step": 4699
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0424075409961195,
+      "learning_rate": 1.7795156663638993e-05,
+      "loss": 0.2487,
+      "step": 4700
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.822807998801817,
+      "learning_rate": 1.7794124925215706e-05,
+      "loss": 0.2308,
+      "step": 4701
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9446368597507817,
+      "learning_rate": 1.7793092975376337e-05,
+      "loss": 0.2051,
+      "step": 4702
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8001210824626251,
+      "learning_rate": 1.779206081414888e-05,
+      "loss": 0.2054,
+      "step": 4703
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7007932947970836,
+      "learning_rate": 1.779102844156134e-05,
+      "loss": 0.2126,
+      "step": 4704
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.018612629375446,
+      "learning_rate": 1.778999585764171e-05,
+      "loss": 0.1843,
+      "step": 4705
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.571513674538329,
+      "learning_rate": 1.7788963062418e-05,
+      "loss": 0.2114,
+      "step": 4706
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7412928909282924,
+      "learning_rate": 1.778793005591823e-05,
+      "loss": 0.1917,
+      "step": 4707
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8497538991645085,
+      "learning_rate": 1.7786896838170414e-05,
+      "loss": 0.1995,
+      "step": 4708
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8581836762327298,
+      "learning_rate": 1.7785863409202587e-05,
+      "loss": 0.2296,
+      "step": 4709
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0205692084242963,
+      "learning_rate": 1.778482976904277e-05,
+      "loss": 0.2222,
+      "step": 4710
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7773003288749312,
+      "learning_rate": 1.7783795917719006e-05,
+      "loss": 0.2123,
+      "step": 4711
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8669931706685519,
+      "learning_rate": 1.7782761855259343e-05,
+      "loss": 0.196,
+      "step": 4712
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.261557946455817,
+      "learning_rate": 1.778172758169182e-05,
+      "loss": 0.2428,
+      "step": 4713
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8542981027231593,
+      "learning_rate": 1.77806930970445e-05,
+      "loss": 0.1949,
+      "step": 4714
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1725172257027152,
+      "learning_rate": 1.7779658401345437e-05,
+      "loss": 0.2447,
+      "step": 4715
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9171805823393477,
+      "learning_rate": 1.7778623494622703e-05,
+      "loss": 0.2291,
+      "step": 4716
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8329476077948494,
+      "learning_rate": 1.7777588376904367e-05,
+      "loss": 0.1999,
+      "step": 4717
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8706417523196315,
+      "learning_rate": 1.777655304821851e-05,
+      "loss": 0.1986,
+      "step": 4718
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9122189195503517,
+      "learning_rate": 1.7775517508593208e-05,
+      "loss": 0.2286,
+      "step": 4719
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9221898541481,
+      "learning_rate": 1.7774481758056553e-05,
+      "loss": 0.2033,
+      "step": 4720
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.913523118983051,
+      "learning_rate": 1.7773445796636647e-05,
+      "loss": 0.2321,
+      "step": 4721
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1853772461053191,
+      "learning_rate": 1.777240962436158e-05,
+      "loss": 0.2175,
+      "step": 4722
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9836678716893172,
+      "learning_rate": 1.7771373241259463e-05,
+      "loss": 0.2354,
+      "step": 4723
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.933538952221682,
+      "learning_rate": 1.777033664735841e-05,
+      "loss": 0.2198,
+      "step": 4724
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9053145119436731,
+      "learning_rate": 1.7769299842686537e-05,
+      "loss": 0.1993,
+      "step": 4725
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4561874749144696,
+      "learning_rate": 1.7768262827271967e-05,
+      "loss": 0.21,
+      "step": 4726
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8923224164256436,
+      "learning_rate": 1.7767225601142827e-05,
+      "loss": 0.2417,
+      "step": 4727
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1688299476678536,
+      "learning_rate": 1.7766188164327255e-05,
+      "loss": 0.2227,
+      "step": 4728
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1615482229124374,
+      "learning_rate": 1.7765150516853393e-05,
+      "loss": 0.2098,
+      "step": 4729
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9914331330545041,
+      "learning_rate": 1.776411265874938e-05,
+      "loss": 0.207,
+      "step": 4730
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.216108796145146,
+      "learning_rate": 1.7763074590043373e-05,
+      "loss": 0.2333,
+      "step": 4731
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0612278807406503,
+      "learning_rate": 1.7762036310763533e-05,
+      "loss": 0.2331,
+      "step": 4732
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9308158763736551,
+      "learning_rate": 1.7760997820938017e-05,
+      "loss": 0.2139,
+      "step": 4733
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.02496181162953,
+      "learning_rate": 1.7759959120594995e-05,
+      "loss": 0.2208,
+      "step": 4734
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9621707948335559,
+      "learning_rate": 1.7758920209762646e-05,
+      "loss": 0.2095,
+      "step": 4735
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8073536972901817,
+      "learning_rate": 1.7757881088469152e-05,
+      "loss": 0.213,
+      "step": 4736
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.602279863576911,
+      "learning_rate": 1.775684175674269e-05,
+      "loss": 0.1964,
+      "step": 4737
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7919440808875956,
+      "learning_rate": 1.7755802214611456e-05,
+      "loss": 0.2221,
+      "step": 4738
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2264881157763754,
+      "learning_rate": 1.7754762462103653e-05,
+      "loss": 0.2061,
+      "step": 4739
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0081452576317012,
+      "learning_rate": 1.775372249924748e-05,
+      "loss": 0.2341,
+      "step": 4740
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4966706046285734,
+      "learning_rate": 1.775268232607114e-05,
+      "loss": 0.1968,
+      "step": 4741
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9282699330047943,
+      "learning_rate": 1.775164194260286e-05,
+      "loss": 0.1917,
+      "step": 4742
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0831685554881305,
+      "learning_rate": 1.7750601348870857e-05,
+      "loss": 0.2245,
+      "step": 4743
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0721042421741023,
+      "learning_rate": 1.774956054490335e-05,
+      "loss": 0.226,
+      "step": 4744
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1269726140008207,
+      "learning_rate": 1.7748519530728578e-05,
+      "loss": 0.2066,
+      "step": 4745
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3184525037031192,
+      "learning_rate": 1.7747478306374774e-05,
+      "loss": 0.1888,
+      "step": 4746
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.447342958685647,
+      "learning_rate": 1.7746436871870185e-05,
+      "loss": 0.2092,
+      "step": 4747
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0632538284053596,
+      "learning_rate": 1.7745395227243057e-05,
+      "loss": 0.2367,
+      "step": 4748
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8571598963074442,
+      "learning_rate": 1.7744353372521645e-05,
+      "loss": 0.1998,
+      "step": 4749
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4045763430768659,
+      "learning_rate": 1.7743311307734212e-05,
+      "loss": 0.2075,
+      "step": 4750
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.140852461397593,
+      "learning_rate": 1.7742269032909022e-05,
+      "loss": 0.2327,
+      "step": 4751
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0450382616949419,
+      "learning_rate": 1.774122654807435e-05,
+      "loss": 0.2048,
+      "step": 4752
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9913868061762731,
+      "learning_rate": 1.7740183853258463e-05,
+      "loss": 0.2349,
+      "step": 4753
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9902220057154458,
+      "learning_rate": 1.773914094848966e-05,
+      "loss": 0.1981,
+      "step": 4754
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9141793025191702,
+      "learning_rate": 1.7738097833796218e-05,
+      "loss": 0.225,
+      "step": 4755
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.122258777024154,
+      "learning_rate": 1.7737054509206437e-05,
+      "loss": 0.2029,
+      "step": 4756
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9388722008012684,
+      "learning_rate": 1.773601097474861e-05,
+      "loss": 0.1799,
+      "step": 4757
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1338319108176849,
+      "learning_rate": 1.7734967230451053e-05,
+      "loss": 0.198,
+      "step": 4758
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1696782476614866,
+      "learning_rate": 1.7733923276342072e-05,
+      "loss": 0.2082,
+      "step": 4759
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9570850227198349,
+      "learning_rate": 1.7732879112449987e-05,
+      "loss": 0.1937,
+      "step": 4760
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9987829452675776,
+      "learning_rate": 1.773183473880312e-05,
+      "loss": 0.2065,
+      "step": 4761
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8658344198195547,
+      "learning_rate": 1.7730790155429796e-05,
+      "loss": 0.2535,
+      "step": 4762
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5105970576810357,
+      "learning_rate": 1.7729745362358354e-05,
+      "loss": 0.2292,
+      "step": 4763
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.030621919076012,
+      "learning_rate": 1.772870035961713e-05,
+      "loss": 0.2424,
+      "step": 4764
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7911133713120762,
+      "learning_rate": 1.772765514723448e-05,
+      "loss": 0.2031,
+      "step": 4765
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7733501456620687,
+      "learning_rate": 1.7726609725238736e-05,
+      "loss": 0.2015,
+      "step": 4766
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7483844203985907,
+      "learning_rate": 1.7725564093658273e-05,
+      "loss": 0.2144,
+      "step": 4767
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.832420970371779,
+      "learning_rate": 1.772451825252145e-05,
+      "loss": 0.1933,
+      "step": 4768
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9269564472066407,
+      "learning_rate": 1.7723472201856632e-05,
+      "loss": 0.2117,
+      "step": 4769
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7062623972950685,
+      "learning_rate": 1.7722425941692193e-05,
+      "loss": 0.2011,
+      "step": 4770
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.300632009889857,
+      "learning_rate": 1.7721379472056512e-05,
+      "loss": 0.235,
+      "step": 4771
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9395077257423332,
+      "learning_rate": 1.772033279297798e-05,
+      "loss": 0.214,
+      "step": 4772
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8855510946435825,
+      "learning_rate": 1.7719285904484984e-05,
+      "loss": 0.2153,
+      "step": 4773
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9062652247205932,
+      "learning_rate": 1.771823880660592e-05,
+      "loss": 0.2036,
+      "step": 4774
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2632706385556125,
+      "learning_rate": 1.7717191499369195e-05,
+      "loss": 0.2172,
+      "step": 4775
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0616280497428603,
+      "learning_rate": 1.7716143982803214e-05,
+      "loss": 0.214,
+      "step": 4776
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0610019253510121,
+      "learning_rate": 1.7715096256936387e-05,
+      "loss": 0.2148,
+      "step": 4777
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9265199693220165,
+      "learning_rate": 1.7714048321797146e-05,
+      "loss": 0.194,
+      "step": 4778
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4833546752386788,
+      "learning_rate": 1.7713000177413905e-05,
+      "loss": 0.2144,
+      "step": 4779
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.119156400854007,
+      "learning_rate": 1.77119518238151e-05,
+      "loss": 0.2137,
+      "step": 4780
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8713331110192771,
+      "learning_rate": 1.7710903261029162e-05,
+      "loss": 0.1895,
+      "step": 4781
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4736802633037827,
+      "learning_rate": 1.770985448908454e-05,
+      "loss": 0.2002,
+      "step": 4782
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9670421103121505,
+      "learning_rate": 1.770880550800968e-05,
+      "loss": 0.2246,
+      "step": 4783
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9997782459952628,
+      "learning_rate": 1.7707756317833037e-05,
+      "loss": 0.2155,
+      "step": 4784
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0048297808434645,
+      "learning_rate": 1.7706706918583065e-05,
+      "loss": 0.22,
+      "step": 4785
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8625274276807509,
+      "learning_rate": 1.7705657310288234e-05,
+      "loss": 0.1931,
+      "step": 4786
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0333429133868393,
+      "learning_rate": 1.7704607492977016e-05,
+      "loss": 0.2056,
+      "step": 4787
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8165586544427031,
+      "learning_rate": 1.770355746667788e-05,
+      "loss": 0.2255,
+      "step": 4788
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9446874307612251,
+      "learning_rate": 1.7702507231419316e-05,
+      "loss": 0.2284,
+      "step": 4789
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.157327733356626,
+      "learning_rate": 1.7701456787229805e-05,
+      "loss": 0.2121,
+      "step": 4790
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9067004989487868,
+      "learning_rate": 1.7700406134137846e-05,
+      "loss": 0.2084,
+      "step": 4791
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5375842708722038,
+      "learning_rate": 1.7699355272171936e-05,
+      "loss": 0.2111,
+      "step": 4792
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.579611023345228,
+      "learning_rate": 1.769830420136058e-05,
+      "loss": 0.2197,
+      "step": 4793
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7815056142120229,
+      "learning_rate": 1.7697252921732288e-05,
+      "loss": 0.2399,
+      "step": 4794
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0464247436420646,
+      "learning_rate": 1.7696201433315572e-05,
+      "loss": 0.2047,
+      "step": 4795
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7803618338517158,
+      "learning_rate": 1.769514973613896e-05,
+      "loss": 0.2282,
+      "step": 4796
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0589890669182576,
+      "learning_rate": 1.7694097830230977e-05,
+      "loss": 0.217,
+      "step": 4797
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8290031373697013,
+      "learning_rate": 1.7693045715620154e-05,
+      "loss": 0.2312,
+      "step": 4798
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9998618652420123,
+      "learning_rate": 1.7691993392335033e-05,
+      "loss": 0.1801,
+      "step": 4799
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.939470955111584,
+      "learning_rate": 1.7690940860404158e-05,
+      "loss": 0.2376,
+      "step": 4800
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7765224504735883,
+      "learning_rate": 1.7689888119856075e-05,
+      "loss": 0.1927,
+      "step": 4801
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.803742550457734,
+      "learning_rate": 1.7688835170719346e-05,
+      "loss": 0.2185,
+      "step": 4802
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0415101084628973,
+      "learning_rate": 1.7687782013022526e-05,
+      "loss": 0.1941,
+      "step": 4803
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7424308880860995,
+      "learning_rate": 1.7686728646794184e-05,
+      "loss": 0.2005,
+      "step": 4804
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8623911571639182,
+      "learning_rate": 1.7685675072062894e-05,
+      "loss": 0.2083,
+      "step": 4805
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8419284534298985,
+      "learning_rate": 1.7684621288857233e-05,
+      "loss": 0.2208,
+      "step": 4806
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7300335906306568,
+      "learning_rate": 1.7683567297205786e-05,
+      "loss": 0.2105,
+      "step": 4807
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.05541219527479,
+      "learning_rate": 1.7682513097137143e-05,
+      "loss": 0.2279,
+      "step": 4808
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0619360379113207,
+      "learning_rate": 1.76814586886799e-05,
+      "loss": 0.2034,
+      "step": 4809
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0246215374768057,
+      "learning_rate": 1.7680404071862653e-05,
+      "loss": 0.2148,
+      "step": 4810
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8031646650335315,
+      "learning_rate": 1.7679349246714012e-05,
+      "loss": 0.2003,
+      "step": 4811
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8335112525439298,
+      "learning_rate": 1.767829421326259e-05,
+      "loss": 0.2031,
+      "step": 4812
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5996760040622504,
+      "learning_rate": 1.7677238971537004e-05,
+      "loss": 0.224,
+      "step": 4813
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0144403532971613,
+      "learning_rate": 1.7676183521565876e-05,
+      "loss": 0.2219,
+      "step": 4814
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0481470431681141,
+      "learning_rate": 1.767512786337784e-05,
+      "loss": 0.2139,
+      "step": 4815
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3381315237653075,
+      "learning_rate": 1.7674071997001525e-05,
+      "loss": 0.2254,
+      "step": 4816
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7657901269899813,
+      "learning_rate": 1.767301592246557e-05,
+      "loss": 0.1899,
+      "step": 4817
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2415438690318472,
+      "learning_rate": 1.767195963979863e-05,
+      "loss": 0.2357,
+      "step": 4818
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.371155851734972,
+      "learning_rate": 1.767090314902935e-05,
+      "loss": 0.2256,
+      "step": 4819
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.027278506127918,
+      "learning_rate": 1.7669846450186384e-05,
+      "loss": 0.2021,
+      "step": 4820
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0454534167228677,
+      "learning_rate": 1.7668789543298407e-05,
+      "loss": 0.2124,
+      "step": 4821
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8698202989216001,
+      "learning_rate": 1.7667732428394077e-05,
+      "loss": 0.2048,
+      "step": 4822
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8752795249443815,
+      "learning_rate": 1.7666675105502073e-05,
+      "loss": 0.2053,
+      "step": 4823
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2287231186117786,
+      "learning_rate": 1.7665617574651074e-05,
+      "loss": 0.2013,
+      "step": 4824
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7811101027066473,
+      "learning_rate": 1.7664559835869763e-05,
+      "loss": 0.1969,
+      "step": 4825
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9831821506037443,
+      "learning_rate": 1.7663501889186837e-05,
+      "loss": 0.2071,
+      "step": 4826
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3321080093873667,
+      "learning_rate": 1.7662443734630987e-05,
+      "loss": 0.2057,
+      "step": 4827
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9328654862358803,
+      "learning_rate": 1.7661385372230918e-05,
+      "loss": 0.198,
+      "step": 4828
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9866929554002803,
+      "learning_rate": 1.766032680201534e-05,
+      "loss": 0.2252,
+      "step": 4829
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8040091299802148,
+      "learning_rate": 1.7659268024012962e-05,
+      "loss": 0.2133,
+      "step": 4830
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.929652719573642,
+      "learning_rate": 1.7658209038252507e-05,
+      "loss": 0.1771,
+      "step": 4831
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9325894915758147,
+      "learning_rate": 1.76571498447627e-05,
+      "loss": 0.228,
+      "step": 4832
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0497336023331287,
+      "learning_rate": 1.765609044357227e-05,
+      "loss": 0.2139,
+      "step": 4833
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0788425842404294,
+      "learning_rate": 1.7655030834709954e-05,
+      "loss": 0.2446,
+      "step": 4834
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8148673513502684,
+      "learning_rate": 1.7653971018204498e-05,
+      "loss": 0.2052,
+      "step": 4835
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.035118517547822,
+      "learning_rate": 1.7652910994084642e-05,
+      "loss": 0.1889,
+      "step": 4836
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8874921404778846,
+      "learning_rate": 1.7651850762379146e-05,
+      "loss": 0.2144,
+      "step": 4837
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.971609414329634,
+      "learning_rate": 1.7650790323116764e-05,
+      "loss": 0.2261,
+      "step": 4838
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9666178939534561,
+      "learning_rate": 1.764972967632626e-05,
+      "loss": 0.2039,
+      "step": 4839
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8557814059940211,
+      "learning_rate": 1.764866882203641e-05,
+      "loss": 0.2122,
+      "step": 4840
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.879809517867278,
+      "learning_rate": 1.7647607760275987e-05,
+      "loss": 0.2397,
+      "step": 4841
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1701570979540563,
+      "learning_rate": 1.764654649107377e-05,
+      "loss": 0.2348,
+      "step": 4842
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9819385381229103,
+      "learning_rate": 1.7645485014458545e-05,
+      "loss": 0.2138,
+      "step": 4843
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0372094524749425,
+      "learning_rate": 1.764442333045911e-05,
+      "loss": 0.2172,
+      "step": 4844
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1991618227826033,
+      "learning_rate": 1.764336143910426e-05,
+      "loss": 0.245,
+      "step": 4845
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2515952300942674,
+      "learning_rate": 1.76422993404228e-05,
+      "loss": 0.2346,
+      "step": 4846
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.162773600935501,
+      "learning_rate": 1.7641237034443535e-05,
+      "loss": 0.2349,
+      "step": 4847
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.198901515461344,
+      "learning_rate": 1.764017452119529e-05,
+      "loss": 0.2354,
+      "step": 4848
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9150902917184182,
+      "learning_rate": 1.7639111800706874e-05,
+      "loss": 0.2176,
+      "step": 4849
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9261690607074005,
+      "learning_rate": 1.7638048873007122e-05,
+      "loss": 0.2096,
+      "step": 4850
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9936123554925016,
+      "learning_rate": 1.7636985738124862e-05,
+      "loss": 0.2199,
+      "step": 4851
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8808245515832294,
+      "learning_rate": 1.7635922396088932e-05,
+      "loss": 0.1955,
+      "step": 4852
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.109091782939819,
+      "learning_rate": 1.7634858846928174e-05,
+      "loss": 0.2231,
+      "step": 4853
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8979865795412102,
+      "learning_rate": 1.7633795090671445e-05,
+      "loss": 0.2311,
+      "step": 4854
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1646120782227802,
+      "learning_rate": 1.7632731127347588e-05,
+      "loss": 0.2348,
+      "step": 4855
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8728535210494023,
+      "learning_rate": 1.763166695698547e-05,
+      "loss": 0.2663,
+      "step": 4856
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9821654410268483,
+      "learning_rate": 1.7630602579613952e-05,
+      "loss": 0.2136,
+      "step": 4857
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0888253463187754,
+      "learning_rate": 1.7629537995261913e-05,
+      "loss": 0.233,
+      "step": 4858
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.010011282244793,
+      "learning_rate": 1.7628473203958217e-05,
+      "loss": 0.1997,
+      "step": 4859
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0148151126522487,
+      "learning_rate": 1.7627408205731762e-05,
+      "loss": 0.1972,
+      "step": 4860
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0225367877080298,
+      "learning_rate": 1.7626343000611424e-05,
+      "loss": 0.2238,
+      "step": 4861
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0413589410630688,
+      "learning_rate": 1.7625277588626105e-05,
+      "loss": 0.1706,
+      "step": 4862
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9075675650831743,
+      "learning_rate": 1.76242119698047e-05,
+      "loss": 0.2213,
+      "step": 4863
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9576335241403465,
+      "learning_rate": 1.7623146144176114e-05,
+      "loss": 0.1945,
+      "step": 4864
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9841155313451155,
+      "learning_rate": 1.7622080111769257e-05,
+      "loss": 0.2088,
+      "step": 4865
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8088195105556512,
+      "learning_rate": 1.762101387261305e-05,
+      "loss": 0.2216,
+      "step": 4866
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7040924319850286,
+      "learning_rate": 1.7619947426736404e-05,
+      "loss": 0.2108,
+      "step": 4867
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8647540788705015,
+      "learning_rate": 1.761888077416826e-05,
+      "loss": 0.2055,
+      "step": 4868
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4917793213736366,
+      "learning_rate": 1.7617813914937544e-05,
+      "loss": 0.1867,
+      "step": 4869
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7589814055129415,
+      "learning_rate": 1.7616746849073195e-05,
+      "loss": 0.2267,
+      "step": 4870
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.064675978697834,
+      "learning_rate": 1.7615679576604157e-05,
+      "loss": 0.2145,
+      "step": 4871
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8149268942899142,
+      "learning_rate": 1.761461209755938e-05,
+      "loss": 0.2243,
+      "step": 4872
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9404075709892658,
+      "learning_rate": 1.761354441196782e-05,
+      "loss": 0.2033,
+      "step": 4873
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.066560634970023,
+      "learning_rate": 1.7612476519858437e-05,
+      "loss": 0.1929,
+      "step": 4874
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8873007469232738,
+      "learning_rate": 1.76114084212602e-05,
+      "loss": 0.2135,
+      "step": 4875
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0803485202137888,
+      "learning_rate": 1.761034011620208e-05,
+      "loss": 0.2024,
+      "step": 4876
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7213845346451273,
+      "learning_rate": 1.7609271604713055e-05,
+      "loss": 0.1998,
+      "step": 4877
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.121037230608576,
+      "learning_rate": 1.7608202886822107e-05,
+      "loss": 0.2315,
+      "step": 4878
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9896700881943471,
+      "learning_rate": 1.7607133962558226e-05,
+      "loss": 0.2267,
+      "step": 4879
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.266664377396315,
+      "learning_rate": 1.7606064831950403e-05,
+      "loss": 0.2078,
+      "step": 4880
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8553061465111234,
+      "learning_rate": 1.7604995495027645e-05,
+      "loss": 0.2116,
+      "step": 4881
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.042777046982744,
+      "learning_rate": 1.7603925951818954e-05,
+      "loss": 0.2112,
+      "step": 4882
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8740583129550602,
+      "learning_rate": 1.7602856202353346e-05,
+      "loss": 0.2023,
+      "step": 4883
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9759766435581416,
+      "learning_rate": 1.760178624665983e-05,
+      "loss": 0.2227,
+      "step": 4884
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8085400387638324,
+      "learning_rate": 1.760071608476743e-05,
+      "loss": 0.2034,
+      "step": 4885
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.833384795728059,
+      "learning_rate": 1.759964571670518e-05,
+      "loss": 0.2134,
+      "step": 4886
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9648949804215201,
+      "learning_rate": 1.7598575142502112e-05,
+      "loss": 0.2157,
+      "step": 4887
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.111719389453267,
+      "learning_rate": 1.7597504362187263e-05,
+      "loss": 0.1994,
+      "step": 4888
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8866865010030395,
+      "learning_rate": 1.759643337578968e-05,
+      "loss": 0.2051,
+      "step": 4889
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8504575300897302,
+      "learning_rate": 1.759536218333841e-05,
+      "loss": 0.232,
+      "step": 4890
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0365543055672188,
+      "learning_rate": 1.7594290784862516e-05,
+      "loss": 0.2141,
+      "step": 4891
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0281837501246553,
+      "learning_rate": 1.7593219180391053e-05,
+      "loss": 0.2039,
+      "step": 4892
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9744470951075427,
+      "learning_rate": 1.759214736995309e-05,
+      "loss": 0.2091,
+      "step": 4893
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9084367183414519,
+      "learning_rate": 1.7591075353577702e-05,
+      "loss": 0.226,
+      "step": 4894
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8317928554365319,
+      "learning_rate": 1.7590003131293967e-05,
+      "loss": 0.2417,
+      "step": 4895
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7541092473554227,
+      "learning_rate": 1.758893070313097e-05,
+      "loss": 0.1997,
+      "step": 4896
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7883788654157217,
+      "learning_rate": 1.7587858069117794e-05,
+      "loss": 0.17,
+      "step": 4897
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2109666508636387,
+      "learning_rate": 1.7586785229283543e-05,
+      "loss": 0.2183,
+      "step": 4898
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.721893397519787,
+      "learning_rate": 1.7585712183657312e-05,
+      "loss": 0.2016,
+      "step": 4899
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7380981007588767,
+      "learning_rate": 1.758463893226821e-05,
+      "loss": 0.2279,
+      "step": 4900
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1129843085791946,
+      "learning_rate": 1.758356547514535e-05,
+      "loss": 0.2218,
+      "step": 4901
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9678562315284823,
+      "learning_rate": 1.7582491812317846e-05,
+      "loss": 0.1974,
+      "step": 4902
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.131902572948813,
+      "learning_rate": 1.7581417943814827e-05,
+      "loss": 0.2106,
+      "step": 4903
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9348090511857947,
+      "learning_rate": 1.7580343869665416e-05,
+      "loss": 0.2027,
+      "step": 4904
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.457110700204861,
+      "learning_rate": 1.757926958989875e-05,
+      "loss": 0.211,
+      "step": 4905
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2125401140409062,
+      "learning_rate": 1.7578195104543964e-05,
+      "loss": 0.186,
+      "step": 4906
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1174812078893102,
+      "learning_rate": 1.7577120413630213e-05,
+      "loss": 0.2141,
+      "step": 4907
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.125396472928048,
+      "learning_rate": 1.757604551718664e-05,
+      "loss": 0.2079,
+      "step": 4908
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.887645588146478,
+      "learning_rate": 1.7574970415242407e-05,
+      "loss": 0.2031,
+      "step": 4909
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2491922728175417,
+      "learning_rate": 1.757389510782667e-05,
+      "loss": 0.2104,
+      "step": 4910
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9282309508640144,
+      "learning_rate": 1.75728195949686e-05,
+      "loss": 0.1882,
+      "step": 4911
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3550172160894849,
+      "learning_rate": 1.7571743876697377e-05,
+      "loss": 0.2003,
+      "step": 4912
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0095560289307193,
+      "learning_rate": 1.7570667953042167e-05,
+      "loss": 0.1877,
+      "step": 4913
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1898247035889686,
+      "learning_rate": 1.7569591824032168e-05,
+      "loss": 0.2217,
+      "step": 4914
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5094226280385097,
+      "learning_rate": 1.7568515489696558e-05,
+      "loss": 0.2195,
+      "step": 4915
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0270852292948054,
+      "learning_rate": 1.7567438950064542e-05,
+      "loss": 0.1776,
+      "step": 4916
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8945831319463787,
+      "learning_rate": 1.7566362205165313e-05,
+      "loss": 0.2195,
+      "step": 4917
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0540510808229446,
+      "learning_rate": 1.7565285255028083e-05,
+      "loss": 0.2182,
+      "step": 4918
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9530058829379024,
+      "learning_rate": 1.756420809968206e-05,
+      "loss": 0.2165,
+      "step": 4919
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9591411145842712,
+      "learning_rate": 1.756313073915647e-05,
+      "loss": 0.2227,
+      "step": 4920
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8538213378369253,
+      "learning_rate": 1.756205317348053e-05,
+      "loss": 0.1935,
+      "step": 4921
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.000411790089436,
+      "learning_rate": 1.756097540268347e-05,
+      "loss": 0.2327,
+      "step": 4922
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0420033351873208,
+      "learning_rate": 1.7559897426794528e-05,
+      "loss": 0.2128,
+      "step": 4923
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9623079713245224,
+      "learning_rate": 1.7558819245842938e-05,
+      "loss": 0.1922,
+      "step": 4924
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1098096767015824,
+      "learning_rate": 1.7557740859857953e-05,
+      "loss": 0.2266,
+      "step": 4925
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3491601834510454,
+      "learning_rate": 1.7556662268868817e-05,
+      "loss": 0.2058,
+      "step": 4926
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9653203134414965,
+      "learning_rate": 1.7555583472904788e-05,
+      "loss": 0.2003,
+      "step": 4927
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.864274668092387,
+      "learning_rate": 1.7554504471995134e-05,
+      "loss": 0.2023,
+      "step": 4928
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0880717825194903,
+      "learning_rate": 1.7553425266169118e-05,
+      "loss": 0.2149,
+      "step": 4929
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0489082281570887,
+      "learning_rate": 1.7552345855456017e-05,
+      "loss": 0.2129,
+      "step": 4930
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.052512902454979,
+      "learning_rate": 1.7551266239885104e-05,
+      "loss": 0.2266,
+      "step": 4931
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6579917251985647,
+      "learning_rate": 1.755018641948567e-05,
+      "loss": 0.1981,
+      "step": 4932
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7821592754300142,
+      "learning_rate": 1.7549106394287004e-05,
+      "loss": 0.2269,
+      "step": 4933
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8870274340321501,
+      "learning_rate": 1.75480261643184e-05,
+      "loss": 0.2051,
+      "step": 4934
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8643188466997564,
+      "learning_rate": 1.7546945729609162e-05,
+      "loss": 0.222,
+      "step": 4935
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8508180991596589,
+      "learning_rate": 1.7545865090188594e-05,
+      "loss": 0.2269,
+      "step": 4936
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0361811340997602,
+      "learning_rate": 1.7544784246086007e-05,
+      "loss": 0.1965,
+      "step": 4937
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7944009776152027,
+      "learning_rate": 1.7543703197330722e-05,
+      "loss": 0.2294,
+      "step": 4938
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2904691970827866,
+      "learning_rate": 1.754262194395206e-05,
+      "loss": 0.1905,
+      "step": 4939
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2425741683662073,
+      "learning_rate": 1.7541540485979357e-05,
+      "loss": 0.2298,
+      "step": 4940
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0170380196630249,
+      "learning_rate": 1.754045882344194e-05,
+      "loss": 0.2172,
+      "step": 4941
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8377036824445884,
+      "learning_rate": 1.753937695636915e-05,
+      "loss": 0.2257,
+      "step": 4942
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4429064752446406,
+      "learning_rate": 1.7538294884790333e-05,
+      "loss": 0.2265,
+      "step": 4943
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.002250270190205,
+      "learning_rate": 1.7537212608734842e-05,
+      "loss": 0.2341,
+      "step": 4944
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0738514608014265,
+      "learning_rate": 1.7536130128232035e-05,
+      "loss": 0.2029,
+      "step": 4945
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8699095271611501,
+      "learning_rate": 1.7535047443311274e-05,
+      "loss": 0.1972,
+      "step": 4946
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.115828821103785,
+      "learning_rate": 1.7533964554001923e-05,
+      "loss": 0.2171,
+      "step": 4947
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8999017703628961,
+      "learning_rate": 1.753288146033336e-05,
+      "loss": 0.2086,
+      "step": 4948
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9623098645662769,
+      "learning_rate": 1.753179816233496e-05,
+      "loss": 0.1941,
+      "step": 4949
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.148386681921377,
+      "learning_rate": 1.7530714660036112e-05,
+      "loss": 0.2169,
+      "step": 4950
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9251444910619677,
+      "learning_rate": 1.7529630953466202e-05,
+      "loss": 0.1907,
+      "step": 4951
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7047593827282227,
+      "learning_rate": 1.7528547042654626e-05,
+      "loss": 0.1885,
+      "step": 4952
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8934978764290634,
+      "learning_rate": 1.7527462927630786e-05,
+      "loss": 0.211,
+      "step": 4953
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0067227280512174,
+      "learning_rate": 1.752637860842409e-05,
+      "loss": 0.1983,
+      "step": 4954
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.594869450914085,
+      "learning_rate": 1.752529408506395e-05,
+      "loss": 0.2195,
+      "step": 4955
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9634578527813551,
+      "learning_rate": 1.7524209357579782e-05,
+      "loss": 0.2178,
+      "step": 4956
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2163076294046304,
+      "learning_rate": 1.752312442600101e-05,
+      "loss": 0.2299,
+      "step": 4957
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.827651648508648,
+      "learning_rate": 1.7522039290357066e-05,
+      "loss": 0.2053,
+      "step": 4958
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8964013881252514,
+      "learning_rate": 1.7520953950677374e-05,
+      "loss": 0.2302,
+      "step": 4959
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9926659228293512,
+      "learning_rate": 1.751986840699139e-05,
+      "loss": 0.1978,
+      "step": 4960
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4285563481945713,
+      "learning_rate": 1.7518782659328545e-05,
+      "loss": 0.2139,
+      "step": 4961
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8479583255530937,
+      "learning_rate": 1.7517696707718297e-05,
+      "loss": 0.204,
+      "step": 4962
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7944980238304211,
+      "learning_rate": 1.7516610552190104e-05,
+      "loss": 0.2052,
+      "step": 4963
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1398659985067,
+      "learning_rate": 1.751552419277342e-05,
+      "loss": 0.2133,
+      "step": 4964
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9269866998395468,
+      "learning_rate": 1.751443762949772e-05,
+      "loss": 0.2054,
+      "step": 4965
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.006432875016119,
+      "learning_rate": 1.7513350862392478e-05,
+      "loss": 0.2559,
+      "step": 4966
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8283899901874466,
+      "learning_rate": 1.7512263891487165e-05,
+      "loss": 0.2087,
+      "step": 4967
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9388658415746001,
+      "learning_rate": 1.7511176716811275e-05,
+      "loss": 0.2167,
+      "step": 4968
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.138438684960691,
+      "learning_rate": 1.7510089338394287e-05,
+      "loss": 0.1984,
+      "step": 4969
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1752742054523158,
+      "learning_rate": 1.7509001756265704e-05,
+      "loss": 0.2174,
+      "step": 4970
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9476066784239713,
+      "learning_rate": 1.7507913970455024e-05,
+      "loss": 0.2176,
+      "step": 4971
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.472689980809985,
+      "learning_rate": 1.750682598099175e-05,
+      "loss": 0.2131,
+      "step": 4972
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0265826255610757,
+      "learning_rate": 1.7505737787905404e-05,
+      "loss": 0.2237,
+      "step": 4973
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5419271677885265,
+      "learning_rate": 1.7504649391225493e-05,
+      "loss": 0.2308,
+      "step": 4974
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8149367596422321,
+      "learning_rate": 1.7503560790981545e-05,
+      "loss": 0.1969,
+      "step": 4975
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.104977220763604,
+      "learning_rate": 1.750247198720308e-05,
+      "loss": 0.2062,
+      "step": 4976
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9566919598399071,
+      "learning_rate": 1.750138297991965e-05,
+      "loss": 0.1797,
+      "step": 4977
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1870863493849544,
+      "learning_rate": 1.7500293769160773e-05,
+      "loss": 0.2253,
+      "step": 4978
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.976047243062067,
+      "learning_rate": 1.749920435495601e-05,
+      "loss": 0.2292,
+      "step": 4979
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8263157669810269,
+      "learning_rate": 1.7498114737334902e-05,
+      "loss": 0.1972,
+      "step": 4980
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9403072898460152,
+      "learning_rate": 1.749702491632701e-05,
+      "loss": 0.22,
+      "step": 4981
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2141379406623052,
+      "learning_rate": 1.749593489196189e-05,
+      "loss": 0.1972,
+      "step": 4982
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7395791963743559,
+      "learning_rate": 1.7494844664269117e-05,
+      "loss": 0.2383,
+      "step": 4983
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8786426046599091,
+      "learning_rate": 1.749375423327826e-05,
+      "loss": 0.2054,
+      "step": 4984
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8585034424431195,
+      "learning_rate": 1.7492663599018893e-05,
+      "loss": 0.2157,
+      "step": 4985
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9268853902764058,
+      "learning_rate": 1.7491572761520604e-05,
+      "loss": 0.2231,
+      "step": 4986
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0354260982717227,
+      "learning_rate": 1.749048172081298e-05,
+      "loss": 0.2163,
+      "step": 4987
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8511031855858545,
+      "learning_rate": 1.7489390476925616e-05,
+      "loss": 0.2072,
+      "step": 4988
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8786659576418835,
+      "learning_rate": 1.7488299029888117e-05,
+      "loss": 0.2133,
+      "step": 4989
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.812193258729004,
+      "learning_rate": 1.7487207379730078e-05,
+      "loss": 0.2303,
+      "step": 4990
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.377958020015114,
+      "learning_rate": 1.7486115526481117e-05,
+      "loss": 0.2166,
+      "step": 4991
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8504237116518911,
+      "learning_rate": 1.748502347017085e-05,
+      "loss": 0.2049,
+      "step": 4992
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9399276635507209,
+      "learning_rate": 1.74839312108289e-05,
+      "loss": 0.1958,
+      "step": 4993
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.834340680227393,
+      "learning_rate": 1.748283874848489e-05,
+      "loss": 0.2146,
+      "step": 4994
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1110556630181991,
+      "learning_rate": 1.748174608316846e-05,
+      "loss": 0.2286,
+      "step": 4995
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.208705629511072,
+      "learning_rate": 1.748065321490924e-05,
+      "loss": 0.1975,
+      "step": 4996
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2830729263319827,
+      "learning_rate": 1.7479560143736885e-05,
+      "loss": 0.1994,
+      "step": 4997
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9408077019695862,
+      "learning_rate": 1.7478466869681035e-05,
+      "loss": 0.198,
+      "step": 4998
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5477988939109983,
+      "learning_rate": 1.7477373392771352e-05,
+      "loss": 0.2293,
+      "step": 4999
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1463803370648873,
+      "learning_rate": 1.747627971303749e-05,
+      "loss": 0.2188,
+      "step": 5000
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8189903246416517,
+      "learning_rate": 1.7475185830509124e-05,
+      "loss": 0.2073,
+      "step": 5001
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2662754100903848,
+      "learning_rate": 1.7474091745215912e-05,
+      "loss": 0.2507,
+      "step": 5002
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1058798998476678,
+      "learning_rate": 1.7472997457187543e-05,
+      "loss": 0.2354,
+      "step": 5003
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9845891354226091,
+      "learning_rate": 1.74719029664537e-05,
+      "loss": 0.2061,
+      "step": 5004
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9242568026610173,
+      "learning_rate": 1.747080827304406e-05,
+      "loss": 0.2387,
+      "step": 5005
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9957505132812781,
+      "learning_rate": 1.746971337698833e-05,
+      "loss": 0.2267,
+      "step": 5006
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.130839216768605,
+      "learning_rate": 1.74686182783162e-05,
+      "loss": 0.2294,
+      "step": 5007
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0601687365795447,
+      "learning_rate": 1.7467522977057375e-05,
+      "loss": 0.2127,
+      "step": 5008
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4708274030519979,
+      "learning_rate": 1.746642747324157e-05,
+      "loss": 0.1969,
+      "step": 5009
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9259996950306866,
+      "learning_rate": 1.74653317668985e-05,
+      "loss": 0.2041,
+      "step": 5010
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.98602895533177,
+      "learning_rate": 1.7464235858057878e-05,
+      "loss": 0.204,
+      "step": 5011
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1042160602475162,
+      "learning_rate": 1.7463139746749443e-05,
+      "loss": 0.2173,
+      "step": 5012
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.328890259856893,
+      "learning_rate": 1.7462043433002915e-05,
+      "loss": 0.2021,
+      "step": 5013
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0417663851102328,
+      "learning_rate": 1.7460946916848042e-05,
+      "loss": 0.2108,
+      "step": 5014
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9022596817041325,
+      "learning_rate": 1.7459850198314562e-05,
+      "loss": 0.1964,
+      "step": 5015
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3093436909694542,
+      "learning_rate": 1.7458753277432223e-05,
+      "loss": 0.2168,
+      "step": 5016
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.579107620746813,
+      "learning_rate": 1.745765615423078e-05,
+      "loss": 0.195,
+      "step": 5017
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9487815613788572,
+      "learning_rate": 1.7456558828739993e-05,
+      "loss": 0.2148,
+      "step": 5018
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2607977547964462,
+      "learning_rate": 1.7455461300989627e-05,
+      "loss": 0.2157,
+      "step": 5019
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9884044516273045,
+      "learning_rate": 1.7454363571009452e-05,
+      "loss": 0.2017,
+      "step": 5020
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9234874314136033,
+      "learning_rate": 1.7453265638829246e-05,
+      "loss": 0.2083,
+      "step": 5021
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9101887407052885,
+      "learning_rate": 1.745216750447878e-05,
+      "loss": 0.2083,
+      "step": 5022
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9324602885160264,
+      "learning_rate": 1.7451069167987858e-05,
+      "loss": 0.2164,
+      "step": 5023
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.327783331629871,
+      "learning_rate": 1.7449970629386265e-05,
+      "loss": 0.1904,
+      "step": 5024
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9224667945636565,
+      "learning_rate": 1.7448871888703792e-05,
+      "loss": 0.1895,
+      "step": 5025
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8781524280150395,
+      "learning_rate": 1.744777294597025e-05,
+      "loss": 0.2284,
+      "step": 5026
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.106374290900814,
+      "learning_rate": 1.744667380121545e-05,
+      "loss": 0.2332,
+      "step": 5027
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2153769878955571,
+      "learning_rate": 1.7445574454469202e-05,
+      "loss": 0.2294,
+      "step": 5028
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.241271721583726,
+      "learning_rate": 1.744447490576132e-05,
+      "loss": 0.2285,
+      "step": 5029
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0396512825817412,
+      "learning_rate": 1.744337515512164e-05,
+      "loss": 0.2046,
+      "step": 5030
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.990253885865504,
+      "learning_rate": 1.744227520257999e-05,
+      "loss": 0.1984,
+      "step": 5031
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9219416918548322,
+      "learning_rate": 1.7441175048166203e-05,
+      "loss": 0.2099,
+      "step": 5032
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9096825504551227,
+      "learning_rate": 1.7440074691910123e-05,
+      "loss": 0.2179,
+      "step": 5033
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9631200723736666,
+      "learning_rate": 1.7438974133841596e-05,
+      "loss": 0.2316,
+      "step": 5034
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.951258868171326,
+      "learning_rate": 1.7437873373990478e-05,
+      "loss": 0.2176,
+      "step": 5035
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9007417553081281,
+      "learning_rate": 1.7436772412386622e-05,
+      "loss": 0.2053,
+      "step": 5036
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0672720680830408,
+      "learning_rate": 1.7435671249059895e-05,
+      "loss": 0.2219,
+      "step": 5037
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.913837085638295,
+      "learning_rate": 1.743456988404017e-05,
+      "loss": 0.2218,
+      "step": 5038
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8086054513570482,
+      "learning_rate": 1.743346831735731e-05,
+      "loss": 0.1971,
+      "step": 5039
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9734111808136677,
+      "learning_rate": 1.7432366549041203e-05,
+      "loss": 0.2252,
+      "step": 5040
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9860366282496003,
+      "learning_rate": 1.7431264579121734e-05,
+      "loss": 0.1978,
+      "step": 5041
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9523717046703011,
+      "learning_rate": 1.7430162407628796e-05,
+      "loss": 0.1968,
+      "step": 5042
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8510264207270812,
+      "learning_rate": 1.742906003459228e-05,
+      "loss": 0.228,
+      "step": 5043
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0463789344345444,
+      "learning_rate": 1.7427957460042092e-05,
+      "loss": 0.1962,
+      "step": 5044
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0830758748704274,
+      "learning_rate": 1.742685468400814e-05,
+      "loss": 0.2348,
+      "step": 5045
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1041907730218539,
+      "learning_rate": 1.7425751706520337e-05,
+      "loss": 0.2205,
+      "step": 5046
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1479770893064867,
+      "learning_rate": 1.7424648527608594e-05,
+      "loss": 0.2145,
+      "step": 5047
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9543567978429918,
+      "learning_rate": 1.742354514730284e-05,
+      "loss": 0.2198,
+      "step": 5048
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2461996880851625,
+      "learning_rate": 1.742244156563301e-05,
+      "loss": 0.2192,
+      "step": 5049
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0231145847760585,
+      "learning_rate": 1.742133778262903e-05,
+      "loss": 0.2013,
+      "step": 5050
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8834962163343502,
+      "learning_rate": 1.7420233798320848e-05,
+      "loss": 0.1999,
+      "step": 5051
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2485676337162055,
+      "learning_rate": 1.74191296127384e-05,
+      "loss": 0.2233,
+      "step": 5052
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.239142632740132,
+      "learning_rate": 1.7418025225911642e-05,
+      "loss": 0.2113,
+      "step": 5053
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9004485805006647,
+      "learning_rate": 1.7416920637870535e-05,
+      "loss": 0.192,
+      "step": 5054
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.938876208690321,
+      "learning_rate": 1.7415815848645032e-05,
+      "loss": 0.2166,
+      "step": 5055
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0012870727151062,
+      "learning_rate": 1.741471085826511e-05,
+      "loss": 0.2053,
+      "step": 5056
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9550687135535754,
+      "learning_rate": 1.7413605666760733e-05,
+      "loss": 0.2296,
+      "step": 5057
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0829094222653786,
+      "learning_rate": 1.7412500274161885e-05,
+      "loss": 0.2173,
+      "step": 5058
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1394625432475365,
+      "learning_rate": 1.741139468049855e-05,
+      "loss": 0.1889,
+      "step": 5059
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.934634491288444,
+      "learning_rate": 1.7410288885800716e-05,
+      "loss": 0.2201,
+      "step": 5060
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0153039446794108,
+      "learning_rate": 1.7409182890098372e-05,
+      "loss": 0.195,
+      "step": 5061
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6814140355594438,
+      "learning_rate": 1.7408076693421528e-05,
+      "loss": 0.2589,
+      "step": 5062
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.06895135418714,
+      "learning_rate": 1.7406970295800188e-05,
+      "loss": 0.2224,
+      "step": 5063
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.964438213306339,
+      "learning_rate": 1.7405863697264357e-05,
+      "loss": 0.2029,
+      "step": 5064
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.388743643425397,
+      "learning_rate": 1.7404756897844054e-05,
+      "loss": 0.2133,
+      "step": 5065
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.285559298704333,
+      "learning_rate": 1.7403649897569302e-05,
+      "loss": 0.2066,
+      "step": 5066
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8019958506858845,
+      "learning_rate": 1.740254269647013e-05,
+      "loss": 0.1981,
+      "step": 5067
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.224315249511999,
+      "learning_rate": 1.7401435294576566e-05,
+      "loss": 0.2235,
+      "step": 5068
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2892782160471126,
+      "learning_rate": 1.7400327691918657e-05,
+      "loss": 0.2211,
+      "step": 5069
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1369492211634333,
+      "learning_rate": 1.7399219888526438e-05,
+      "loss": 0.207,
+      "step": 5070
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8704337063333867,
+      "learning_rate": 1.7398111884429966e-05,
+      "loss": 0.1868,
+      "step": 5071
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9232480538530867,
+      "learning_rate": 1.7397003679659285e-05,
+      "loss": 0.1932,
+      "step": 5072
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.90226151158112,
+      "learning_rate": 1.7395895274244464e-05,
+      "loss": 0.2037,
+      "step": 5073
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0136440681877585,
+      "learning_rate": 1.7394786668215564e-05,
+      "loss": 0.1827,
+      "step": 5074
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1129209039716363,
+      "learning_rate": 1.739367786160266e-05,
+      "loss": 0.207,
+      "step": 5075
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.204496096714119,
+      "learning_rate": 1.7392568854435828e-05,
+      "loss": 0.2223,
+      "step": 5076
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9851629720093069,
+      "learning_rate": 1.7391459646745145e-05,
+      "loss": 0.2275,
+      "step": 5077
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8517560781459264,
+      "learning_rate": 1.7390350238560706e-05,
+      "loss": 0.1947,
+      "step": 5078
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9295862908238195,
+      "learning_rate": 1.7389240629912594e-05,
+      "loss": 0.2254,
+      "step": 5079
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9968261336024243,
+      "learning_rate": 1.7388130820830914e-05,
+      "loss": 0.2028,
+      "step": 5080
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0777716957127221,
+      "learning_rate": 1.738702081134577e-05,
+      "loss": 0.2021,
+      "step": 5081
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8685885199262032,
+      "learning_rate": 1.738591060148727e-05,
+      "loss": 0.2068,
+      "step": 5082
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.37429320393285,
+      "learning_rate": 1.738480019128553e-05,
+      "loss": 0.225,
+      "step": 5083
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.068890387233058,
+      "learning_rate": 1.7383689580770662e-05,
+      "loss": 0.2211,
+      "step": 5084
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0787377651617818,
+      "learning_rate": 1.73825787699728e-05,
+      "loss": 0.212,
+      "step": 5085
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9387024177777231,
+      "learning_rate": 1.738146775892207e-05,
+      "loss": 0.2207,
+      "step": 5086
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8724959929584131,
+      "learning_rate": 1.738035654764861e-05,
+      "loss": 0.2258,
+      "step": 5087
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0341884160681272,
+      "learning_rate": 1.7379245136182563e-05,
+      "loss": 0.1936,
+      "step": 5088
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9305601735593984,
+      "learning_rate": 1.7378133524554076e-05,
+      "loss": 0.2028,
+      "step": 5089
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9216192771869709,
+      "learning_rate": 1.73770217127933e-05,
+      "loss": 0.2057,
+      "step": 5090
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8275065447729806,
+      "learning_rate": 1.737590970093039e-05,
+      "loss": 0.2191,
+      "step": 5091
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8141071722166489,
+      "learning_rate": 1.737479748899552e-05,
+      "loss": 0.2119,
+      "step": 5092
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9472226924117711,
+      "learning_rate": 1.7373685077018844e-05,
+      "loss": 0.2188,
+      "step": 5093
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.047574692637895,
+      "learning_rate": 1.7372572465030545e-05,
+      "loss": 0.2104,
+      "step": 5094
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9203578125292801,
+      "learning_rate": 1.7371459653060806e-05,
+      "loss": 0.212,
+      "step": 5095
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8055320516963432,
+      "learning_rate": 1.7370346641139805e-05,
+      "loss": 0.2044,
+      "step": 5096
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.842063707877496,
+      "learning_rate": 1.7369233429297734e-05,
+      "loss": 0.2212,
+      "step": 5097
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.021204666571421,
+      "learning_rate": 1.7368120017564792e-05,
+      "loss": 0.2016,
+      "step": 5098
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.310478716770612,
+      "learning_rate": 1.7367006405971177e-05,
+      "loss": 0.2464,
+      "step": 5099
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4498907688492093,
+      "learning_rate": 1.7365892594547097e-05,
+      "loss": 0.202,
+      "step": 5100
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.196400429856508,
+      "learning_rate": 1.7364778583322765e-05,
+      "loss": 0.2247,
+      "step": 5101
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.923844439015725,
+      "learning_rate": 1.7363664372328398e-05,
+      "loss": 0.2332,
+      "step": 5102
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9109385255003308,
+      "learning_rate": 1.736254996159422e-05,
+      "loss": 0.207,
+      "step": 5103
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0074463288140154,
+      "learning_rate": 1.7361435351150456e-05,
+      "loss": 0.2194,
+      "step": 5104
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9833506313364749,
+      "learning_rate": 1.7360320541027342e-05,
+      "loss": 0.2277,
+      "step": 5105
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7589870505508941,
+      "learning_rate": 1.7359205531255123e-05,
+      "loss": 0.2,
+      "step": 5106
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9118865622399176,
+      "learning_rate": 1.735809032186403e-05,
+      "loss": 0.2307,
+      "step": 5107
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9536475657355715,
+      "learning_rate": 1.7356974912884327e-05,
+      "loss": 0.2089,
+      "step": 5108
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8118782600313672,
+      "learning_rate": 1.7355859304346262e-05,
+      "loss": 0.2264,
+      "step": 5109
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9823121461102297,
+      "learning_rate": 1.7354743496280103e-05,
+      "loss": 0.2034,
+      "step": 5110
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.903102188316509,
+      "learning_rate": 1.7353627488716106e-05,
+      "loss": 0.2166,
+      "step": 5111
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0695530528902415,
+      "learning_rate": 1.7352511281684548e-05,
+      "loss": 0.2328,
+      "step": 5112
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0200358766120496,
+      "learning_rate": 1.7351394875215707e-05,
+      "loss": 0.1933,
+      "step": 5113
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1098375204015731,
+      "learning_rate": 1.7350278269339867e-05,
+      "loss": 0.2238,
+      "step": 5114
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9829062711668459,
+      "learning_rate": 1.7349161464087312e-05,
+      "loss": 0.2239,
+      "step": 5115
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9539037518616844,
+      "learning_rate": 1.7348044459488334e-05,
+      "loss": 0.1879,
+      "step": 5116
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9057009263172614,
+      "learning_rate": 1.734692725557324e-05,
+      "loss": 0.2127,
+      "step": 5117
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9057268038085381,
+      "learning_rate": 1.734580985237233e-05,
+      "loss": 0.2221,
+      "step": 5118
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8635195105524243,
+      "learning_rate": 1.7344692249915907e-05,
+      "loss": 0.2007,
+      "step": 5119
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5332147512028922,
+      "learning_rate": 1.7343574448234294e-05,
+      "loss": 0.1949,
+      "step": 5120
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.92037678511703,
+      "learning_rate": 1.7342456447357813e-05,
+      "loss": 0.2245,
+      "step": 5121
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3309735135939107,
+      "learning_rate": 1.7341338247316785e-05,
+      "loss": 0.2805,
+      "step": 5122
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0574651253476115,
+      "learning_rate": 1.734021984814154e-05,
+      "loss": 0.2256,
+      "step": 5123
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0460865418770287,
+      "learning_rate": 1.7339101249862418e-05,
+      "loss": 0.1988,
+      "step": 5124
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1198594990219086,
+      "learning_rate": 1.7337982452509757e-05,
+      "loss": 0.2306,
+      "step": 5125
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8641486909806005,
+      "learning_rate": 1.7336863456113912e-05,
+      "loss": 0.1677,
+      "step": 5126
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9883756135069262,
+      "learning_rate": 1.7335744260705233e-05,
+      "loss": 0.1941,
+      "step": 5127
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.038637478516568,
+      "learning_rate": 1.733462486631407e-05,
+      "loss": 0.2101,
+      "step": 5128
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7833925304111683,
+      "learning_rate": 1.73335052729708e-05,
+      "loss": 0.1937,
+      "step": 5129
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1097990773913913,
+      "learning_rate": 1.733238548070578e-05,
+      "loss": 0.2329,
+      "step": 5130
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9441953595842283,
+      "learning_rate": 1.7331265489549392e-05,
+      "loss": 0.2259,
+      "step": 5131
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0517679100815562,
+      "learning_rate": 1.7330145299532014e-05,
+      "loss": 0.1908,
+      "step": 5132
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.865174450000809,
+      "learning_rate": 1.7329024910684033e-05,
+      "loss": 0.2338,
+      "step": 5133
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4011916507400322,
+      "learning_rate": 1.7327904323035833e-05,
+      "loss": 0.1923,
+      "step": 5134
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.733663183504703,
+      "learning_rate": 1.7326783536617817e-05,
+      "loss": 0.2025,
+      "step": 5135
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2850760579337643,
+      "learning_rate": 1.7325662551460382e-05,
+      "loss": 0.2243,
+      "step": 5136
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8465786613373185,
+      "learning_rate": 1.7324541367593938e-05,
+      "loss": 0.2175,
+      "step": 5137
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8575404475842421,
+      "learning_rate": 1.7323419985048895e-05,
+      "loss": 0.2039,
+      "step": 5138
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.410440921873463,
+      "learning_rate": 1.732229840385567e-05,
+      "loss": 0.1937,
+      "step": 5139
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7958882699551637,
+      "learning_rate": 1.732117662404469e-05,
+      "loss": 0.2388,
+      "step": 5140
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3727292538167888,
+      "learning_rate": 1.7320054645646376e-05,
+      "loss": 0.2214,
+      "step": 5141
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3762093404062987,
+      "learning_rate": 1.7318932468691172e-05,
+      "loss": 0.2023,
+      "step": 5142
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9535075619962287,
+      "learning_rate": 1.7317810093209507e-05,
+      "loss": 0.1909,
+      "step": 5143
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0330727306732685,
+      "learning_rate": 1.731668751923183e-05,
+      "loss": 0.2154,
+      "step": 5144
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8244231717387553,
+      "learning_rate": 1.7315564746788592e-05,
+      "loss": 0.2179,
+      "step": 5145
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9174505172121481,
+      "learning_rate": 1.731444177591025e-05,
+      "loss": 0.2175,
+      "step": 5146
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9969529480674448,
+      "learning_rate": 1.7313318606627258e-05,
+      "loss": 0.2279,
+      "step": 5147
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2100766546934472,
+      "learning_rate": 1.7312195238970088e-05,
+      "loss": 0.1959,
+      "step": 5148
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.412580848356862,
+      "learning_rate": 1.7311071672969206e-05,
+      "loss": 0.2057,
+      "step": 5149
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9606173224523024,
+      "learning_rate": 1.7309947908655096e-05,
+      "loss": 0.2338,
+      "step": 5150
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0224410815863882,
+      "learning_rate": 1.7308823946058237e-05,
+      "loss": 0.2161,
+      "step": 5151
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.16150172069741,
+      "learning_rate": 1.7307699785209108e-05,
+      "loss": 0.1782,
+      "step": 5152
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9738594797923933,
+      "learning_rate": 1.7306575426138213e-05,
+      "loss": 0.229,
+      "step": 5153
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9858951663213912,
+      "learning_rate": 1.730545086887605e-05,
+      "loss": 0.2131,
+      "step": 5154
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8323012000824099,
+      "learning_rate": 1.730432611345312e-05,
+      "loss": 0.2032,
+      "step": 5155
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8765788794254222,
+      "learning_rate": 1.730320115989993e-05,
+      "loss": 0.2255,
+      "step": 5156
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8654955411348836,
+      "learning_rate": 1.7302076008246993e-05,
+      "loss": 0.1953,
+      "step": 5157
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1368959010757456,
+      "learning_rate": 1.7300950658524836e-05,
+      "loss": 0.2067,
+      "step": 5158
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0193148376193706,
+      "learning_rate": 1.729982511076398e-05,
+      "loss": 0.2171,
+      "step": 5159
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.084242175875511,
+      "learning_rate": 1.7298699364994952e-05,
+      "loss": 0.2046,
+      "step": 5160
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9197173344240579,
+      "learning_rate": 1.7297573421248294e-05,
+      "loss": 0.2357,
+      "step": 5161
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1499815341386719,
+      "learning_rate": 1.729644727955454e-05,
+      "loss": 0.192,
+      "step": 5162
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2309580105851474,
+      "learning_rate": 1.7295320939944247e-05,
+      "loss": 0.2381,
+      "step": 5163
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1276580222219936,
+      "learning_rate": 1.729419440244796e-05,
+      "loss": 0.2211,
+      "step": 5164
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.034431671433726,
+      "learning_rate": 1.729306766709624e-05,
+      "loss": 0.202,
+      "step": 5165
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8705311980286389,
+      "learning_rate": 1.7291940733919645e-05,
+      "loss": 0.2149,
+      "step": 5166
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8392682674202115,
+      "learning_rate": 1.7290813602948748e-05,
+      "loss": 0.2088,
+      "step": 5167
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8684905385029499,
+      "learning_rate": 1.7289686274214116e-05,
+      "loss": 0.2112,
+      "step": 5168
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1264314185559987,
+      "learning_rate": 1.7288558747746335e-05,
+      "loss": 0.2012,
+      "step": 5169
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.978250214493829,
+      "learning_rate": 1.7287431023575988e-05,
+      "loss": 0.2155,
+      "step": 5170
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9360368977001948,
+      "learning_rate": 1.728630310173366e-05,
+      "loss": 0.2361,
+      "step": 5171
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8958072123216545,
+      "learning_rate": 1.7285174982249947e-05,
+      "loss": 0.2179,
+      "step": 5172
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.941632424104645,
+      "learning_rate": 1.7284046665155456e-05,
+      "loss": 0.2368,
+      "step": 5173
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8263123681636818,
+      "learning_rate": 1.7282918150480786e-05,
+      "loss": 0.204,
+      "step": 5174
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.914557956965905,
+      "learning_rate": 1.728178943825655e-05,
+      "loss": 0.1889,
+      "step": 5175
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9954073324336042,
+      "learning_rate": 1.7280660528513362e-05,
+      "loss": 0.2202,
+      "step": 5176
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7576450292859586,
+      "learning_rate": 1.727953142128185e-05,
+      "loss": 0.176,
+      "step": 5177
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8326638995643959,
+      "learning_rate": 1.727840211659263e-05,
+      "loss": 0.2173,
+      "step": 5178
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9662341961029348,
+      "learning_rate": 1.727727261447635e-05,
+      "loss": 0.208,
+      "step": 5179
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1586971973050766,
+      "learning_rate": 1.7276142914963635e-05,
+      "loss": 0.2183,
+      "step": 5180
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2003375851518439,
+      "learning_rate": 1.727501301808513e-05,
+      "loss": 0.2245,
+      "step": 5181
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9655590656767455,
+      "learning_rate": 1.7273882923871492e-05,
+      "loss": 0.1938,
+      "step": 5182
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3799769940471969,
+      "learning_rate": 1.7272752632353365e-05,
+      "loss": 0.223,
+      "step": 5183
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8849271828718694,
+      "learning_rate": 1.727162214356141e-05,
+      "loss": 0.2208,
+      "step": 5184
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8298189179543192,
+      "learning_rate": 1.72704914575263e-05,
+      "loss": 0.1903,
+      "step": 5185
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9025161258137357,
+      "learning_rate": 1.7269360574278694e-05,
+      "loss": 0.2043,
+      "step": 5186
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9525245007094877,
+      "learning_rate": 1.7268229493849273e-05,
+      "loss": 0.1838,
+      "step": 5187
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7302099244104475,
+      "learning_rate": 1.7267098216268715e-05,
+      "loss": 0.2172,
+      "step": 5188
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.539391710700766,
+      "learning_rate": 1.726596674156771e-05,
+      "loss": 0.249,
+      "step": 5189
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.207634354231271,
+      "learning_rate": 1.7264835069776945e-05,
+      "loss": 0.2087,
+      "step": 5190
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7730743543155159,
+      "learning_rate": 1.726370320092712e-05,
+      "loss": 0.2335,
+      "step": 5191
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1122396560369499,
+      "learning_rate": 1.7262571135048934e-05,
+      "loss": 0.2164,
+      "step": 5192
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3805955610238752,
+      "learning_rate": 1.7261438872173096e-05,
+      "loss": 0.2053,
+      "step": 5193
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0468968152705946,
+      "learning_rate": 1.7260306412330317e-05,
+      "loss": 0.207,
+      "step": 5194
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8783581393917873,
+      "learning_rate": 1.725917375555132e-05,
+      "loss": 0.2215,
+      "step": 5195
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3218838017025807,
+      "learning_rate": 1.7258040901866824e-05,
+      "loss": 0.209,
+      "step": 5196
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3075623769540963,
+      "learning_rate": 1.725690785130756e-05,
+      "loss": 0.2252,
+      "step": 5197
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2522311782873228,
+      "learning_rate": 1.7255774603904253e-05,
+      "loss": 0.1953,
+      "step": 5198
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9995002246188047,
+      "learning_rate": 1.7254641159687657e-05,
+      "loss": 0.2169,
+      "step": 5199
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8811425044279627,
+      "learning_rate": 1.725350751868851e-05,
+      "loss": 0.1869,
+      "step": 5200
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0446346367689991,
+      "learning_rate": 1.725237368093756e-05,
+      "loss": 0.1975,
+      "step": 5201
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8818232613937096,
+      "learning_rate": 1.7251239646465562e-05,
+      "loss": 0.2087,
+      "step": 5202
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.070969961207307,
+      "learning_rate": 1.7250105415303283e-05,
+      "loss": 0.2288,
+      "step": 5203
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8885154254095409,
+      "learning_rate": 1.7248970987481484e-05,
+      "loss": 0.2268,
+      "step": 5204
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0101696219659841,
+      "learning_rate": 1.7247836363030935e-05,
+      "loss": 0.1933,
+      "step": 5205
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0928642196121847,
+      "learning_rate": 1.724670154198242e-05,
+      "loss": 0.2081,
+      "step": 5206
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8613315400435708,
+      "learning_rate": 1.7245566524366713e-05,
+      "loss": 0.2135,
+      "step": 5207
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8440440917985735,
+      "learning_rate": 1.7244431310214604e-05,
+      "loss": 0.234,
+      "step": 5208
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2084843767372881,
+      "learning_rate": 1.724329589955689e-05,
+      "loss": 0.1991,
+      "step": 5209
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2607093167904153,
+      "learning_rate": 1.7242160292424362e-05,
+      "loss": 0.2204,
+      "step": 5210
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9823639926148594,
+      "learning_rate": 1.724102448884783e-05,
+      "loss": 0.2363,
+      "step": 5211
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9200310958626141,
+      "learning_rate": 1.7239888488858097e-05,
+      "loss": 0.2204,
+      "step": 5212
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.950300322829358,
+      "learning_rate": 1.723875229248598e-05,
+      "loss": 0.2235,
+      "step": 5213
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0896260359504228,
+      "learning_rate": 1.72376158997623e-05,
+      "loss": 0.2371,
+      "step": 5214
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9389995805140761,
+      "learning_rate": 1.7236479310717878e-05,
+      "loss": 0.2168,
+      "step": 5215
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.970088000932802,
+      "learning_rate": 1.723534252538355e-05,
+      "loss": 0.2205,
+      "step": 5216
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1210916950770902,
+      "learning_rate": 1.7234205543790143e-05,
+      "loss": 0.2066,
+      "step": 5217
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8368801598300282,
+      "learning_rate": 1.7233068365968505e-05,
+      "loss": 0.2272,
+      "step": 5218
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.007554450271422,
+      "learning_rate": 1.723193099194948e-05,
+      "loss": 0.2177,
+      "step": 5219
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.127711066292918,
+      "learning_rate": 1.7230793421763914e-05,
+      "loss": 0.1948,
+      "step": 5220
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4235497287302739,
+      "learning_rate": 1.722965565544267e-05,
+      "loss": 0.1932,
+      "step": 5221
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.263286240902716,
+      "learning_rate": 1.722851769301661e-05,
+      "loss": 0.2192,
+      "step": 5222
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8378216432441894,
+      "learning_rate": 1.7227379534516594e-05,
+      "loss": 0.1854,
+      "step": 5223
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.87419967006041,
+      "learning_rate": 1.7226241179973505e-05,
+      "loss": 0.2109,
+      "step": 5224
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7866001766190502,
+      "learning_rate": 1.7225102629418217e-05,
+      "loss": 0.2234,
+      "step": 5225
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.030357376854233,
+      "learning_rate": 1.7223963882881606e-05,
+      "loss": 0.1997,
+      "step": 5226
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9039436771158188,
+      "learning_rate": 1.722282494039457e-05,
+      "loss": 0.2306,
+      "step": 5227
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8311084196052098,
+      "learning_rate": 1.7221685801988003e-05,
+      "loss": 0.1968,
+      "step": 5228
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0155689933704966,
+      "learning_rate": 1.7220546467692797e-05,
+      "loss": 0.1919,
+      "step": 5229
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7621332743514103,
+      "learning_rate": 1.721940693753986e-05,
+      "loss": 0.1836,
+      "step": 5230
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9408075793675481,
+      "learning_rate": 1.7218267211560103e-05,
+      "loss": 0.2248,
+      "step": 5231
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.978169976254573,
+      "learning_rate": 1.7217127289784437e-05,
+      "loss": 0.2243,
+      "step": 5232
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9131227974824523,
+      "learning_rate": 1.7215987172243788e-05,
+      "loss": 0.2016,
+      "step": 5233
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0586303575143619,
+      "learning_rate": 1.721484685896908e-05,
+      "loss": 0.2054,
+      "step": 5234
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7416757017851802,
+      "learning_rate": 1.7213706349991243e-05,
+      "loss": 0.2054,
+      "step": 5235
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2013216628849288,
+      "learning_rate": 1.721256564534122e-05,
+      "loss": 0.2013,
+      "step": 5236
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4472292421550423,
+      "learning_rate": 1.7211424745049935e-05,
+      "loss": 0.2044,
+      "step": 5237
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1853024744443268,
+      "learning_rate": 1.7210283649148355e-05,
+      "loss": 0.2098,
+      "step": 5238
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.941182147562369,
+      "learning_rate": 1.720914235766742e-05,
+      "loss": 0.2263,
+      "step": 5239
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3061098361084038,
+      "learning_rate": 1.7208000870638094e-05,
+      "loss": 0.2085,
+      "step": 5240
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8182517884818539,
+      "learning_rate": 1.7206859188091334e-05,
+      "loss": 0.2026,
+      "step": 5241
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8225484731142085,
+      "learning_rate": 1.7205717310058115e-05,
+      "loss": 0.2006,
+      "step": 5242
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9473367325916228,
+      "learning_rate": 1.7204575236569403e-05,
+      "loss": 0.2237,
+      "step": 5243
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4132935672607803,
+      "learning_rate": 1.7203432967656185e-05,
+      "loss": 0.165,
+      "step": 5244
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9167655284706728,
+      "learning_rate": 1.7202290503349436e-05,
+      "loss": 0.1945,
+      "step": 5245
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9433089720769943,
+      "learning_rate": 1.7201147843680156e-05,
+      "loss": 0.2056,
+      "step": 5246
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5687627971666505,
+      "learning_rate": 1.7200004988679332e-05,
+      "loss": 0.2077,
+      "step": 5247
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0573084994768132,
+      "learning_rate": 1.7198861938377965e-05,
+      "loss": 0.2293,
+      "step": 5248
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7877683252059432,
+      "learning_rate": 1.719771869280706e-05,
+      "loss": 0.1943,
+      "step": 5249
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0970146564562546,
+      "learning_rate": 1.719657525199763e-05,
+      "loss": 0.2058,
+      "step": 5250
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9594541757438699,
+      "learning_rate": 1.7195431615980692e-05,
+      "loss": 0.2311,
+      "step": 5251
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9958431420570283,
+      "learning_rate": 1.719428778478726e-05,
+      "loss": 0.221,
+      "step": 5252
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1668962676306558,
+      "learning_rate": 1.719314375844837e-05,
+      "loss": 0.2285,
+      "step": 5253
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0370614060296535,
+      "learning_rate": 1.719199953699505e-05,
+      "loss": 0.2159,
+      "step": 5254
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8308609951209482,
+      "learning_rate": 1.7190855120458333e-05,
+      "loss": 0.2,
+      "step": 5255
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0656123715589458,
+      "learning_rate": 1.7189710508869266e-05,
+      "loss": 0.2023,
+      "step": 5256
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.883717672221715,
+      "learning_rate": 1.7188565702258893e-05,
+      "loss": 0.235,
+      "step": 5257
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8786535948719226,
+      "learning_rate": 1.7187420700658273e-05,
+      "loss": 0.1991,
+      "step": 5258
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0064167074518728,
+      "learning_rate": 1.718627550409846e-05,
+      "loss": 0.2273,
+      "step": 5259
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2174073349081103,
+      "learning_rate": 1.7185130112610518e-05,
+      "loss": 0.2033,
+      "step": 5260
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0550775089638038,
+      "learning_rate": 1.7183984526225517e-05,
+      "loss": 0.2257,
+      "step": 5261
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.843216437626411,
+      "learning_rate": 1.7182838744974525e-05,
+      "loss": 0.2314,
+      "step": 5262
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9792723396409431,
+      "learning_rate": 1.7181692768888632e-05,
+      "loss": 0.1869,
+      "step": 5263
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8365306985513885,
+      "learning_rate": 1.7180546597998913e-05,
+      "loss": 0.2035,
+      "step": 5264
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1671518444866484,
+      "learning_rate": 1.7179400232336462e-05,
+      "loss": 0.2152,
+      "step": 5265
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0216996902622235,
+      "learning_rate": 1.7178253671932378e-05,
+      "loss": 0.1933,
+      "step": 5266
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9408021541356948,
+      "learning_rate": 1.7177106916817754e-05,
+      "loss": 0.202,
+      "step": 5267
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.076468275502716,
+      "learning_rate": 1.7175959967023703e-05,
+      "loss": 0.2212,
+      "step": 5268
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8578660342363781,
+      "learning_rate": 1.717481282258133e-05,
+      "loss": 0.1909,
+      "step": 5269
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.398141662645059,
+      "learning_rate": 1.7173665483521757e-05,
+      "loss": 0.2328,
+      "step": 5270
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1515910090053505,
+      "learning_rate": 1.7172517949876098e-05,
+      "loss": 0.2135,
+      "step": 5271
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8627620038988313,
+      "learning_rate": 1.7171370221675486e-05,
+      "loss": 0.1891,
+      "step": 5272
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.196025450833706,
+      "learning_rate": 1.7170222298951053e-05,
+      "loss": 0.2216,
+      "step": 5273
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8852782147065706,
+      "learning_rate": 1.7169074181733934e-05,
+      "loss": 0.1954,
+      "step": 5274
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1495828840508941,
+      "learning_rate": 1.7167925870055273e-05,
+      "loss": 0.1853,
+      "step": 5275
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0454671880927513,
+      "learning_rate": 1.716677736394622e-05,
+      "loss": 0.206,
+      "step": 5276
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1007704616492229,
+      "learning_rate": 1.7165628663437923e-05,
+      "loss": 0.1847,
+      "step": 5277
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9550909862275008,
+      "learning_rate": 1.7164479768561546e-05,
+      "loss": 0.1888,
+      "step": 5278
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2416770226952552,
+      "learning_rate": 1.7163330679348248e-05,
+      "loss": 0.1962,
+      "step": 5279
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9840509198815576,
+      "learning_rate": 1.7162181395829204e-05,
+      "loss": 0.2141,
+      "step": 5280
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1730068768818405,
+      "learning_rate": 1.7161031918035584e-05,
+      "loss": 0.2345,
+      "step": 5281
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0294990181578525,
+      "learning_rate": 1.715988224599857e-05,
+      "loss": 0.2259,
+      "step": 5282
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8718280425482781,
+      "learning_rate": 1.7158732379749342e-05,
+      "loss": 0.2206,
+      "step": 5283
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7973950150881813,
+      "learning_rate": 1.71575823193191e-05,
+      "loss": 0.2011,
+      "step": 5284
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8958789502079094,
+      "learning_rate": 1.7156432064739024e-05,
+      "loss": 0.235,
+      "step": 5285
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9867998258952296,
+      "learning_rate": 1.7155281616040333e-05,
+      "loss": 0.2324,
+      "step": 5286
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9559868928584956,
+      "learning_rate": 1.715413097325422e-05,
+      "loss": 0.2009,
+      "step": 5287
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9716198690195488,
+      "learning_rate": 1.71529801364119e-05,
+      "loss": 0.1964,
+      "step": 5288
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1129669967352243,
+      "learning_rate": 1.715182910554459e-05,
+      "loss": 0.2745,
+      "step": 5289
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9123072409362243,
+      "learning_rate": 1.7150677880683515e-05,
+      "loss": 0.2095,
+      "step": 5290
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0062131124197695,
+      "learning_rate": 1.7149526461859897e-05,
+      "loss": 0.2296,
+      "step": 5291
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0380383036645535,
+      "learning_rate": 1.7148374849104965e-05,
+      "loss": 0.2074,
+      "step": 5292
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5639900438854735,
+      "learning_rate": 1.7147223042449968e-05,
+      "loss": 0.1836,
+      "step": 5293
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7888321865748255,
+      "learning_rate": 1.7146071041926138e-05,
+      "loss": 0.1985,
+      "step": 5294
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.433020224907867,
+      "learning_rate": 1.714491884756473e-05,
+      "loss": 0.252,
+      "step": 5295
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9519082354925571,
+      "learning_rate": 1.7143766459396993e-05,
+      "loss": 0.2113,
+      "step": 5296
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0719575839448903,
+      "learning_rate": 1.7142613877454186e-05,
+      "loss": 0.2011,
+      "step": 5297
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0089385606249057,
+      "learning_rate": 1.714146110176758e-05,
+      "loss": 0.2314,
+      "step": 5298
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2971638756917503,
+      "learning_rate": 1.714030813236843e-05,
+      "loss": 0.2287,
+      "step": 5299
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3336062167999805,
+      "learning_rate": 1.7139154969288026e-05,
+      "loss": 0.2174,
+      "step": 5300
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0890198977625634,
+      "learning_rate": 1.7138001612557636e-05,
+      "loss": 0.2095,
+      "step": 5301
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2127147798939326,
+      "learning_rate": 1.7136848062208552e-05,
+      "loss": 0.2234,
+      "step": 5302
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.4830948947069063,
+      "learning_rate": 1.7135694318272057e-05,
+      "loss": 0.2158,
+      "step": 5303
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1718966869611827,
+      "learning_rate": 1.7134540380779453e-05,
+      "loss": 0.2205,
+      "step": 5304
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.302483243747173,
+      "learning_rate": 1.713338624976204e-05,
+      "loss": 0.209,
+      "step": 5305
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9086727219113531,
+      "learning_rate": 1.713223192525112e-05,
+      "loss": 0.191,
+      "step": 5306
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.11751268899965,
+      "learning_rate": 1.7131077407278008e-05,
+      "loss": 0.2278,
+      "step": 5307
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0514724800840336,
+      "learning_rate": 1.7129922695874016e-05,
+      "loss": 0.2207,
+      "step": 5308
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9040266788733141,
+      "learning_rate": 1.712876779107047e-05,
+      "loss": 0.2056,
+      "step": 5309
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9642309087364644,
+      "learning_rate": 1.7127612692898695e-05,
+      "loss": 0.1947,
+      "step": 5310
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9401415784326851,
+      "learning_rate": 1.7126457401390023e-05,
+      "loss": 0.2014,
+      "step": 5311
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.027856827944775,
+      "learning_rate": 1.712530191657579e-05,
+      "loss": 0.2217,
+      "step": 5312
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.926767366522984,
+      "learning_rate": 1.712414623848734e-05,
+      "loss": 0.1958,
+      "step": 5313
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0775741073707663,
+      "learning_rate": 1.712299036715602e-05,
+      "loss": 0.234,
+      "step": 5314
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0525832503522692,
+      "learning_rate": 1.712183430261319e-05,
+      "loss": 0.1927,
+      "step": 5315
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8435504630743897,
+      "learning_rate": 1.71206780448902e-05,
+      "loss": 0.1961,
+      "step": 5316
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8021204498469657,
+      "learning_rate": 1.711952159401841e-05,
+      "loss": 0.1926,
+      "step": 5317
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0147291827406735,
+      "learning_rate": 1.71183649500292e-05,
+      "loss": 0.1947,
+      "step": 5318
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0304586584190933,
+      "learning_rate": 1.711720811295394e-05,
+      "loss": 0.2174,
+      "step": 5319
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9570126657180933,
+      "learning_rate": 1.7116051082824003e-05,
+      "loss": 0.2029,
+      "step": 5320
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9625640745359515,
+      "learning_rate": 1.711489385967078e-05,
+      "loss": 0.1857,
+      "step": 5321
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9279472961440479,
+      "learning_rate": 1.7113736443525662e-05,
+      "loss": 0.2104,
+      "step": 5322
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9756088976113342,
+      "learning_rate": 1.7112578834420036e-05,
+      "loss": 0.2136,
+      "step": 5323
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0709776914176408,
+      "learning_rate": 1.7111421032385313e-05,
+      "loss": 0.2043,
+      "step": 5324
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.225056753679554,
+      "learning_rate": 1.711026303745289e-05,
+      "loss": 0.222,
+      "step": 5325
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9022739276514431,
+      "learning_rate": 1.710910484965418e-05,
+      "loss": 0.2045,
+      "step": 5326
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7728402557707814,
+      "learning_rate": 1.71079464690206e-05,
+      "loss": 0.2314,
+      "step": 5327
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8080405879850907,
+      "learning_rate": 1.7106787895583573e-05,
+      "loss": 0.232,
+      "step": 5328
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7797907753458592,
+      "learning_rate": 1.710562912937452e-05,
+      "loss": 0.2095,
+      "step": 5329
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5049046662803585,
+      "learning_rate": 1.710447017042488e-05,
+      "loss": 0.2189,
+      "step": 5330
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8657038274319842,
+      "learning_rate": 1.710331101876608e-05,
+      "loss": 0.2169,
+      "step": 5331
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9621461414357568,
+      "learning_rate": 1.7102151674429567e-05,
+      "loss": 0.211,
+      "step": 5332
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3494794196811704,
+      "learning_rate": 1.7100992137446792e-05,
+      "loss": 0.1997,
+      "step": 5333
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0959570082489183,
+      "learning_rate": 1.7099832407849203e-05,
+      "loss": 0.2073,
+      "step": 5334
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0047881510760754,
+      "learning_rate": 1.709867248566826e-05,
+      "loss": 0.2302,
+      "step": 5335
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.292243114304556,
+      "learning_rate": 1.7097512370935422e-05,
+      "loss": 0.256,
+      "step": 5336
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8680614701378598,
+      "learning_rate": 1.7096352063682163e-05,
+      "loss": 0.2063,
+      "step": 5337
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4413192657041078,
+      "learning_rate": 1.709519156393995e-05,
+      "loss": 0.2173,
+      "step": 5338
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9953817989447833,
+      "learning_rate": 1.709403087174027e-05,
+      "loss": 0.2115,
+      "step": 5339
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0349009846998212,
+      "learning_rate": 1.70928699871146e-05,
+      "loss": 0.2315,
+      "step": 5340
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.977550830932778,
+      "learning_rate": 1.709170891009443e-05,
+      "loss": 0.1951,
+      "step": 5341
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9936609207937576,
+      "learning_rate": 1.7090547640711256e-05,
+      "loss": 0.2179,
+      "step": 5342
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0861275402229364,
+      "learning_rate": 1.7089386178996576e-05,
+      "loss": 0.2168,
+      "step": 5343
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7740154237094764,
+      "learning_rate": 1.70882245249819e-05,
+      "loss": 0.2033,
+      "step": 5344
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.137396135861429,
+      "learning_rate": 1.7087062678698726e-05,
+      "loss": 0.2073,
+      "step": 5345
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9734578826182253,
+      "learning_rate": 1.7085900640178582e-05,
+      "loss": 0.2231,
+      "step": 5346
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.023946483249804,
+      "learning_rate": 1.7084738409452982e-05,
+      "loss": 0.2204,
+      "step": 5347
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9838807350978188,
+      "learning_rate": 1.7083575986553448e-05,
+      "loss": 0.2084,
+      "step": 5348
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.038334702873617,
+      "learning_rate": 1.708241337151152e-05,
+      "loss": 0.2181,
+      "step": 5349
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9705124443322362,
+      "learning_rate": 1.708125056435873e-05,
+      "loss": 0.2053,
+      "step": 5350
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0641275795654905,
+      "learning_rate": 1.7080087565126613e-05,
+      "loss": 0.2133,
+      "step": 5351
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8789516998052838,
+      "learning_rate": 1.707892437384673e-05,
+      "loss": 0.2174,
+      "step": 5352
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9768442549284431,
+      "learning_rate": 1.7077760990550617e-05,
+      "loss": 0.2115,
+      "step": 5353
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0576896748411604,
+      "learning_rate": 1.7076597415269836e-05,
+      "loss": 0.2261,
+      "step": 5354
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.04442745089693,
+      "learning_rate": 1.7075433648035952e-05,
+      "loss": 0.2209,
+      "step": 5355
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9242450796659158,
+      "learning_rate": 1.707426968888053e-05,
+      "loss": 0.2035,
+      "step": 5356
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8041864219716768,
+      "learning_rate": 1.7073105537835145e-05,
+      "loss": 0.2223,
+      "step": 5357
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0618635761951816,
+      "learning_rate": 1.7071941194931372e-05,
+      "loss": 0.1791,
+      "step": 5358
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8489978326658181,
+      "learning_rate": 1.7070776660200797e-05,
+      "loss": 0.2106,
+      "step": 5359
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0697877918430914,
+      "learning_rate": 1.7069611933675006e-05,
+      "loss": 0.2105,
+      "step": 5360
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.06294677154257,
+      "learning_rate": 1.7068447015385587e-05,
+      "loss": 0.236,
+      "step": 5361
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2730267833270381,
+      "learning_rate": 1.706728190536415e-05,
+      "loss": 0.2269,
+      "step": 5362
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1527764705868002,
+      "learning_rate": 1.7066116603642285e-05,
+      "loss": 0.2155,
+      "step": 5363
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9380764122116448,
+      "learning_rate": 1.706495111025161e-05,
+      "loss": 0.2209,
+      "step": 5364
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.816402075469522,
+      "learning_rate": 1.706378542522374e-05,
+      "loss": 0.2285,
+      "step": 5365
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.097092753302864,
+      "learning_rate": 1.706261954859029e-05,
+      "loss": 0.223,
+      "step": 5366
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2072634018517636,
+      "learning_rate": 1.7061453480382885e-05,
+      "loss": 0.2088,
+      "step": 5367
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1616017699973211,
+      "learning_rate": 1.7060287220633158e-05,
+      "loss": 0.1903,
+      "step": 5368
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0131673326230828,
+      "learning_rate": 1.7059120769372737e-05,
+      "loss": 0.204,
+      "step": 5369
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0226302459685408,
+      "learning_rate": 1.7057954126633268e-05,
+      "loss": 0.1971,
+      "step": 5370
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1991289041476931,
+      "learning_rate": 1.7056787292446396e-05,
+      "loss": 0.2156,
+      "step": 5371
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9740162925818302,
+      "learning_rate": 1.7055620266843776e-05,
+      "loss": 0.2157,
+      "step": 5372
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3082793580892642,
+      "learning_rate": 1.705445304985705e-05,
+      "loss": 0.1786,
+      "step": 5373
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9009398830787593,
+      "learning_rate": 1.7053285641517886e-05,
+      "loss": 0.1898,
+      "step": 5374
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.333310758399028,
+      "learning_rate": 1.7052118041857954e-05,
+      "loss": 0.2291,
+      "step": 5375
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8219067225406175,
+      "learning_rate": 1.7050950250908923e-05,
+      "loss": 0.2086,
+      "step": 5376
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0504821356893528,
+      "learning_rate": 1.7049782268702464e-05,
+      "loss": 0.2185,
+      "step": 5377
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5062762094220585,
+      "learning_rate": 1.7048614095270264e-05,
+      "loss": 0.2142,
+      "step": 5378
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1517461171713312,
+      "learning_rate": 1.704744573064401e-05,
+      "loss": 0.2252,
+      "step": 5379
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0666768851989106,
+      "learning_rate": 1.704627717485539e-05,
+      "loss": 0.2154,
+      "step": 5380
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9800546805869343,
+      "learning_rate": 1.7045108427936104e-05,
+      "loss": 0.2039,
+      "step": 5381
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.285558315197764,
+      "learning_rate": 1.7043939489917858e-05,
+      "loss": 0.217,
+      "step": 5382
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9672045321469797,
+      "learning_rate": 1.7042770360832353e-05,
+      "loss": 0.2049,
+      "step": 5383
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3148918349554373,
+      "learning_rate": 1.7041601040711303e-05,
+      "loss": 0.2531,
+      "step": 5384
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9287482626051508,
+      "learning_rate": 1.7040431529586427e-05,
+      "loss": 0.1955,
+      "step": 5385
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0439263751970198,
+      "learning_rate": 1.7039261827489452e-05,
+      "loss": 0.2057,
+      "step": 5386
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0966870805891058,
+      "learning_rate": 1.7038091934452098e-05,
+      "loss": 0.2182,
+      "step": 5387
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0429284976399116,
+      "learning_rate": 1.7036921850506104e-05,
+      "loss": 0.2444,
+      "step": 5388
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9220008646181936,
+      "learning_rate": 1.7035751575683208e-05,
+      "loss": 0.2106,
+      "step": 5389
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2418202057946766,
+      "learning_rate": 1.7034581110015156e-05,
+      "loss": 0.2111,
+      "step": 5390
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9337888990910338,
+      "learning_rate": 1.7033410453533687e-05,
+      "loss": 0.2124,
+      "step": 5391
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2735895782979454,
+      "learning_rate": 1.7032239606270567e-05,
+      "loss": 0.1977,
+      "step": 5392
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8751710465726789,
+      "learning_rate": 1.7031068568257548e-05,
+      "loss": 0.1978,
+      "step": 5393
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8455923178009026,
+      "learning_rate": 1.7029897339526404e-05,
+      "loss": 0.2147,
+      "step": 5394
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.826334770638359,
+      "learning_rate": 1.702872592010889e-05,
+      "loss": 0.2223,
+      "step": 5395
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0356782106426614,
+      "learning_rate": 1.702755431003679e-05,
+      "loss": 0.216,
+      "step": 5396
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0489926641495397,
+      "learning_rate": 1.7026382509341885e-05,
+      "loss": 0.2214,
+      "step": 5397
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9830309908645947,
+      "learning_rate": 1.7025210518055954e-05,
+      "loss": 0.2126,
+      "step": 5398
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8860550610226454,
+      "learning_rate": 1.7024038336210794e-05,
+      "loss": 0.1982,
+      "step": 5399
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0477678845976919,
+      "learning_rate": 1.7022865963838195e-05,
+      "loss": 0.1764,
+      "step": 5400
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9756090995833591,
+      "learning_rate": 1.7021693400969962e-05,
+      "loss": 0.2074,
+      "step": 5401
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0863202596740222,
+      "learning_rate": 1.7020520647637894e-05,
+      "loss": 0.2004,
+      "step": 5402
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6093505669717603,
+      "learning_rate": 1.701934770387381e-05,
+      "loss": 0.2044,
+      "step": 5403
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1912297537469665,
+      "learning_rate": 1.7018174569709523e-05,
+      "loss": 0.214,
+      "step": 5404
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0552055408422767,
+      "learning_rate": 1.7017001245176857e-05,
+      "loss": 0.2056,
+      "step": 5405
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9568556007681108,
+      "learning_rate": 1.7015827730307637e-05,
+      "loss": 0.2009,
+      "step": 5406
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0428933384888612,
+      "learning_rate": 1.701465402513369e-05,
+      "loss": 0.1951,
+      "step": 5407
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0154243359082187,
+      "learning_rate": 1.7013480129686857e-05,
+      "loss": 0.2184,
+      "step": 5408
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1022197933165605,
+      "learning_rate": 1.701230604399898e-05,
+      "loss": 0.197,
+      "step": 5409
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0560707879020101,
+      "learning_rate": 1.7011131768101906e-05,
+      "loss": 0.2162,
+      "step": 5410
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2588170514748378,
+      "learning_rate": 1.7009957302027484e-05,
+      "loss": 0.2096,
+      "step": 5411
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0979029434665841,
+      "learning_rate": 1.7008782645807578e-05,
+      "loss": 0.2112,
+      "step": 5412
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9524557231896512,
+      "learning_rate": 1.7007607799474045e-05,
+      "loss": 0.2133,
+      "step": 5413
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1895621186357845,
+      "learning_rate": 1.7006432763058753e-05,
+      "loss": 0.2,
+      "step": 5414
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.190375294075288,
+      "learning_rate": 1.7005257536593577e-05,
+      "loss": 0.2149,
+      "step": 5415
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9149426868161031,
+      "learning_rate": 1.7004082120110396e-05,
+      "loss": 0.2138,
+      "step": 5416
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.885553028586094,
+      "learning_rate": 1.7002906513641094e-05,
+      "loss": 0.2094,
+      "step": 5417
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1377344849438038,
+      "learning_rate": 1.7001730717217554e-05,
+      "loss": 0.2188,
+      "step": 5418
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0142042105802078,
+      "learning_rate": 1.700055473087167e-05,
+      "loss": 0.2463,
+      "step": 5419
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9577845607969336,
+      "learning_rate": 1.699937855463535e-05,
+      "loss": 0.2003,
+      "step": 5420
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7987401617507317,
+      "learning_rate": 1.699820218854049e-05,
+      "loss": 0.196,
+      "step": 5421
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9466080898054804,
+      "learning_rate": 1.6997025632618996e-05,
+      "loss": 0.2141,
+      "step": 5422
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9591314964338855,
+      "learning_rate": 1.6995848886902794e-05,
+      "loss": 0.2099,
+      "step": 5423
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0143917225888177,
+      "learning_rate": 1.699467195142379e-05,
+      "loss": 0.2189,
+      "step": 5424
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1287154569913822,
+      "learning_rate": 1.6993494826213917e-05,
+      "loss": 0.2119,
+      "step": 5425
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9108997932383712,
+      "learning_rate": 1.6992317511305103e-05,
+      "loss": 0.2298,
+      "step": 5426
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0312293374855317,
+      "learning_rate": 1.6991140006729277e-05,
+      "loss": 0.2044,
+      "step": 5427
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2243294060862493,
+      "learning_rate": 1.6989962312518384e-05,
+      "loss": 0.2113,
+      "step": 5428
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3494977850309595,
+      "learning_rate": 1.698878442870437e-05,
+      "loss": 0.2011,
+      "step": 5429
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6045698540700097,
+      "learning_rate": 1.6987606355319184e-05,
+      "loss": 0.2081,
+      "step": 5430
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9385287988505853,
+      "learning_rate": 1.698642809239478e-05,
+      "loss": 0.189,
+      "step": 5431
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.064849914586746,
+      "learning_rate": 1.698524963996312e-05,
+      "loss": 0.238,
+      "step": 5432
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9401539962006684,
+      "learning_rate": 1.698407099805617e-05,
+      "loss": 0.1883,
+      "step": 5433
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2014313002921773,
+      "learning_rate": 1.69828921667059e-05,
+      "loss": 0.1861,
+      "step": 5434
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0598940147011735,
+      "learning_rate": 1.6981713145944284e-05,
+      "loss": 0.2166,
+      "step": 5435
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1006504747535542,
+      "learning_rate": 1.6980533935803306e-05,
+      "loss": 0.2054,
+      "step": 5436
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8626033245045626,
+      "learning_rate": 1.6979354536314946e-05,
+      "loss": 0.2281,
+      "step": 5437
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8596006588035439,
+      "learning_rate": 1.6978174947511206e-05,
+      "loss": 0.2064,
+      "step": 5438
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4517008204949062,
+      "learning_rate": 1.6976995169424072e-05,
+      "loss": 0.2104,
+      "step": 5439
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8330081037185146,
+      "learning_rate": 1.6975815202085556e-05,
+      "loss": 0.191,
+      "step": 5440
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2590492794567847,
+      "learning_rate": 1.6974635045527652e-05,
+      "loss": 0.2134,
+      "step": 5441
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.002732843409242,
+      "learning_rate": 1.6973454699782382e-05,
+      "loss": 0.2081,
+      "step": 5442
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7695496820937182,
+      "learning_rate": 1.697227416488176e-05,
+      "loss": 0.224,
+      "step": 5443
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7855777505010354,
+      "learning_rate": 1.6971093440857808e-05,
+      "loss": 0.1951,
+      "step": 5444
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9924584904825682,
+      "learning_rate": 1.6969912527742547e-05,
+      "loss": 0.2015,
+      "step": 5445
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1286658064526591,
+      "learning_rate": 1.696873142556802e-05,
+      "loss": 0.2269,
+      "step": 5446
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1905921151254606,
+      "learning_rate": 1.6967550134366256e-05,
+      "loss": 0.2266,
+      "step": 5447
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9938986283838219,
+      "learning_rate": 1.6966368654169305e-05,
+      "loss": 0.2068,
+      "step": 5448
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9412002207458522,
+      "learning_rate": 1.696518698500921e-05,
+      "loss": 0.1996,
+      "step": 5449
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8128848611817997,
+      "learning_rate": 1.696400512691802e-05,
+      "loss": 0.2133,
+      "step": 5450
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7118661473447155,
+      "learning_rate": 1.6962823079927803e-05,
+      "loss": 0.2049,
+      "step": 5451
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9659168027327373,
+      "learning_rate": 1.696164084407062e-05,
+      "loss": 0.1916,
+      "step": 5452
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.07456804180084,
+      "learning_rate": 1.6960458419378528e-05,
+      "loss": 0.2201,
+      "step": 5453
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1286968393726051,
+      "learning_rate": 1.695927580588361e-05,
+      "loss": 0.2281,
+      "step": 5454
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9546368608833864,
+      "learning_rate": 1.6958093003617942e-05,
+      "loss": 0.2243,
+      "step": 5455
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8286795365597308,
+      "learning_rate": 1.6956910012613612e-05,
+      "loss": 0.211,
+      "step": 5456
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1624253033639174,
+      "learning_rate": 1.6955726832902705e-05,
+      "loss": 0.1959,
+      "step": 5457
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8258952586684022,
+      "learning_rate": 1.6954543464517313e-05,
+      "loss": 0.2002,
+      "step": 5458
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9861277252982845,
+      "learning_rate": 1.6953359907489538e-05,
+      "loss": 0.1835,
+      "step": 5459
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.869053520172057,
+      "learning_rate": 1.695217616185148e-05,
+      "loss": 0.2033,
+      "step": 5460
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.466873269888362,
+      "learning_rate": 1.6950992227635252e-05,
+      "loss": 0.214,
+      "step": 5461
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9216917177291614,
+      "learning_rate": 1.6949808104872965e-05,
+      "loss": 0.208,
+      "step": 5462
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1146734573826524,
+      "learning_rate": 1.6948623793596744e-05,
+      "loss": 0.2057,
+      "step": 5463
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9795441070290263,
+      "learning_rate": 1.694743929383871e-05,
+      "loss": 0.2031,
+      "step": 5464
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9643666484708967,
+      "learning_rate": 1.6946254605630995e-05,
+      "loss": 0.2045,
+      "step": 5465
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9659569712091012,
+      "learning_rate": 1.6945069729005726e-05,
+      "loss": 0.1829,
+      "step": 5466
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8573246043762399,
+      "learning_rate": 1.6943884663995055e-05,
+      "loss": 0.2012,
+      "step": 5467
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1065850167937314,
+      "learning_rate": 1.6942699410631114e-05,
+      "loss": 0.1991,
+      "step": 5468
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7600803308859483,
+      "learning_rate": 1.6941513968946063e-05,
+      "loss": 0.2072,
+      "step": 5469
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9756031387165406,
+      "learning_rate": 1.6940328338972053e-05,
+      "loss": 0.2149,
+      "step": 5470
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7830774903264235,
+      "learning_rate": 1.6939142520741243e-05,
+      "loss": 0.1987,
+      "step": 5471
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.047546367636695,
+      "learning_rate": 1.6937956514285797e-05,
+      "loss": 0.2123,
+      "step": 5472
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2211101494147778,
+      "learning_rate": 1.6936770319637896e-05,
+      "loss": 0.1906,
+      "step": 5473
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9542329622300272,
+      "learning_rate": 1.6935583936829706e-05,
+      "loss": 0.2119,
+      "step": 5474
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.99799970671572,
+      "learning_rate": 1.693439736589341e-05,
+      "loss": 0.2135,
+      "step": 5475
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2219047082819476,
+      "learning_rate": 1.693321060686119e-05,
+      "loss": 0.2051,
+      "step": 5476
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0655680430773613,
+      "learning_rate": 1.6932023659765248e-05,
+      "loss": 0.2079,
+      "step": 5477
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8407756911500547,
+      "learning_rate": 1.6930836524637766e-05,
+      "loss": 0.2046,
+      "step": 5478
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0019352759531008,
+      "learning_rate": 1.6929649201510953e-05,
+      "loss": 0.2061,
+      "step": 5479
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0336266052542888,
+      "learning_rate": 1.692846169041702e-05,
+      "loss": 0.2073,
+      "step": 5480
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1592968084273145,
+      "learning_rate": 1.6927273991388164e-05,
+      "loss": 0.2137,
+      "step": 5481
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0472714014176197,
+      "learning_rate": 1.6926086104456613e-05,
+      "loss": 0.1932,
+      "step": 5482
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8884190104412825,
+      "learning_rate": 1.6924898029654585e-05,
+      "loss": 0.1962,
+      "step": 5483
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.440890522067889,
+      "learning_rate": 1.692370976701431e-05,
+      "loss": 0.2311,
+      "step": 5484
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9975412180086698,
+      "learning_rate": 1.6922521316568014e-05,
+      "loss": 0.1937,
+      "step": 5485
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9162614283067528,
+      "learning_rate": 1.6921332678347936e-05,
+      "loss": 0.2477,
+      "step": 5486
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.885433998276788,
+      "learning_rate": 1.6920143852386316e-05,
+      "loss": 0.1839,
+      "step": 5487
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9263433712536561,
+      "learning_rate": 1.6918954838715408e-05,
+      "loss": 0.2161,
+      "step": 5488
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.053505012946194,
+      "learning_rate": 1.6917765637367455e-05,
+      "loss": 0.2018,
+      "step": 5489
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3169890952877474,
+      "learning_rate": 1.691657624837472e-05,
+      "loss": 0.2152,
+      "step": 5490
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7815313992450333,
+      "learning_rate": 1.6915386671769463e-05,
+      "loss": 0.1914,
+      "step": 5491
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.070490158653429,
+      "learning_rate": 1.6914196907583952e-05,
+      "loss": 0.231,
+      "step": 5492
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8897247721716537,
+      "learning_rate": 1.6913006955850462e-05,
+      "loss": 0.2147,
+      "step": 5493
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2170042600343527,
+      "learning_rate": 1.6911816816601266e-05,
+      "loss": 0.2403,
+      "step": 5494
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.202057324771667,
+      "learning_rate": 1.691062648986865e-05,
+      "loss": 0.21,
+      "step": 5495
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0040241770319058,
+      "learning_rate": 1.69094359756849e-05,
+      "loss": 0.2084,
+      "step": 5496
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8620668771810918,
+      "learning_rate": 1.6908245274082306e-05,
+      "loss": 0.2105,
+      "step": 5497
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2031020687777687,
+      "learning_rate": 1.690705438509317e-05,
+      "loss": 0.2091,
+      "step": 5498
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2348820456964273,
+      "learning_rate": 1.6905863308749793e-05,
+      "loss": 0.1794,
+      "step": 5499
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0425667322586452,
+      "learning_rate": 1.6904672045084485e-05,
+      "loss": 0.2207,
+      "step": 5500
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0625047037067803,
+      "learning_rate": 1.6903480594129557e-05,
+      "loss": 0.1962,
+      "step": 5501
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1736637841443045,
+      "learning_rate": 1.6902288955917328e-05,
+      "loss": 0.2155,
+      "step": 5502
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.101980767830775,
+      "learning_rate": 1.690109713048012e-05,
+      "loss": 0.2163,
+      "step": 5503
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.937151736434946,
+      "learning_rate": 1.6899905117850266e-05,
+      "loss": 0.1959,
+      "step": 5504
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8079687624922969,
+      "learning_rate": 1.6898712918060093e-05,
+      "loss": 0.1853,
+      "step": 5505
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8451325110889567,
+      "learning_rate": 1.6897520531141944e-05,
+      "loss": 0.1939,
+      "step": 5506
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9312220398678531,
+      "learning_rate": 1.6896327957128162e-05,
+      "loss": 0.2163,
+      "step": 5507
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9727108910026903,
+      "learning_rate": 1.689513519605109e-05,
+      "loss": 0.1939,
+      "step": 5508
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7698826201117085,
+      "learning_rate": 1.689394224794309e-05,
+      "loss": 0.2068,
+      "step": 5509
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7688375248720871,
+      "learning_rate": 1.689274911283652e-05,
+      "loss": 0.196,
+      "step": 5510
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0006142145056547,
+      "learning_rate": 1.6891555790763735e-05,
+      "loss": 0.2128,
+      "step": 5511
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9641337809883561,
+      "learning_rate": 1.6890362281757117e-05,
+      "loss": 0.2134,
+      "step": 5512
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8013917063334224,
+      "learning_rate": 1.6889168585849027e-05,
+      "loss": 0.2173,
+      "step": 5513
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8104111308085543,
+      "learning_rate": 1.688797470307185e-05,
+      "loss": 0.1928,
+      "step": 5514
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8837868269865045,
+      "learning_rate": 1.6886780633457975e-05,
+      "loss": 0.2148,
+      "step": 5515
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9583015710960666,
+      "learning_rate": 1.688558637703978e-05,
+      "loss": 0.1941,
+      "step": 5516
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8237325286578135,
+      "learning_rate": 1.688439193384967e-05,
+      "loss": 0.2069,
+      "step": 5517
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.285974324365605,
+      "learning_rate": 1.688319730392004e-05,
+      "loss": 0.2282,
+      "step": 5518
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1747843825560265,
+      "learning_rate": 1.6882002487283293e-05,
+      "loss": 0.2305,
+      "step": 5519
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.165339327087881,
+      "learning_rate": 1.688080748397184e-05,
+      "loss": 0.2201,
+      "step": 5520
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0824986772041116,
+      "learning_rate": 1.6879612294018092e-05,
+      "loss": 0.2254,
+      "step": 5521
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5266727806258882,
+      "learning_rate": 1.687841691745448e-05,
+      "loss": 0.2098,
+      "step": 5522
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0719337185200175,
+      "learning_rate": 1.6877221354313413e-05,
+      "loss": 0.1957,
+      "step": 5523
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9076648477056829,
+      "learning_rate": 1.6876025604627335e-05,
+      "loss": 0.1939,
+      "step": 5524
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7657803013254962,
+      "learning_rate": 1.6874829668428667e-05,
+      "loss": 0.2222,
+      "step": 5525
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3297842157111655,
+      "learning_rate": 1.6873633545749858e-05,
+      "loss": 0.2118,
+      "step": 5526
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8951088432162874,
+      "learning_rate": 1.6872437236623352e-05,
+      "loss": 0.2007,
+      "step": 5527
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2104995388774757,
+      "learning_rate": 1.68712407410816e-05,
+      "loss": 0.1915,
+      "step": 5528
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1709013879049428,
+      "learning_rate": 1.6870044059157052e-05,
+      "loss": 0.2232,
+      "step": 5529
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.236545303278789,
+      "learning_rate": 1.686884719088217e-05,
+      "loss": 0.2066,
+      "step": 5530
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0617137566736374,
+      "learning_rate": 1.6867650136289425e-05,
+      "loss": 0.2365,
+      "step": 5531
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3100906770551048,
+      "learning_rate": 1.686645289541128e-05,
+      "loss": 0.2079,
+      "step": 5532
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9947827430940916,
+      "learning_rate": 1.686525546828021e-05,
+      "loss": 0.1977,
+      "step": 5533
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6173318374449783,
+      "learning_rate": 1.6864057854928696e-05,
+      "loss": 0.216,
+      "step": 5534
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8658319621042003,
+      "learning_rate": 1.686286005538923e-05,
+      "loss": 0.1902,
+      "step": 5535
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9118008242410937,
+      "learning_rate": 1.6861662069694292e-05,
+      "loss": 0.2222,
+      "step": 5536
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.287722536221362,
+      "learning_rate": 1.686046389787639e-05,
+      "loss": 0.209,
+      "step": 5537
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6445443639482815,
+      "learning_rate": 1.6859265539968014e-05,
+      "loss": 0.2136,
+      "step": 5538
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2204081239684939,
+      "learning_rate": 1.6858066996001673e-05,
+      "loss": 0.2172,
+      "step": 5539
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0577270671622425,
+      "learning_rate": 1.6856868266009874e-05,
+      "loss": 0.1842,
+      "step": 5540
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.24709712484204,
+      "learning_rate": 1.6855669350025138e-05,
+      "loss": 0.2265,
+      "step": 5541
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0953891982266504,
+      "learning_rate": 1.6854470248079983e-05,
+      "loss": 0.2333,
+      "step": 5542
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9000461617024391,
+      "learning_rate": 1.685327096020694e-05,
+      "loss": 0.221,
+      "step": 5543
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9692217209212479,
+      "learning_rate": 1.685207148643853e-05,
+      "loss": 0.1821,
+      "step": 5544
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9836872882390156,
+      "learning_rate": 1.6850871826807297e-05,
+      "loss": 0.2122,
+      "step": 5545
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1278818571254685,
+      "learning_rate": 1.6849671981345775e-05,
+      "loss": 0.1943,
+      "step": 5546
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1074602416317132,
+      "learning_rate": 1.6848471950086517e-05,
+      "loss": 0.224,
+      "step": 5547
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.005589817604628,
+      "learning_rate": 1.684727173306207e-05,
+      "loss": 0.2247,
+      "step": 5548
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9020066072063216,
+      "learning_rate": 1.684607133030499e-05,
+      "loss": 0.2225,
+      "step": 5549
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7223123330717156,
+      "learning_rate": 1.684487074184784e-05,
+      "loss": 0.1739,
+      "step": 5550
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7908254662329712,
+      "learning_rate": 1.6843669967723183e-05,
+      "loss": 0.2086,
+      "step": 5551
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9050190208821911,
+      "learning_rate": 1.6842469007963592e-05,
+      "loss": 0.1978,
+      "step": 5552
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1064952641542272,
+      "learning_rate": 1.6841267862601644e-05,
+      "loss": 0.199,
+      "step": 5553
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3459381134914081,
+      "learning_rate": 1.6840066531669915e-05,
+      "loss": 0.2242,
+      "step": 5554
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8388784954524903,
+      "learning_rate": 1.6838865015200995e-05,
+      "loss": 0.21,
+      "step": 5555
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9451542750441775,
+      "learning_rate": 1.683766331322748e-05,
+      "loss": 0.2214,
+      "step": 5556
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8709892311875053,
+      "learning_rate": 1.683646142578196e-05,
+      "loss": 0.2011,
+      "step": 5557
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9928480478934592,
+      "learning_rate": 1.6835259352897035e-05,
+      "loss": 0.2044,
+      "step": 5558
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9502941509055366,
+      "learning_rate": 1.6834057094605314e-05,
+      "loss": 0.2281,
+      "step": 5559
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0910863752995716,
+      "learning_rate": 1.683285465093941e-05,
+      "loss": 0.1846,
+      "step": 5560
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5288457696683095,
+      "learning_rate": 1.683165202193194e-05,
+      "loss": 0.1984,
+      "step": 5561
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8747400245648658,
+      "learning_rate": 1.683044920761552e-05,
+      "loss": 0.2311,
+      "step": 5562
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0847339869897128,
+      "learning_rate": 1.682924620802278e-05,
+      "loss": 0.1875,
+      "step": 5563
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.024929122243032,
+      "learning_rate": 1.682804302318635e-05,
+      "loss": 0.1864,
+      "step": 5564
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6855645827959116,
+      "learning_rate": 1.6826839653138872e-05,
+      "loss": 0.1951,
+      "step": 5565
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9463126278497938,
+      "learning_rate": 1.6825636097912976e-05,
+      "loss": 0.2164,
+      "step": 5566
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7997011520507737,
+      "learning_rate": 1.682443235754132e-05,
+      "loss": 0.1905,
+      "step": 5567
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3388113262501908,
+      "learning_rate": 1.682322843205655e-05,
+      "loss": 0.2205,
+      "step": 5568
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9707915728471659,
+      "learning_rate": 1.6822024321491323e-05,
+      "loss": 0.2247,
+      "step": 5569
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8366771355218784,
+      "learning_rate": 1.6820820025878298e-05,
+      "loss": 0.196,
+      "step": 5570
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1658522644977178,
+      "learning_rate": 1.6819615545250146e-05,
+      "loss": 0.2072,
+      "step": 5571
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0906354617526341,
+      "learning_rate": 1.681841087963954e-05,
+      "loss": 0.2709,
+      "step": 5572
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7832304082448883,
+      "learning_rate": 1.681720602907915e-05,
+      "loss": 0.1903,
+      "step": 5573
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9031727797125559,
+      "learning_rate": 1.6816000993601668e-05,
+      "loss": 0.1975,
+      "step": 5574
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8269271957462649,
+      "learning_rate": 1.6814795773239766e-05,
+      "loss": 0.1893,
+      "step": 5575
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.409115031127353,
+      "learning_rate": 1.681359036802615e-05,
+      "loss": 0.2361,
+      "step": 5576
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0714923444580238,
+      "learning_rate": 1.681238477799351e-05,
+      "loss": 0.2124,
+      "step": 5577
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4448729012174302,
+      "learning_rate": 1.6811179003174546e-05,
+      "loss": 0.2037,
+      "step": 5578
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9301528109817786,
+      "learning_rate": 1.6809973043601962e-05,
+      "loss": 0.1971,
+      "step": 5579
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1091802621706934,
+      "learning_rate": 1.680876689930848e-05,
+      "loss": 0.2285,
+      "step": 5580
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7086801941850318,
+      "learning_rate": 1.680756057032681e-05,
+      "loss": 0.1964,
+      "step": 5581
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0176127550275247,
+      "learning_rate": 1.680635405668968e-05,
+      "loss": 0.2213,
+      "step": 5582
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4627825962559355,
+      "learning_rate": 1.6805147358429806e-05,
+      "loss": 0.1863,
+      "step": 5583
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7935095943088147,
+      "learning_rate": 1.6803940475579926e-05,
+      "loss": 0.1861,
+      "step": 5584
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2494953657874224,
+      "learning_rate": 1.680273340817278e-05,
+      "loss": 0.2112,
+      "step": 5585
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.154399835221216,
+      "learning_rate": 1.68015261562411e-05,
+      "loss": 0.2179,
+      "step": 5586
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9786789665166755,
+      "learning_rate": 1.6800318719817647e-05,
+      "loss": 0.2403,
+      "step": 5587
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7028235365975033,
+      "learning_rate": 1.679911109893516e-05,
+      "loss": 0.1846,
+      "step": 5588
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1778933740553295,
+      "learning_rate": 1.67979032936264e-05,
+      "loss": 0.2242,
+      "step": 5589
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.398224553047776,
+      "learning_rate": 1.679669530392413e-05,
+      "loss": 0.2028,
+      "step": 5590
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3966578774852827,
+      "learning_rate": 1.679548712986111e-05,
+      "loss": 0.2242,
+      "step": 5591
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9520399610008522,
+      "learning_rate": 1.6794278771470127e-05,
+      "loss": 0.1944,
+      "step": 5592
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8628424629593858,
+      "learning_rate": 1.6793070228783946e-05,
+      "loss": 0.1972,
+      "step": 5593
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8946564228844422,
+      "learning_rate": 1.679186150183535e-05,
+      "loss": 0.2161,
+      "step": 5594
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.660295366849027,
+      "learning_rate": 1.6790652590657125e-05,
+      "loss": 0.2274,
+      "step": 5595
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1028149430210776,
+      "learning_rate": 1.678944349528207e-05,
+      "loss": 0.2041,
+      "step": 5596
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9057071652801462,
+      "learning_rate": 1.6788234215742974e-05,
+      "loss": 0.224,
+      "step": 5597
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8486970283199143,
+      "learning_rate": 1.6787024752072642e-05,
+      "loss": 0.2148,
+      "step": 5598
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1979543365735743,
+      "learning_rate": 1.678581510430388e-05,
+      "loss": 0.2268,
+      "step": 5599
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0996428666993319,
+      "learning_rate": 1.6784605272469502e-05,
+      "loss": 0.2101,
+      "step": 5600
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1156512400914844,
+      "learning_rate": 1.6783395256602318e-05,
+      "loss": 0.2249,
+      "step": 5601
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0102994579614044,
+      "learning_rate": 1.6782185056735157e-05,
+      "loss": 0.2209,
+      "step": 5602
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1022746944347066,
+      "learning_rate": 1.6780974672900845e-05,
+      "loss": 0.1736,
+      "step": 5603
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8559003285821831,
+      "learning_rate": 1.677976410513221e-05,
+      "loss": 0.2073,
+      "step": 5604
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.316426081816249,
+      "learning_rate": 1.6778553353462092e-05,
+      "loss": 0.2115,
+      "step": 5605
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3715287544859591,
+      "learning_rate": 1.677734241792333e-05,
+      "loss": 0.2162,
+      "step": 5606
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8883864765063251,
+      "learning_rate": 1.677613129854877e-05,
+      "loss": 0.203,
+      "step": 5607
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8644309330656957,
+      "learning_rate": 1.6774919995371272e-05,
+      "loss": 0.1833,
+      "step": 5608
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8490153399647028,
+      "learning_rate": 1.6773708508423683e-05,
+      "loss": 0.2016,
+      "step": 5609
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9282538393726152,
+      "learning_rate": 1.6772496837738866e-05,
+      "loss": 0.2365,
+      "step": 5610
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0473637461227412,
+      "learning_rate": 1.6771284983349693e-05,
+      "loss": 0.2206,
+      "step": 5611
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0559113462080079,
+      "learning_rate": 1.6770072945289034e-05,
+      "loss": 0.2596,
+      "step": 5612
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7196028723600773,
+      "learning_rate": 1.676886072358976e-05,
+      "loss": 0.2199,
+      "step": 5613
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.091513271080775,
+      "learning_rate": 1.676764831828476e-05,
+      "loss": 0.2019,
+      "step": 5614
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9225670399473616,
+      "learning_rate": 1.6766435729406913e-05,
+      "loss": 0.2347,
+      "step": 5615
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.99503266690628,
+      "learning_rate": 1.6765222956989117e-05,
+      "loss": 0.2061,
+      "step": 5616
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.878108183400604,
+      "learning_rate": 1.6764010001064268e-05,
+      "loss": 0.1947,
+      "step": 5617
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.881611187195488,
+      "learning_rate": 1.6762796861665262e-05,
+      "loss": 0.2059,
+      "step": 5618
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8924377250902428,
+      "learning_rate": 1.6761583538825013e-05,
+      "loss": 0.2228,
+      "step": 5619
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8815121230345735,
+      "learning_rate": 1.6760370032576424e-05,
+      "loss": 0.2137,
+      "step": 5620
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8511565240034018,
+      "learning_rate": 1.6759156342952422e-05,
+      "loss": 0.2009,
+      "step": 5621
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1423106856019447,
+      "learning_rate": 1.6757942469985917e-05,
+      "loss": 0.2177,
+      "step": 5622
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.370018827440741,
+      "learning_rate": 1.6756728413709843e-05,
+      "loss": 0.2134,
+      "step": 5623
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3487268044836134,
+      "learning_rate": 1.6755514174157127e-05,
+      "loss": 0.2188,
+      "step": 5624
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.090300940977795,
+      "learning_rate": 1.675429975136071e-05,
+      "loss": 0.2023,
+      "step": 5625
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9414580517820581,
+      "learning_rate": 1.675308514535353e-05,
+      "loss": 0.2219,
+      "step": 5626
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8682789778569923,
+      "learning_rate": 1.6751870356168534e-05,
+      "loss": 0.1966,
+      "step": 5627
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8623789355882996,
+      "learning_rate": 1.6750655383838674e-05,
+      "loss": 0.2179,
+      "step": 5628
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9338422907144169,
+      "learning_rate": 1.6749440228396903e-05,
+      "loss": 0.2091,
+      "step": 5629
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1181363031884917,
+      "learning_rate": 1.6748224889876188e-05,
+      "loss": 0.2387,
+      "step": 5630
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.623438042143102,
+      "learning_rate": 1.674700936830949e-05,
+      "loss": 0.2126,
+      "step": 5631
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.910801099431657,
+      "learning_rate": 1.6745793663729785e-05,
+      "loss": 0.238,
+      "step": 5632
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9199691919612899,
+      "learning_rate": 1.674457777617004e-05,
+      "loss": 0.1937,
+      "step": 5633
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9524353356354959,
+      "learning_rate": 1.6743361705663246e-05,
+      "loss": 0.2159,
+      "step": 5634
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8468532219803875,
+      "learning_rate": 1.6742145452242383e-05,
+      "loss": 0.2097,
+      "step": 5635
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8347163633932896,
+      "learning_rate": 1.6740929015940442e-05,
+      "loss": 0.209,
+      "step": 5636
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0916796899646724,
+      "learning_rate": 1.6739712396790424e-05,
+      "loss": 0.2056,
+      "step": 5637
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9631455506402163,
+      "learning_rate": 1.673849559482533e-05,
+      "loss": 0.1959,
+      "step": 5638
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9242865697867108,
+      "learning_rate": 1.6737278610078153e-05,
+      "loss": 0.1952,
+      "step": 5639
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.074821995609312,
+      "learning_rate": 1.6736061442581922e-05,
+      "loss": 0.2034,
+      "step": 5640
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9970315145305882,
+      "learning_rate": 1.673484409236964e-05,
+      "loss": 0.1959,
+      "step": 5641
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8036034716113344,
+      "learning_rate": 1.673362655947433e-05,
+      "loss": 0.2198,
+      "step": 5642
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9439196709366106,
+      "learning_rate": 1.673240884392902e-05,
+      "loss": 0.1962,
+      "step": 5643
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.881953076253939,
+      "learning_rate": 1.6731190945766742e-05,
+      "loss": 0.2086,
+      "step": 5644
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0536395032248527,
+      "learning_rate": 1.672997286502053e-05,
+      "loss": 0.2236,
+      "step": 5645
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3389673197902134,
+      "learning_rate": 1.672875460172342e-05,
+      "loss": 0.2045,
+      "step": 5646
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9475155150541045,
+      "learning_rate": 1.6727536155908466e-05,
+      "loss": 0.2071,
+      "step": 5647
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9946118684590498,
+      "learning_rate": 1.672631752760871e-05,
+      "loss": 0.2139,
+      "step": 5648
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1607991460269982,
+      "learning_rate": 1.6725098716857212e-05,
+      "loss": 0.2155,
+      "step": 5649
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8941861618108317,
+      "learning_rate": 1.672387972368703e-05,
+      "loss": 0.2113,
+      "step": 5650
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.086806213120535,
+      "learning_rate": 1.6722660548131235e-05,
+      "loss": 0.2116,
+      "step": 5651
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7956815228190831,
+      "learning_rate": 1.6721441190222893e-05,
+      "loss": 0.1886,
+      "step": 5652
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0590701485177547,
+      "learning_rate": 1.6720221649995076e-05,
+      "loss": 0.2069,
+      "step": 5653
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7979807415368692,
+      "learning_rate": 1.6719001927480867e-05,
+      "loss": 0.2057,
+      "step": 5654
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0005418175143537,
+      "learning_rate": 1.6717782022713353e-05,
+      "loss": 0.2211,
+      "step": 5655
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9628323917247162,
+      "learning_rate": 1.671656193572562e-05,
+      "loss": 0.2178,
+      "step": 5656
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5734424180049371,
+      "learning_rate": 1.671534166655077e-05,
+      "loss": 0.1901,
+      "step": 5657
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.55494116484686,
+      "learning_rate": 1.6714121215221894e-05,
+      "loss": 0.229,
+      "step": 5658
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9325780349684512,
+      "learning_rate": 1.67129005817721e-05,
+      "loss": 0.2171,
+      "step": 5659
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2796034374258602,
+      "learning_rate": 1.67116797662345e-05,
+      "loss": 0.2123,
+      "step": 5660
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0749231903118135,
+      "learning_rate": 1.6710458768642207e-05,
+      "loss": 0.2215,
+      "step": 5661
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1093769231076407,
+      "learning_rate": 1.670923758902834e-05,
+      "loss": 0.2224,
+      "step": 5662
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3186374701895731,
+      "learning_rate": 1.6708016227426026e-05,
+      "loss": 0.1889,
+      "step": 5663
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.067109564591355,
+      "learning_rate": 1.6706794683868392e-05,
+      "loss": 0.2334,
+      "step": 5664
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2771290744730772,
+      "learning_rate": 1.6705572958388576e-05,
+      "loss": 0.2373,
+      "step": 5665
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.994192818216985,
+      "learning_rate": 1.6704351051019713e-05,
+      "loss": 0.2172,
+      "step": 5666
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8792652678481304,
+      "learning_rate": 1.6703128961794947e-05,
+      "loss": 0.1987,
+      "step": 5667
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3800397889807778,
+      "learning_rate": 1.670190669074743e-05,
+      "loss": 0.2263,
+      "step": 5668
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.840686344309017,
+      "learning_rate": 1.670068423791032e-05,
+      "loss": 0.1914,
+      "step": 5669
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8809991020409779,
+      "learning_rate": 1.6699461603316765e-05,
+      "loss": 0.1922,
+      "step": 5670
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1474302610352636,
+      "learning_rate": 1.669823878699994e-05,
+      "loss": 0.2132,
+      "step": 5671
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9949574631829177,
+      "learning_rate": 1.669701578899301e-05,
+      "loss": 0.2054,
+      "step": 5672
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.378673043957621,
+      "learning_rate": 1.6695792609329148e-05,
+      "loss": 0.1907,
+      "step": 5673
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9551872363706768,
+      "learning_rate": 1.669456924804153e-05,
+      "loss": 0.1831,
+      "step": 5674
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.921282116685391,
+      "learning_rate": 1.6693345705163343e-05,
+      "loss": 0.1964,
+      "step": 5675
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.415101477622893,
+      "learning_rate": 1.669212198072778e-05,
+      "loss": 0.2267,
+      "step": 5676
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2433915586599227,
+      "learning_rate": 1.669089807476803e-05,
+      "loss": 0.2407,
+      "step": 5677
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1426110298544896,
+      "learning_rate": 1.668967398731729e-05,
+      "loss": 0.2146,
+      "step": 5678
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8677298754657188,
+      "learning_rate": 1.6688449718408763e-05,
+      "loss": 0.2143,
+      "step": 5679
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.845779757673886,
+      "learning_rate": 1.6687225268075665e-05,
+      "loss": 0.2377,
+      "step": 5680
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7281009544445249,
+      "learning_rate": 1.6686000636351197e-05,
+      "loss": 0.1995,
+      "step": 5681
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8926328124730796,
+      "learning_rate": 1.6684775823268592e-05,
+      "loss": 0.1922,
+      "step": 5682
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.923995341210268,
+      "learning_rate": 1.668355082886106e-05,
+      "loss": 0.1937,
+      "step": 5683
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.391058018212231,
+      "learning_rate": 1.6682325653161833e-05,
+      "loss": 0.2005,
+      "step": 5684
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9626507421652647,
+      "learning_rate": 1.668110029620415e-05,
+      "loss": 0.2245,
+      "step": 5685
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0274059668926714,
+      "learning_rate": 1.6679874758021238e-05,
+      "loss": 0.184,
+      "step": 5686
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0835223734839512,
+      "learning_rate": 1.6678649038646353e-05,
+      "loss": 0.2178,
+      "step": 5687
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0583879888167973,
+      "learning_rate": 1.667742313811273e-05,
+      "loss": 0.2051,
+      "step": 5688
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.826642326401598,
+      "learning_rate": 1.667619705645363e-05,
+      "loss": 0.2006,
+      "step": 5689
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8127180877352143,
+      "learning_rate": 1.667497079370231e-05,
+      "loss": 0.2119,
+      "step": 5690
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4911742939755106,
+      "learning_rate": 1.6673744349892027e-05,
+      "loss": 0.2051,
+      "step": 5691
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8507874783495339,
+      "learning_rate": 1.6672517725056052e-05,
+      "loss": 0.1919,
+      "step": 5692
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2710015881210284,
+      "learning_rate": 1.6671290919227656e-05,
+      "loss": 0.2389,
+      "step": 5693
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9906515157794544,
+      "learning_rate": 1.667006393244012e-05,
+      "loss": 0.2384,
+      "step": 5694
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9793754164793437,
+      "learning_rate": 1.666883676472672e-05,
+      "loss": 0.2152,
+      "step": 5695
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8065468973363352,
+      "learning_rate": 1.666760941612075e-05,
+      "loss": 0.2056,
+      "step": 5696
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9731836300048301,
+      "learning_rate": 1.666638188665549e-05,
+      "loss": 0.2213,
+      "step": 5697
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.226280971426743,
+      "learning_rate": 1.6665154176364252e-05,
+      "loss": 0.2196,
+      "step": 5698
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9443958794853763,
+      "learning_rate": 1.666392628528033e-05,
+      "loss": 0.1855,
+      "step": 5699
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1573762089894577,
+      "learning_rate": 1.666269821343703e-05,
+      "loss": 0.201,
+      "step": 5700
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9768856011391813,
+      "learning_rate": 1.666146996086766e-05,
+      "loss": 0.2271,
+      "step": 5701
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8118714858889647,
+      "learning_rate": 1.6660241527605546e-05,
+      "loss": 0.1818,
+      "step": 5702
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8388414321070643,
+      "learning_rate": 1.6659012913684005e-05,
+      "loss": 0.2118,
+      "step": 5703
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8937549781822656,
+      "learning_rate": 1.665778411913636e-05,
+      "loss": 0.2302,
+      "step": 5704
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0990565637143503,
+      "learning_rate": 1.6656555143995946e-05,
+      "loss": 0.211,
+      "step": 5705
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0010147981590252,
+      "learning_rate": 1.66553259882961e-05,
+      "loss": 0.2307,
+      "step": 5706
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8837306750998283,
+      "learning_rate": 1.6654096652070157e-05,
+      "loss": 0.1806,
+      "step": 5707
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8684906796406004,
+      "learning_rate": 1.665286713535147e-05,
+      "loss": 0.2114,
+      "step": 5708
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9092706245114834,
+      "learning_rate": 1.6651637438173382e-05,
+      "loss": 0.1945,
+      "step": 5709
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0686983902228964,
+      "learning_rate": 1.665040756056926e-05,
+      "loss": 0.23,
+      "step": 5710
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1221867067651907,
+      "learning_rate": 1.6649177502572447e-05,
+      "loss": 0.2209,
+      "step": 5711
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0704699851234865,
+      "learning_rate": 1.6647947264216328e-05,
+      "loss": 0.1899,
+      "step": 5712
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.991710562833805,
+      "learning_rate": 1.664671684553426e-05,
+      "loss": 0.1844,
+      "step": 5713
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8965807307602568,
+      "learning_rate": 1.6645486246559622e-05,
+      "loss": 0.1953,
+      "step": 5714
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0461114963512261,
+      "learning_rate": 1.6644255467325793e-05,
+      "loss": 0.2339,
+      "step": 5715
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8634193165469928,
+      "learning_rate": 1.6643024507866158e-05,
+      "loss": 0.1947,
+      "step": 5716
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8478908905893784,
+      "learning_rate": 1.664179336821411e-05,
+      "loss": 0.2118,
+      "step": 5717
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9551220717526766,
+      "learning_rate": 1.6640562048403044e-05,
+      "loss": 0.198,
+      "step": 5718
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1053131099679867,
+      "learning_rate": 1.6639330548466356e-05,
+      "loss": 0.2285,
+      "step": 5719
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8436838705449303,
+      "learning_rate": 1.6638098868437453e-05,
+      "loss": 0.1869,
+      "step": 5720
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.41614560693824,
+      "learning_rate": 1.663686700834974e-05,
+      "loss": 0.2166,
+      "step": 5721
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.451927439591581,
+      "learning_rate": 1.6635634968236637e-05,
+      "loss": 0.2365,
+      "step": 5722
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.993119411046374,
+      "learning_rate": 1.663440274813156e-05,
+      "loss": 0.1974,
+      "step": 5723
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.097312693258662,
+      "learning_rate": 1.6633170348067935e-05,
+      "loss": 0.1944,
+      "step": 5724
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1593390713884788,
+      "learning_rate": 1.663193776807919e-05,
+      "loss": 0.2147,
+      "step": 5725
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.159035057231313,
+      "learning_rate": 1.6630705008198757e-05,
+      "loss": 0.2045,
+      "step": 5726
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9332523049846795,
+      "learning_rate": 1.6629472068460077e-05,
+      "loss": 0.1989,
+      "step": 5727
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2142552198529883,
+      "learning_rate": 1.662823894889659e-05,
+      "loss": 0.2172,
+      "step": 5728
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.49516565416373,
+      "learning_rate": 1.6627005649541746e-05,
+      "loss": 0.2127,
+      "step": 5729
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4027560988903203,
+      "learning_rate": 1.6625772170429005e-05,
+      "loss": 0.1918,
+      "step": 5730
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.202650465639793,
+      "learning_rate": 1.6624538511591817e-05,
+      "loss": 0.2219,
+      "step": 5731
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2748809823636216,
+      "learning_rate": 1.6623304673063647e-05,
+      "loss": 0.2256,
+      "step": 5732
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0268563268112219,
+      "learning_rate": 1.6622070654877966e-05,
+      "loss": 0.2014,
+      "step": 5733
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1245407843284232,
+      "learning_rate": 1.6620836457068242e-05,
+      "loss": 0.2189,
+      "step": 5734
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9168931745402805,
+      "learning_rate": 1.6619602079667956e-05,
+      "loss": 0.2112,
+      "step": 5735
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8845526859215986,
+      "learning_rate": 1.661836752271059e-05,
+      "loss": 0.2323,
+      "step": 5736
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0410697350692784,
+      "learning_rate": 1.6617132786229634e-05,
+      "loss": 0.1992,
+      "step": 5737
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0815951639272545,
+      "learning_rate": 1.661589787025857e-05,
+      "loss": 0.2061,
+      "step": 5738
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.133386163528876,
+      "learning_rate": 1.6614662774830908e-05,
+      "loss": 0.2232,
+      "step": 5739
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3203570744289945,
+      "learning_rate": 1.6613427499980143e-05,
+      "loss": 0.1992,
+      "step": 5740
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8843085131207743,
+      "learning_rate": 1.6612192045739787e-05,
+      "loss": 0.2039,
+      "step": 5741
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8025452416194425,
+      "learning_rate": 1.6610956412143346e-05,
+      "loss": 0.1985,
+      "step": 5742
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.489947977642656,
+      "learning_rate": 1.6609720599224337e-05,
+      "loss": 0.2007,
+      "step": 5743
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9234053910717891,
+      "learning_rate": 1.6608484607016283e-05,
+      "loss": 0.2143,
+      "step": 5744
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9372514810803722,
+      "learning_rate": 1.6607248435552714e-05,
+      "loss": 0.2024,
+      "step": 5745
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.107377154971928,
+      "learning_rate": 1.6606012084867158e-05,
+      "loss": 0.2152,
+      "step": 5746
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.922792485971133,
+      "learning_rate": 1.660477555499315e-05,
+      "loss": 0.2083,
+      "step": 5747
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9716374962394986,
+      "learning_rate": 1.660353884596423e-05,
+      "loss": 0.1889,
+      "step": 5748
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.03480208202746,
+      "learning_rate": 1.6602301957813945e-05,
+      "loss": 0.23,
+      "step": 5749
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2601246611472028,
+      "learning_rate": 1.6601064890575852e-05,
+      "loss": 0.2107,
+      "step": 5750
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9725700733302011,
+      "learning_rate": 1.6599827644283496e-05,
+      "loss": 0.1852,
+      "step": 5751
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8676443914009231,
+      "learning_rate": 1.6598590218970448e-05,
+      "loss": 0.2128,
+      "step": 5752
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9559585801515682,
+      "learning_rate": 1.6597352614670265e-05,
+      "loss": 0.1967,
+      "step": 5753
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9523708018419746,
+      "learning_rate": 1.6596114831416516e-05,
+      "loss": 0.1927,
+      "step": 5754
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5704200689560264,
+      "learning_rate": 1.6594876869242785e-05,
+      "loss": 0.2105,
+      "step": 5755
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2809466782681533,
+      "learning_rate": 1.659363872818264e-05,
+      "loss": 0.1994,
+      "step": 5756
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7852549758135308,
+      "learning_rate": 1.6592400408269678e-05,
+      "loss": 0.1898,
+      "step": 5757
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.468007426452819,
+      "learning_rate": 1.659116190953748e-05,
+      "loss": 0.1984,
+      "step": 5758
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0512057713606473,
+      "learning_rate": 1.6589923232019646e-05,
+      "loss": 0.236,
+      "step": 5759
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8335140202496528,
+      "learning_rate": 1.6588684375749767e-05,
+      "loss": 0.2051,
+      "step": 5760
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8435524615441553,
+      "learning_rate": 1.6587445340761456e-05,
+      "loss": 0.1985,
+      "step": 5761
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.136818437779603,
+      "learning_rate": 1.658620612708832e-05,
+      "loss": 0.203,
+      "step": 5762
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5529887857017122,
+      "learning_rate": 1.6584966734763966e-05,
+      "loss": 0.1925,
+      "step": 5763
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0151767977894686,
+      "learning_rate": 1.6583727163822016e-05,
+      "loss": 0.2128,
+      "step": 5764
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8870410825868733,
+      "learning_rate": 1.6582487414296097e-05,
+      "loss": 0.2046,
+      "step": 5765
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3136083487300818,
+      "learning_rate": 1.6581247486219837e-05,
+      "loss": 0.2081,
+      "step": 5766
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.075196880227764,
+      "learning_rate": 1.6580007379626868e-05,
+      "loss": 0.2084,
+      "step": 5767
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.928292447081866,
+      "learning_rate": 1.6578767094550826e-05,
+      "loss": 0.2121,
+      "step": 5768
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1057184105594184,
+      "learning_rate": 1.6577526631025352e-05,
+      "loss": 0.2214,
+      "step": 5769
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.084890084007609,
+      "learning_rate": 1.65762859890841e-05,
+      "loss": 0.2286,
+      "step": 5770
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0065522632622652,
+      "learning_rate": 1.6575045168760716e-05,
+      "loss": 0.1941,
+      "step": 5771
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0942944625729116,
+      "learning_rate": 1.6573804170088866e-05,
+      "loss": 0.2055,
+      "step": 5772
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2163497184581087,
+      "learning_rate": 1.65725629931022e-05,
+      "loss": 0.2141,
+      "step": 5773
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2837384138894168,
+      "learning_rate": 1.65713216378344e-05,
+      "loss": 0.2065,
+      "step": 5774
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1628408812538042,
+      "learning_rate": 1.6570080104319122e-05,
+      "loss": 0.2181,
+      "step": 5775
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0705948505531206,
+      "learning_rate": 1.656883839259005e-05,
+      "loss": 0.2109,
+      "step": 5776
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6840407787429668,
+      "learning_rate": 1.656759650268087e-05,
+      "loss": 0.2008,
+      "step": 5777
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2944722190769227,
+      "learning_rate": 1.6566354434625262e-05,
+      "loss": 0.2134,
+      "step": 5778
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7357663578551865,
+      "learning_rate": 1.656511218845692e-05,
+      "loss": 0.1994,
+      "step": 5779
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3509628119207966,
+      "learning_rate": 1.6563869764209538e-05,
+      "loss": 0.1859,
+      "step": 5780
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.133614725165837,
+      "learning_rate": 1.656262716191682e-05,
+      "loss": 0.2061,
+      "step": 5781
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0005438686506853,
+      "learning_rate": 1.6561384381612463e-05,
+      "loss": 0.1961,
+      "step": 5782
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2009133423409168,
+      "learning_rate": 1.656014142333019e-05,
+      "loss": 0.2023,
+      "step": 5783
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1461548411705642,
+      "learning_rate": 1.6558898287103708e-05,
+      "loss": 0.2175,
+      "step": 5784
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.102960060107423,
+      "learning_rate": 1.6557654972966743e-05,
+      "loss": 0.2301,
+      "step": 5785
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8208184710818014,
+      "learning_rate": 1.6556411480953012e-05,
+      "loss": 0.2057,
+      "step": 5786
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9678370278571667,
+      "learning_rate": 1.655516781109625e-05,
+      "loss": 0.2112,
+      "step": 5787
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7878003614739206,
+      "learning_rate": 1.6553923963430193e-05,
+      "loss": 0.1782,
+      "step": 5788
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0807032380326957,
+      "learning_rate": 1.655267993798858e-05,
+      "loss": 0.1907,
+      "step": 5789
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0574042335885612,
+      "learning_rate": 1.655143573480515e-05,
+      "loss": 0.2407,
+      "step": 5790
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9989306136577616,
+      "learning_rate": 1.6550191353913657e-05,
+      "loss": 0.212,
+      "step": 5791
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9239187621127245,
+      "learning_rate": 1.654894679534785e-05,
+      "loss": 0.1757,
+      "step": 5792
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9822578991598856,
+      "learning_rate": 1.6547702059141497e-05,
+      "loss": 0.1973,
+      "step": 5793
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0885857577961602,
+      "learning_rate": 1.6546457145328354e-05,
+      "loss": 0.2046,
+      "step": 5794
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1196107951161263,
+      "learning_rate": 1.654521205394219e-05,
+      "loss": 0.1986,
+      "step": 5795
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0144618703541814,
+      "learning_rate": 1.654396678501678e-05,
+      "loss": 0.1774,
+      "step": 5796
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9975357074513678,
+      "learning_rate": 1.65427213385859e-05,
+      "loss": 0.1916,
+      "step": 5797
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.027741848788133,
+      "learning_rate": 1.6541475714683337e-05,
+      "loss": 0.197,
+      "step": 5798
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7796913200996538,
+      "learning_rate": 1.6540229913342875e-05,
+      "loss": 0.1971,
+      "step": 5799
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7289629178138828,
+      "learning_rate": 1.6538983934598304e-05,
+      "loss": 0.201,
+      "step": 5800
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2443187942554474,
+      "learning_rate": 1.653773777848343e-05,
+      "loss": 0.2208,
+      "step": 5801
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.013422787334009,
+      "learning_rate": 1.6536491445032044e-05,
+      "loss": 0.1913,
+      "step": 5802
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.358326711355817,
+      "learning_rate": 1.6535244934277962e-05,
+      "loss": 0.2023,
+      "step": 5803
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5021160626864412,
+      "learning_rate": 1.653399824625499e-05,
+      "loss": 0.2047,
+      "step": 5804
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8694947879982603,
+      "learning_rate": 1.653275138099695e-05,
+      "loss": 0.24,
+      "step": 5805
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0982253887611164,
+      "learning_rate": 1.6531504338537653e-05,
+      "loss": 0.2039,
+      "step": 5806
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9651108431730049,
+      "learning_rate": 1.6530257118910936e-05,
+      "loss": 0.1938,
+      "step": 5807
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2459276680152085,
+      "learning_rate": 1.6529009722150626e-05,
+      "loss": 0.2199,
+      "step": 5808
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0178850451064618,
+      "learning_rate": 1.652776214829056e-05,
+      "loss": 0.2025,
+      "step": 5809
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1226569227170007,
+      "learning_rate": 1.6526514397364575e-05,
+      "loss": 0.2068,
+      "step": 5810
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5468864133936497,
+      "learning_rate": 1.652526646940652e-05,
+      "loss": 0.203,
+      "step": 5811
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.046251258894582,
+      "learning_rate": 1.6524018364450243e-05,
+      "loss": 0.1911,
+      "step": 5812
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9299981519638938,
+      "learning_rate": 1.6522770082529596e-05,
+      "loss": 0.2273,
+      "step": 5813
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.867663303689732,
+      "learning_rate": 1.6521521623678445e-05,
+      "loss": 0.2123,
+      "step": 5814
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8664470400619952,
+      "learning_rate": 1.6520272987930652e-05,
+      "loss": 0.215,
+      "step": 5815
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9126660340051115,
+      "learning_rate": 1.6519024175320083e-05,
+      "loss": 0.2117,
+      "step": 5816
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0413653605044608,
+      "learning_rate": 1.651777518588062e-05,
+      "loss": 0.2017,
+      "step": 5817
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3946594861416581,
+      "learning_rate": 1.6516526019646134e-05,
+      "loss": 0.1966,
+      "step": 5818
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3855714810992754,
+      "learning_rate": 1.651527667665051e-05,
+      "loss": 0.2036,
+      "step": 5819
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8732515849837948,
+      "learning_rate": 1.6514027156927645e-05,
+      "loss": 0.2048,
+      "step": 5820
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.534939302966109,
+      "learning_rate": 1.6512777460511416e-05,
+      "loss": 0.2151,
+      "step": 5821
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8623181831359924,
+      "learning_rate": 1.6511527587435736e-05,
+      "loss": 0.1974,
+      "step": 5822
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0794220973605955,
+      "learning_rate": 1.6510277537734503e-05,
+      "loss": 0.2174,
+      "step": 5823
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5431023442504375,
+      "learning_rate": 1.6509027311441622e-05,
+      "loss": 0.2106,
+      "step": 5824
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0567470028378723,
+      "learning_rate": 1.6507776908591008e-05,
+      "loss": 0.2049,
+      "step": 5825
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9985409094547987,
+      "learning_rate": 1.6506526329216577e-05,
+      "loss": 0.226,
+      "step": 5826
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9380862280445232,
+      "learning_rate": 1.6505275573352256e-05,
+      "loss": 0.2239,
+      "step": 5827
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.02642850762763,
+      "learning_rate": 1.6504024641031962e-05,
+      "loss": 0.22,
+      "step": 5828
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5253979388506016,
+      "learning_rate": 1.6502773532289636e-05,
+      "loss": 0.2042,
+      "step": 5829
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3658204535216976,
+      "learning_rate": 1.650152224715921e-05,
+      "loss": 0.1972,
+      "step": 5830
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7792321176824304,
+      "learning_rate": 1.6500270785674622e-05,
+      "loss": 0.1887,
+      "step": 5831
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0631302674580914,
+      "learning_rate": 1.6499019147869826e-05,
+      "loss": 0.1867,
+      "step": 5832
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9993820993819318,
+      "learning_rate": 1.649776733377877e-05,
+      "loss": 0.2004,
+      "step": 5833
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1175435765314348,
+      "learning_rate": 1.6496515343435402e-05,
+      "loss": 0.215,
+      "step": 5834
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7984194198142733,
+      "learning_rate": 1.6495263176873693e-05,
+      "loss": 0.1992,
+      "step": 5835
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8747750342396716,
+      "learning_rate": 1.6494010834127606e-05,
+      "loss": 0.21,
+      "step": 5836
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0164685258600892,
+      "learning_rate": 1.6492758315231105e-05,
+      "loss": 0.2239,
+      "step": 5837
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2855641459344684,
+      "learning_rate": 1.6491505620218164e-05,
+      "loss": 0.2095,
+      "step": 5838
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0643574580669881,
+      "learning_rate": 1.649025274912277e-05,
+      "loss": 0.2103,
+      "step": 5839
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9935202396359508,
+      "learning_rate": 1.6488999701978905e-05,
+      "loss": 0.2034,
+      "step": 5840
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8713947124605383,
+      "learning_rate": 1.6487746478820553e-05,
+      "loss": 0.2141,
+      "step": 5841
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8350889530294335,
+      "learning_rate": 1.6486493079681717e-05,
+      "loss": 0.1887,
+      "step": 5842
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9041295959036052,
+      "learning_rate": 1.6485239504596388e-05,
+      "loss": 0.1954,
+      "step": 5843
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9224791675877131,
+      "learning_rate": 1.6483985753598568e-05,
+      "loss": 0.1839,
+      "step": 5844
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2302875736385745,
+      "learning_rate": 1.6482731826722268e-05,
+      "loss": 0.2014,
+      "step": 5845
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9097035251363624,
+      "learning_rate": 1.6481477724001505e-05,
+      "loss": 0.2199,
+      "step": 5846
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9888881253592569,
+      "learning_rate": 1.648022344547029e-05,
+      "loss": 0.2227,
+      "step": 5847
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9414438514171718,
+      "learning_rate": 1.647896899116265e-05,
+      "loss": 0.2229,
+      "step": 5848
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0139577064085958,
+      "learning_rate": 1.647771436111261e-05,
+      "loss": 0.2129,
+      "step": 5849
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8495467191827688,
+      "learning_rate": 1.64764595553542e-05,
+      "loss": 0.1753,
+      "step": 5850
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1991959064883873,
+      "learning_rate": 1.647520457392146e-05,
+      "loss": 0.2039,
+      "step": 5851
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8550431663988788,
+      "learning_rate": 1.647394941684843e-05,
+      "loss": 0.221,
+      "step": 5852
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8606637731862601,
+      "learning_rate": 1.6472694084169155e-05,
+      "loss": 0.1815,
+      "step": 5853
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0369770786766133,
+      "learning_rate": 1.6471438575917688e-05,
+      "loss": 0.2189,
+      "step": 5854
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0865897305584387,
+      "learning_rate": 1.6470182892128085e-05,
+      "loss": 0.2183,
+      "step": 5855
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9732329093878587,
+      "learning_rate": 1.6468927032834407e-05,
+      "loss": 0.2007,
+      "step": 5856
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9412361804662748,
+      "learning_rate": 1.6467670998070715e-05,
+      "loss": 0.2058,
+      "step": 5857
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2031680501967463,
+      "learning_rate": 1.6466414787871084e-05,
+      "loss": 0.1962,
+      "step": 5858
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9115571475165246,
+      "learning_rate": 1.6465158402269585e-05,
+      "loss": 0.2157,
+      "step": 5859
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7158282006426111,
+      "learning_rate": 1.64639018413003e-05,
+      "loss": 0.1991,
+      "step": 5860
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8445846725413154,
+      "learning_rate": 1.6462645104997313e-05,
+      "loss": 0.1883,
+      "step": 5861
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8719816706104393,
+      "learning_rate": 1.646138819339471e-05,
+      "loss": 0.2028,
+      "step": 5862
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9688970258918401,
+      "learning_rate": 1.646013110652659e-05,
+      "loss": 0.2212,
+      "step": 5863
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.888364909503381,
+      "learning_rate": 1.645887384442705e-05,
+      "loss": 0.2056,
+      "step": 5864
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8982603781397506,
+      "learning_rate": 1.645761640713019e-05,
+      "loss": 0.2103,
+      "step": 5865
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3715133981948604,
+      "learning_rate": 1.645635879467012e-05,
+      "loss": 0.2104,
+      "step": 5866
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.002168238431376,
+      "learning_rate": 1.6455101007080955e-05,
+      "loss": 0.2034,
+      "step": 5867
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9555106086590374,
+      "learning_rate": 1.645384304439681e-05,
+      "loss": 0.2587,
+      "step": 5868
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1883950255284266,
+      "learning_rate": 1.6452584906651807e-05,
+      "loss": 0.2111,
+      "step": 5869
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0928930058177924,
+      "learning_rate": 1.6451326593880072e-05,
+      "loss": 0.2152,
+      "step": 5870
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.022071758104498,
+      "learning_rate": 1.6450068106115745e-05,
+      "loss": 0.2027,
+      "step": 5871
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9165368983625263,
+      "learning_rate": 1.644880944339295e-05,
+      "loss": 0.1926,
+      "step": 5872
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9855256169629955,
+      "learning_rate": 1.6447550605745836e-05,
+      "loss": 0.2126,
+      "step": 5873
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9811264576955663,
+      "learning_rate": 1.644629159320855e-05,
+      "loss": 0.2095,
+      "step": 5874
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8399754500955879,
+      "learning_rate": 1.644503240581524e-05,
+      "loss": 0.1823,
+      "step": 5875
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0565131500331548,
+      "learning_rate": 1.6443773043600058e-05,
+      "loss": 0.2225,
+      "step": 5876
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2131651665261163,
+      "learning_rate": 1.6442513506597175e-05,
+      "loss": 0.2317,
+      "step": 5877
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9404071637103668,
+      "learning_rate": 1.6441253794840745e-05,
+      "loss": 0.1882,
+      "step": 5878
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9782996232409031,
+      "learning_rate": 1.6439993908364942e-05,
+      "loss": 0.1891,
+      "step": 5879
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9048113744479341,
+      "learning_rate": 1.643873384720394e-05,
+      "loss": 0.2286,
+      "step": 5880
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0659766855293744,
+      "learning_rate": 1.643747361139192e-05,
+      "loss": 0.2317,
+      "step": 5881
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8068037627212562,
+      "learning_rate": 1.6436213200963065e-05,
+      "loss": 0.198,
+      "step": 5882
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.877210694130185,
+      "learning_rate": 1.643495261595156e-05,
+      "loss": 0.211,
+      "step": 5883
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8623766726369708,
+      "learning_rate": 1.6433691856391608e-05,
+      "loss": 0.2107,
+      "step": 5884
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0274789198072334,
+      "learning_rate": 1.6432430922317396e-05,
+      "loss": 0.2024,
+      "step": 5885
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3082314211100265,
+      "learning_rate": 1.6431169813763134e-05,
+      "loss": 0.2281,
+      "step": 5886
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9357881374870074,
+      "learning_rate": 1.6429908530763027e-05,
+      "loss": 0.2018,
+      "step": 5887
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8417090098117563,
+      "learning_rate": 1.6428647073351287e-05,
+      "loss": 0.2225,
+      "step": 5888
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8117492998568088,
+      "learning_rate": 1.6427385441562135e-05,
+      "loss": 0.1911,
+      "step": 5889
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9145794378876565,
+      "learning_rate": 1.6426123635429787e-05,
+      "loss": 0.2137,
+      "step": 5890
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2038725188113313,
+      "learning_rate": 1.6424861654988477e-05,
+      "loss": 0.2157,
+      "step": 5891
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9292827245859995,
+      "learning_rate": 1.6423599500272424e-05,
+      "loss": 0.1931,
+      "step": 5892
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9946593686418843,
+      "learning_rate": 1.6422337171315878e-05,
+      "loss": 0.1904,
+      "step": 5893
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3859397288247326,
+      "learning_rate": 1.642107466815307e-05,
+      "loss": 0.2022,
+      "step": 5894
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8991319027869209,
+      "learning_rate": 1.6419811990818252e-05,
+      "loss": 0.2132,
+      "step": 5895
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0301125345203759,
+      "learning_rate": 1.6418549139345667e-05,
+      "loss": 0.2348,
+      "step": 5896
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0126074489129862,
+      "learning_rate": 1.641728611376958e-05,
+      "loss": 0.215,
+      "step": 5897
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8804663931110788,
+      "learning_rate": 1.641602291412424e-05,
+      "loss": 0.1978,
+      "step": 5898
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9066843573868038,
+      "learning_rate": 1.641475954044392e-05,
+      "loss": 0.214,
+      "step": 5899
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8398951594937323,
+      "learning_rate": 1.641349599276288e-05,
+      "loss": 0.1948,
+      "step": 5900
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.88342216108323,
+      "learning_rate": 1.64122322711154e-05,
+      "loss": 0.2018,
+      "step": 5901
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8609414098283307,
+      "learning_rate": 1.6410968375535762e-05,
+      "loss": 0.2192,
+      "step": 5902
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7894700819728515,
+      "learning_rate": 1.640970430605824e-05,
+      "loss": 0.2002,
+      "step": 5903
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.948260892075998,
+      "learning_rate": 1.640844006271713e-05,
+      "loss": 0.2033,
+      "step": 5904
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9018493655463135,
+      "learning_rate": 1.640717564554672e-05,
+      "loss": 0.2039,
+      "step": 5905
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8166650211310612,
+      "learning_rate": 1.6405911054581307e-05,
+      "loss": 0.1891,
+      "step": 5906
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9446127345677011,
+      "learning_rate": 1.6404646289855194e-05,
+      "loss": 0.2209,
+      "step": 5907
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8743844759657704,
+      "learning_rate": 1.640338135140269e-05,
+      "loss": 0.1924,
+      "step": 5908
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4321428005760437,
+      "learning_rate": 1.640211623925811e-05,
+      "loss": 0.2296,
+      "step": 5909
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9025069904442802,
+      "learning_rate": 1.640085095345576e-05,
+      "loss": 0.2115,
+      "step": 5910
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8714708162046169,
+      "learning_rate": 1.6399585494029968e-05,
+      "loss": 0.1909,
+      "step": 5911
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1490013763920408,
+      "learning_rate": 1.639831986101506e-05,
+      "loss": 0.2036,
+      "step": 5912
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9508681180299058,
+      "learning_rate": 1.639705405444536e-05,
+      "loss": 0.1905,
+      "step": 5913
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.235841414052421,
+      "learning_rate": 1.6395788074355212e-05,
+      "loss": 0.1929,
+      "step": 5914
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1104003104638973,
+      "learning_rate": 1.639452192077895e-05,
+      "loss": 0.1946,
+      "step": 5915
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9560855958680096,
+      "learning_rate": 1.6393255593750917e-05,
+      "loss": 0.2202,
+      "step": 5916
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8003227650607686,
+      "learning_rate": 1.6391989093305468e-05,
+      "loss": 0.1872,
+      "step": 5917
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.093372947587186,
+      "learning_rate": 1.6390722419476952e-05,
+      "loss": 0.215,
+      "step": 5918
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8234618930446632,
+      "learning_rate": 1.638945557229973e-05,
+      "loss": 0.2064,
+      "step": 5919
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0258032410734896,
+      "learning_rate": 1.6388188551808166e-05,
+      "loss": 0.1988,
+      "step": 5920
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1774833686716462,
+      "learning_rate": 1.6386921358036624e-05,
+      "loss": 0.1894,
+      "step": 5921
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8458940053240332,
+      "learning_rate": 1.638565399101948e-05,
+      "loss": 0.211,
+      "step": 5922
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0143614561029222,
+      "learning_rate": 1.6384386450791114e-05,
+      "loss": 0.2019,
+      "step": 5923
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9030318476331984,
+      "learning_rate": 1.6383118737385903e-05,
+      "loss": 0.1939,
+      "step": 5924
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9946142914549149,
+      "learning_rate": 1.6381850850838232e-05,
+      "loss": 0.1889,
+      "step": 5925
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9871772763534351,
+      "learning_rate": 1.63805827911825e-05,
+      "loss": 0.1854,
+      "step": 5926
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9040610810639874,
+      "learning_rate": 1.63793145584531e-05,
+      "loss": 0.1925,
+      "step": 5927
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9290544683412078,
+      "learning_rate": 1.637804615268443e-05,
+      "loss": 0.2287,
+      "step": 5928
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9562484980139982,
+      "learning_rate": 1.63767775739109e-05,
+      "loss": 0.2036,
+      "step": 5929
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8777655403386205,
+      "learning_rate": 1.6375508822166917e-05,
+      "loss": 0.2296,
+      "step": 5930
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8735047939100099,
+      "learning_rate": 1.63742398974869e-05,
+      "loss": 0.1851,
+      "step": 5931
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.935282896007759,
+      "learning_rate": 1.6372970799905262e-05,
+      "loss": 0.2209,
+      "step": 5932
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0901327351521781,
+      "learning_rate": 1.6371701529456433e-05,
+      "loss": 0.2119,
+      "step": 5933
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1528605527009306,
+      "learning_rate": 1.6370432086174837e-05,
+      "loss": 0.1906,
+      "step": 5934
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8206100964092897,
+      "learning_rate": 1.6369162470094915e-05,
+      "loss": 0.2065,
+      "step": 5935
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8480007652060755,
+      "learning_rate": 1.63678926812511e-05,
+      "loss": 0.1984,
+      "step": 5936
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2839298260001515,
+      "learning_rate": 1.6366622719677834e-05,
+      "loss": 0.2135,
+      "step": 5937
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9534265566922148,
+      "learning_rate": 1.6365352585409572e-05,
+      "loss": 0.2263,
+      "step": 5938
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.962950305698948,
+      "learning_rate": 1.636408227848076e-05,
+      "loss": 0.2058,
+      "step": 5939
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9968989531366338,
+      "learning_rate": 1.6362811798925852e-05,
+      "loss": 0.2056,
+      "step": 5940
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9642269498234763,
+      "learning_rate": 1.636154114677932e-05,
+      "loss": 0.2174,
+      "step": 5941
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8536369510964444,
+      "learning_rate": 1.636027032207562e-05,
+      "loss": 0.1977,
+      "step": 5942
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.701056712938503,
+      "learning_rate": 1.6358999324849235e-05,
+      "loss": 0.2213,
+      "step": 5943
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7754497041155105,
+      "learning_rate": 1.635772815513463e-05,
+      "loss": 0.1935,
+      "step": 5944
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9182076271424672,
+      "learning_rate": 1.635645681296629e-05,
+      "loss": 0.1837,
+      "step": 5945
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9513656758902006,
+      "learning_rate": 1.63551852983787e-05,
+      "loss": 0.2034,
+      "step": 5946
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8099373859425659,
+      "learning_rate": 1.635391361140635e-05,
+      "loss": 0.2035,
+      "step": 5947
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2689296573815851,
+      "learning_rate": 1.6352641752083734e-05,
+      "loss": 0.2376,
+      "step": 5948
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3730141177994373,
+      "learning_rate": 1.6351369720445353e-05,
+      "loss": 0.2268,
+      "step": 5949
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8013748927283166,
+      "learning_rate": 1.6350097516525705e-05,
+      "loss": 0.2,
+      "step": 5950
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.116575591861395,
+      "learning_rate": 1.634882514035931e-05,
+      "loss": 0.1904,
+      "step": 5951
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0059605669840697,
+      "learning_rate": 1.6347552591980672e-05,
+      "loss": 0.2203,
+      "step": 5952
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9044501685247852,
+      "learning_rate": 1.634627987142431e-05,
+      "loss": 0.2197,
+      "step": 5953
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1633662000849947,
+      "learning_rate": 1.6345006978724748e-05,
+      "loss": 0.2302,
+      "step": 5954
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0048216985971365,
+      "learning_rate": 1.6343733913916516e-05,
+      "loss": 0.199,
+      "step": 5955
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7956734026992229,
+      "learning_rate": 1.634246067703414e-05,
+      "loss": 0.2101,
+      "step": 5956
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0101755028948003,
+      "learning_rate": 1.6341187268112162e-05,
+      "loss": 0.2084,
+      "step": 5957
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9222892642766556,
+      "learning_rate": 1.633991368718512e-05,
+      "loss": 0.2193,
+      "step": 5958
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9677295069555725,
+      "learning_rate": 1.6338639934287563e-05,
+      "loss": 0.2085,
+      "step": 5959
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9158364680199044,
+      "learning_rate": 1.6337366009454037e-05,
+      "loss": 0.1924,
+      "step": 5960
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.944337670541037,
+      "learning_rate": 1.6336091912719102e-05,
+      "loss": 0.2005,
+      "step": 5961
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8085176627955762,
+      "learning_rate": 1.6334817644117316e-05,
+      "loss": 0.2191,
+      "step": 5962
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4056936577366979,
+      "learning_rate": 1.633354320368324e-05,
+      "loss": 0.1941,
+      "step": 5963
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8614378964196189,
+      "learning_rate": 1.6332268591451454e-05,
+      "loss": 0.2031,
+      "step": 5964
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.934286179464845,
+      "learning_rate": 1.633099380745652e-05,
+      "loss": 0.1998,
+      "step": 5965
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9754800276459915,
+      "learning_rate": 1.6329718851733024e-05,
+      "loss": 0.2147,
+      "step": 5966
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1766704671644548,
+      "learning_rate": 1.6328443724315544e-05,
+      "loss": 0.2143,
+      "step": 5967
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0562421455698277,
+      "learning_rate": 1.6327168425238672e-05,
+      "loss": 0.213,
+      "step": 5968
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9481384551774131,
+      "learning_rate": 1.6325892954536997e-05,
+      "loss": 0.1783,
+      "step": 5969
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1198994360550176,
+      "learning_rate": 1.6324617312245123e-05,
+      "loss": 0.1809,
+      "step": 5970
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.865446262601342,
+      "learning_rate": 1.632334149839764e-05,
+      "loss": 0.167,
+      "step": 5971
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8359926893079918,
+      "learning_rate": 1.632206551302917e-05,
+      "loss": 0.2142,
+      "step": 5972
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1003436066740737,
+      "learning_rate": 1.632078935617431e-05,
+      "loss": 0.1987,
+      "step": 5973
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1175808214952012,
+      "learning_rate": 1.6319513027867683e-05,
+      "loss": 0.2127,
+      "step": 5974
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.09109343841727,
+      "learning_rate": 1.631823652814391e-05,
+      "loss": 0.2195,
+      "step": 5975
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5938819437245713,
+      "learning_rate": 1.631695985703761e-05,
+      "loss": 0.1932,
+      "step": 5976
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1572532065360304,
+      "learning_rate": 1.631568301458342e-05,
+      "loss": 0.2291,
+      "step": 5977
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7935200691091274,
+      "learning_rate": 1.6314406000815975e-05,
+      "loss": 0.1829,
+      "step": 5978
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1058627197706676,
+      "learning_rate": 1.6313128815769904e-05,
+      "loss": 0.2027,
+      "step": 5979
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8970087572102301,
+      "learning_rate": 1.631185145947986e-05,
+      "loss": 0.2225,
+      "step": 5980
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8393393193043988,
+      "learning_rate": 1.631057393198049e-05,
+      "loss": 0.2061,
+      "step": 5981
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9851589297321419,
+      "learning_rate": 1.6309296233306446e-05,
+      "loss": 0.2127,
+      "step": 5982
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3751830177191615,
+      "learning_rate": 1.6308018363492385e-05,
+      "loss": 0.2221,
+      "step": 5983
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8238273156343242,
+      "learning_rate": 1.630674032257297e-05,
+      "loss": 0.1916,
+      "step": 5984
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0898527791933963,
+      "learning_rate": 1.6305462110582863e-05,
+      "loss": 0.2429,
+      "step": 5985
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.302930626821737,
+      "learning_rate": 1.6304183727556747e-05,
+      "loss": 0.2166,
+      "step": 5986
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.39479291669915,
+      "learning_rate": 1.6302905173529285e-05,
+      "loss": 0.2118,
+      "step": 5987
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4736219557226864,
+      "learning_rate": 1.6301626448535168e-05,
+      "loss": 0.2022,
+      "step": 5988
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9127817268931459,
+      "learning_rate": 1.6300347552609074e-05,
+      "loss": 0.1959,
+      "step": 5989
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.008387597169717,
+      "learning_rate": 1.62990684857857e-05,
+      "loss": 0.227,
+      "step": 5990
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8037508248908429,
+      "learning_rate": 1.6297789248099736e-05,
+      "loss": 0.21,
+      "step": 5991
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.087320987869786,
+      "learning_rate": 1.6296509839585885e-05,
+      "loss": 0.2035,
+      "step": 5992
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8294639592735905,
+      "learning_rate": 1.6295230260278847e-05,
+      "loss": 0.2123,
+      "step": 5993
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7337293869270486,
+      "learning_rate": 1.6293950510213335e-05,
+      "loss": 0.2082,
+      "step": 5994
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8497313029047259,
+      "learning_rate": 1.6292670589424057e-05,
+      "loss": 0.2148,
+      "step": 5995
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6162721450291733,
+      "learning_rate": 1.6291390497945737e-05,
+      "loss": 0.2115,
+      "step": 5996
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8330036247733196,
+      "learning_rate": 1.6290110235813094e-05,
+      "loss": 0.1815,
+      "step": 5997
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1318812757855268,
+      "learning_rate": 1.6288829803060853e-05,
+      "loss": 0.2048,
+      "step": 5998
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.82097750858233,
+      "learning_rate": 1.6287549199723745e-05,
+      "loss": 0.2189,
+      "step": 5999
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8236429171653487,
+      "learning_rate": 1.6286268425836517e-05,
+      "loss": 0.2,
+      "step": 6000
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5845029614639,
+      "learning_rate": 1.62849874814339e-05,
+      "loss": 0.1795,
+      "step": 6001
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4522997519291503,
+      "learning_rate": 1.6283706366550646e-05,
+      "loss": 0.2155,
+      "step": 6002
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9965723060845276,
+      "learning_rate": 1.6282425081221498e-05,
+      "loss": 0.1913,
+      "step": 6003
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0240651787267885,
+      "learning_rate": 1.6281143625481214e-05,
+      "loss": 0.1995,
+      "step": 6004
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7303063496135936,
+      "learning_rate": 1.627986199936456e-05,
+      "loss": 0.2153,
+      "step": 6005
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9488526266709439,
+      "learning_rate": 1.6278580202906287e-05,
+      "loss": 0.2074,
+      "step": 6006
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8874494026802576,
+      "learning_rate": 1.6277298236141177e-05,
+      "loss": 0.234,
+      "step": 6007
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9477817405838913,
+      "learning_rate": 1.6276016099103995e-05,
+      "loss": 0.2199,
+      "step": 6008
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1014385615064697,
+      "learning_rate": 1.6274733791829522e-05,
+      "loss": 0.2199,
+      "step": 6009
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0774681703386346,
+      "learning_rate": 1.627345131435254e-05,
+      "loss": 0.2538,
+      "step": 6010
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2211872301986042,
+      "learning_rate": 1.6272168666707838e-05,
+      "loss": 0.2199,
+      "step": 6011
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7944085611231179,
+      "learning_rate": 1.627088584893021e-05,
+      "loss": 0.1968,
+      "step": 6012
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1626975384193183,
+      "learning_rate": 1.6269602861054442e-05,
+      "loss": 0.2084,
+      "step": 6013
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9266890007568821,
+      "learning_rate": 1.6268319703115348e-05,
+      "loss": 0.1952,
+      "step": 6014
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0407450828399152,
+      "learning_rate": 1.6267036375147728e-05,
+      "loss": 0.2085,
+      "step": 6015
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9540943475088866,
+      "learning_rate": 1.6265752877186386e-05,
+      "loss": 0.1963,
+      "step": 6016
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9119042826028682,
+      "learning_rate": 1.626446920926615e-05,
+      "loss": 0.2015,
+      "step": 6017
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3939619942230101,
+      "learning_rate": 1.626318537142183e-05,
+      "loss": 0.1915,
+      "step": 6018
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4344270785747297,
+      "learning_rate": 1.6261901363688257e-05,
+      "loss": 0.2255,
+      "step": 6019
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.079480238045351,
+      "learning_rate": 1.626061718610025e-05,
+      "loss": 0.2304,
+      "step": 6020
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4324024219544331,
+      "learning_rate": 1.625933283869265e-05,
+      "loss": 0.1926,
+      "step": 6021
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7344142079737424,
+      "learning_rate": 1.6258048321500294e-05,
+      "loss": 0.1956,
+      "step": 6022
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1070084317838647,
+      "learning_rate": 1.6256763634558024e-05,
+      "loss": 0.2114,
+      "step": 6023
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9764029510305736,
+      "learning_rate": 1.625547877790069e-05,
+      "loss": 0.1983,
+      "step": 6024
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9096497335946024,
+      "learning_rate": 1.6254193751563137e-05,
+      "loss": 0.2244,
+      "step": 6025
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9647081068711773,
+      "learning_rate": 1.6252908555580223e-05,
+      "loss": 0.2021,
+      "step": 6026
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.003658864196987,
+      "learning_rate": 1.6251623189986815e-05,
+      "loss": 0.2282,
+      "step": 6027
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0027742910910584,
+      "learning_rate": 1.6250337654817774e-05,
+      "loss": 0.2084,
+      "step": 6028
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0873519198431463,
+      "learning_rate": 1.624905195010797e-05,
+      "loss": 0.2172,
+      "step": 6029
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0149186797500493,
+      "learning_rate": 1.6247766075892283e-05,
+      "loss": 0.2034,
+      "step": 6030
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9063023009686626,
+      "learning_rate": 1.624648003220558e-05,
+      "loss": 0.1857,
+      "step": 6031
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7650021001433587,
+      "learning_rate": 1.624519381908276e-05,
+      "loss": 0.2074,
+      "step": 6032
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3530888182017393,
+      "learning_rate": 1.6243907436558705e-05,
+      "loss": 0.2268,
+      "step": 6033
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.911985352429917,
+      "learning_rate": 1.624262088466831e-05,
+      "loss": 0.2375,
+      "step": 6034
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6303058569455007,
+      "learning_rate": 1.6241334163446465e-05,
+      "loss": 0.2074,
+      "step": 6035
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.459715495070587,
+      "learning_rate": 1.6240047272928082e-05,
+      "loss": 0.2149,
+      "step": 6036
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8229690605766699,
+      "learning_rate": 1.6238760213148064e-05,
+      "loss": 0.1893,
+      "step": 6037
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6615636863545942,
+      "learning_rate": 1.6237472984141322e-05,
+      "loss": 0.2066,
+      "step": 6038
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8825843487781814,
+      "learning_rate": 1.623618558594277e-05,
+      "loss": 0.1996,
+      "step": 6039
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8986305451461882,
+      "learning_rate": 1.6234898018587336e-05,
+      "loss": 0.193,
+      "step": 6040
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8603017400911558,
+      "learning_rate": 1.623361028210994e-05,
+      "loss": 0.2074,
+      "step": 6041
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0757625435404399,
+      "learning_rate": 1.6232322376545516e-05,
+      "loss": 0.1834,
+      "step": 6042
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9255175015178886,
+      "learning_rate": 1.623103430192899e-05,
+      "loss": 0.1739,
+      "step": 6043
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0095454531080827,
+      "learning_rate": 1.6229746058295312e-05,
+      "loss": 0.2135,
+      "step": 6044
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7730502584757073,
+      "learning_rate": 1.6228457645679414e-05,
+      "loss": 0.2339,
+      "step": 6045
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9101283209796647,
+      "learning_rate": 1.6227169064116255e-05,
+      "loss": 0.2184,
+      "step": 6046
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4470438726261499,
+      "learning_rate": 1.622588031364078e-05,
+      "loss": 0.2147,
+      "step": 6047
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6151800077895164,
+      "learning_rate": 1.6224591394287954e-05,
+      "loss": 0.2114,
+      "step": 6048
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.141700862481167,
+      "learning_rate": 1.6223302306092733e-05,
+      "loss": 0.2097,
+      "step": 6049
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.6258373665253605,
+      "learning_rate": 1.6222013049090086e-05,
+      "loss": 0.2116,
+      "step": 6050
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0270473789404706,
+      "learning_rate": 1.6220723623314983e-05,
+      "loss": 0.2229,
+      "step": 6051
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9431739891813424,
+      "learning_rate": 1.6219434028802402e-05,
+      "loss": 0.2079,
+      "step": 6052
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0931954007419473,
+      "learning_rate": 1.621814426558732e-05,
+      "loss": 0.225,
+      "step": 6053
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5694243876987546,
+      "learning_rate": 1.6216854333704725e-05,
+      "loss": 0.2058,
+      "step": 6054
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.858906512456257,
+      "learning_rate": 1.6215564233189606e-05,
+      "loss": 0.2106,
+      "step": 6055
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8471245246956891,
+      "learning_rate": 1.621427396407695e-05,
+      "loss": 0.2,
+      "step": 6056
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0547280866759583,
+      "learning_rate": 1.6212983526401767e-05,
+      "loss": 0.2011,
+      "step": 6057
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.119564287962234,
+      "learning_rate": 1.6211692920199054e-05,
+      "loss": 0.1756,
+      "step": 6058
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1099677356408273,
+      "learning_rate": 1.621040214550382e-05,
+      "loss": 0.177,
+      "step": 6059
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0913921470754946,
+      "learning_rate": 1.6209111202351076e-05,
+      "loss": 0.22,
+      "step": 6060
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0658146194201514,
+      "learning_rate": 1.620782009077584e-05,
+      "loss": 0.2421,
+      "step": 6061
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5260663580908203,
+      "learning_rate": 1.6206528810813135e-05,
+      "loss": 0.1937,
+      "step": 6062
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9794394801934251,
+      "learning_rate": 1.6205237362497982e-05,
+      "loss": 0.2261,
+      "step": 6063
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5247938545986877,
+      "learning_rate": 1.6203945745865418e-05,
+      "loss": 0.2084,
+      "step": 6064
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.242093586082393,
+      "learning_rate": 1.6202653960950474e-05,
+      "loss": 0.2221,
+      "step": 6065
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6609521159811682,
+      "learning_rate": 1.6201362007788193e-05,
+      "loss": 0.1981,
+      "step": 6066
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9717974454332525,
+      "learning_rate": 1.6200069886413612e-05,
+      "loss": 0.2035,
+      "step": 6067
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.040278624958183,
+      "learning_rate": 1.6198777596861792e-05,
+      "loss": 0.2517,
+      "step": 6068
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.060313397045809,
+      "learning_rate": 1.6197485139167775e-05,
+      "loss": 0.2244,
+      "step": 6069
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4185748386978991,
+      "learning_rate": 1.619619251336663e-05,
+      "loss": 0.1927,
+      "step": 6070
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1902479620651243,
+      "learning_rate": 1.6194899719493404e-05,
+      "loss": 0.2443,
+      "step": 6071
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0869353007806424,
+      "learning_rate": 1.619360675758318e-05,
+      "loss": 0.1993,
+      "step": 6072
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4174221590710505,
+      "learning_rate": 1.619231362767102e-05,
+      "loss": 0.1729,
+      "step": 6073
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0794084394279622,
+      "learning_rate": 1.6191020329792003e-05,
+      "loss": 0.1664,
+      "step": 6074
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7769533876934619,
+      "learning_rate": 1.6189726863981212e-05,
+      "loss": 0.1774,
+      "step": 6075
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0305866506804409,
+      "learning_rate": 1.618843323027373e-05,
+      "loss": 0.2087,
+      "step": 6076
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8876921764605629,
+      "learning_rate": 1.6187139428704645e-05,
+      "loss": 0.2175,
+      "step": 6077
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0822377892878023,
+      "learning_rate": 1.6185845459309053e-05,
+      "loss": 0.1939,
+      "step": 6078
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9176729470876167,
+      "learning_rate": 1.6184551322122056e-05,
+      "loss": 0.1904,
+      "step": 6079
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1716604783757618,
+      "learning_rate": 1.6183257017178754e-05,
+      "loss": 0.1996,
+      "step": 6080
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2684319856473547,
+      "learning_rate": 1.6181962544514257e-05,
+      "loss": 0.1879,
+      "step": 6081
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8193278183743143,
+      "learning_rate": 1.6180667904163675e-05,
+      "loss": 0.1868,
+      "step": 6082
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5171040957386068,
+      "learning_rate": 1.617937309616213e-05,
+      "loss": 0.219,
+      "step": 6083
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.33064610781682,
+      "learning_rate": 1.6178078120544735e-05,
+      "loss": 0.205,
+      "step": 6084
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6137351930166788,
+      "learning_rate": 1.6176782977346626e-05,
+      "loss": 0.1999,
+      "step": 6085
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5597204402629181,
+      "learning_rate": 1.6175487666602928e-05,
+      "loss": 0.215,
+      "step": 6086
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1470285061130765,
+      "learning_rate": 1.617419218834878e-05,
+      "loss": 0.2121,
+      "step": 6087
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2800202663819764,
+      "learning_rate": 1.617289654261932e-05,
+      "loss": 0.2111,
+      "step": 6088
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.545652925213808,
+      "learning_rate": 1.6171600729449693e-05,
+      "loss": 0.2338,
+      "step": 6089
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1290088033332193,
+      "learning_rate": 1.617030474887505e-05,
+      "loss": 0.1965,
+      "step": 6090
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.969297537706917,
+      "learning_rate": 1.616900860093054e-05,
+      "loss": 0.2065,
+      "step": 6091
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.190717952530125,
+      "learning_rate": 1.616771228565132e-05,
+      "loss": 0.195,
+      "step": 6092
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1595524076758887,
+      "learning_rate": 1.616641580307256e-05,
+      "loss": 0.2136,
+      "step": 6093
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6403050277216027,
+      "learning_rate": 1.616511915322942e-05,
+      "loss": 0.1861,
+      "step": 6094
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6437870280410332,
+      "learning_rate": 1.6163822336157076e-05,
+      "loss": 0.21,
+      "step": 6095
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8419726499255795,
+      "learning_rate": 1.6162525351890702e-05,
+      "loss": 0.207,
+      "step": 6096
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1212531960451904,
+      "learning_rate": 1.6161228200465485e-05,
+      "loss": 0.1988,
+      "step": 6097
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.051091374373422,
+      "learning_rate": 1.61599308819166e-05,
+      "loss": 0.2186,
+      "step": 6098
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9688417139595524,
+      "learning_rate": 1.6158633396279243e-05,
+      "loss": 0.243,
+      "step": 6099
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2062547757471067,
+      "learning_rate": 1.6157335743588606e-05,
+      "loss": 0.2107,
+      "step": 6100
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0249173195322316,
+      "learning_rate": 1.6156037923879893e-05,
+      "loss": 0.2127,
+      "step": 6101
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3317006096755801,
+      "learning_rate": 1.61547399371883e-05,
+      "loss": 0.2066,
+      "step": 6102
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0908772213653302,
+      "learning_rate": 1.6153441783549043e-05,
+      "loss": 0.2107,
+      "step": 6103
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.061902894350475,
+      "learning_rate": 1.6152143462997325e-05,
+      "loss": 0.2165,
+      "step": 6104
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4954267996419317,
+      "learning_rate": 1.615084497556837e-05,
+      "loss": 0.2195,
+      "step": 6105
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5649819771869884,
+      "learning_rate": 1.61495463212974e-05,
+      "loss": 0.2019,
+      "step": 6106
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2063094496662974,
+      "learning_rate": 1.6148247500219635e-05,
+      "loss": 0.2021,
+      "step": 6107
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.995772199176662,
+      "learning_rate": 1.614694851237031e-05,
+      "loss": 0.1937,
+      "step": 6108
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9327743983007237,
+      "learning_rate": 1.614564935778466e-05,
+      "loss": 0.2104,
+      "step": 6109
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4621831264462208,
+      "learning_rate": 1.6144350036497925e-05,
+      "loss": 0.1952,
+      "step": 6110
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.882534748644571,
+      "learning_rate": 1.6143050548545348e-05,
+      "loss": 0.1974,
+      "step": 6111
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1978580185486694,
+      "learning_rate": 1.6141750893962176e-05,
+      "loss": 0.1929,
+      "step": 6112
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1465567091981104,
+      "learning_rate": 1.6140451072783666e-05,
+      "loss": 0.2049,
+      "step": 6113
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2664074830992735,
+      "learning_rate": 1.613915108504507e-05,
+      "loss": 0.1923,
+      "step": 6114
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0980669945835926,
+      "learning_rate": 1.613785093078166e-05,
+      "loss": 0.1896,
+      "step": 6115
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.631937385976018,
+      "learning_rate": 1.6136550610028694e-05,
+      "loss": 0.1885,
+      "step": 6116
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5889117710056704,
+      "learning_rate": 1.6135250122821445e-05,
+      "loss": 0.1946,
+      "step": 6117
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5134914654859153,
+      "learning_rate": 1.6133949469195194e-05,
+      "loss": 0.2162,
+      "step": 6118
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2476327546290005,
+      "learning_rate": 1.6132648649185214e-05,
+      "loss": 0.2031,
+      "step": 6119
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8453200099035407,
+      "learning_rate": 1.6131347662826793e-05,
+      "loss": 0.1868,
+      "step": 6120
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.319105677189608,
+      "learning_rate": 1.613004651015522e-05,
+      "loss": 0.2438,
+      "step": 6121
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0640604006619943,
+      "learning_rate": 1.612874519120579e-05,
+      "loss": 0.2214,
+      "step": 6122
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4071140029611668,
+      "learning_rate": 1.61274437060138e-05,
+      "loss": 0.2094,
+      "step": 6123
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4588025197496317,
+      "learning_rate": 1.6126142054614556e-05,
+      "loss": 0.2015,
+      "step": 6124
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1702119223872676,
+      "learning_rate": 1.612484023704336e-05,
+      "loss": 0.1994,
+      "step": 6125
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3620761809710489,
+      "learning_rate": 1.6123538253335526e-05,
+      "loss": 0.1852,
+      "step": 6126
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4131413570897076,
+      "learning_rate": 1.612223610352637e-05,
+      "loss": 0.2007,
+      "step": 6127
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.835054207929354,
+      "learning_rate": 1.612093378765122e-05,
+      "loss": 0.1974,
+      "step": 6128
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6071319397296138,
+      "learning_rate": 1.611963130574539e-05,
+      "loss": 0.2054,
+      "step": 6129
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7980661677245373,
+      "learning_rate": 1.6118328657844216e-05,
+      "loss": 0.2101,
+      "step": 6130
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5776819747938176,
+      "learning_rate": 1.6117025843983032e-05,
+      "loss": 0.2358,
+      "step": 6131
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.877337042985936,
+      "learning_rate": 1.6115722864197178e-05,
+      "loss": 0.2195,
+      "step": 6132
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5470479776051027,
+      "learning_rate": 1.6114419718521994e-05,
+      "loss": 0.2126,
+      "step": 6133
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.749177645710888,
+      "learning_rate": 1.6113116406992833e-05,
+      "loss": 0.1907,
+      "step": 6134
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4421813486978317,
+      "learning_rate": 1.611181292964504e-05,
+      "loss": 0.207,
+      "step": 6135
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.520289919949835,
+      "learning_rate": 1.611050928651398e-05,
+      "loss": 0.2052,
+      "step": 6136
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.293791051110462,
+      "learning_rate": 1.6109205477635006e-05,
+      "loss": 0.2182,
+      "step": 6137
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.513386645098428,
+      "learning_rate": 1.610790150304349e-05,
+      "loss": 0.2319,
+      "step": 6138
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0876460560077863,
+      "learning_rate": 1.61065973627748e-05,
+      "loss": 0.2333,
+      "step": 6139
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1635359350738135,
+      "learning_rate": 1.6105293056864314e-05,
+      "loss": 0.1949,
+      "step": 6140
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9290528759839747,
+      "learning_rate": 1.6103988585347408e-05,
+      "loss": 0.2172,
+      "step": 6141
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5256634967346925,
+      "learning_rate": 1.6102683948259467e-05,
+      "loss": 0.1897,
+      "step": 6142
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5949036318919134,
+      "learning_rate": 1.610137914563588e-05,
+      "loss": 0.2081,
+      "step": 6143
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.454851335074028,
+      "learning_rate": 1.6100074177512038e-05,
+      "loss": 0.199,
+      "step": 6144
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7598947260802447,
+      "learning_rate": 1.6098769043923338e-05,
+      "loss": 0.2103,
+      "step": 6145
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4634098370805022,
+      "learning_rate": 1.6097463744905187e-05,
+      "loss": 0.1987,
+      "step": 6146
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2310446246519797,
+      "learning_rate": 1.6096158280492984e-05,
+      "loss": 0.1591,
+      "step": 6147
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3524333215013153,
+      "learning_rate": 1.6094852650722145e-05,
+      "loss": 0.2018,
+      "step": 6148
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6759301777793807,
+      "learning_rate": 1.6093546855628085e-05,
+      "loss": 0.1788,
+      "step": 6149
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2458571119083635,
+      "learning_rate": 1.609224089524622e-05,
+      "loss": 0.2098,
+      "step": 6150
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5002867548310084,
+      "learning_rate": 1.6090934769611976e-05,
+      "loss": 0.2265,
+      "step": 6151
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1122774148111543,
+      "learning_rate": 1.6089628478760785e-05,
+      "loss": 0.2059,
+      "step": 6152
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1362389543734666,
+      "learning_rate": 1.6088322022728076e-05,
+      "loss": 0.2541,
+      "step": 6153
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.268390507330005,
+      "learning_rate": 1.608701540154929e-05,
+      "loss": 0.1765,
+      "step": 6154
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6803598721739206,
+      "learning_rate": 1.6085708615259863e-05,
+      "loss": 0.1931,
+      "step": 6155
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.066069526341508,
+      "learning_rate": 1.608440166389525e-05,
+      "loss": 0.1848,
+      "step": 6156
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1775493264874153,
+      "learning_rate": 1.6083094547490895e-05,
+      "loss": 0.1911,
+      "step": 6157
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.316939659878754,
+      "learning_rate": 1.6081787266082258e-05,
+      "loss": 0.204,
+      "step": 6158
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1886676203084303,
+      "learning_rate": 1.60804798197048e-05,
+      "loss": 0.193,
+      "step": 6159
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0711579452613742,
+      "learning_rate": 1.6079172208393985e-05,
+      "loss": 0.1887,
+      "step": 6160
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3376326440975657,
+      "learning_rate": 1.607786443218528e-05,
+      "loss": 0.1986,
+      "step": 6161
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1083942748161777,
+      "learning_rate": 1.6076556491114152e-05,
+      "loss": 0.2099,
+      "step": 6162
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.614406990008451,
+      "learning_rate": 1.6075248385216093e-05,
+      "loss": 0.2101,
+      "step": 6163
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0930356453004402,
+      "learning_rate": 1.6073940114526574e-05,
+      "loss": 0.2174,
+      "step": 6164
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0492328588433775,
+      "learning_rate": 1.607263167908109e-05,
+      "loss": 0.2359,
+      "step": 6165
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2424165398353795,
+      "learning_rate": 1.6071323078915128e-05,
+      "loss": 0.2419,
+      "step": 6166
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8259912298969864,
+      "learning_rate": 1.6070014314064183e-05,
+      "loss": 0.2406,
+      "step": 6167
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8474954794357589,
+      "learning_rate": 1.6068705384563757e-05,
+      "loss": 0.2092,
+      "step": 6168
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0906208809203402,
+      "learning_rate": 1.6067396290449356e-05,
+      "loss": 0.2131,
+      "step": 6169
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6127793598314828,
+      "learning_rate": 1.6066087031756485e-05,
+      "loss": 0.2096,
+      "step": 6170
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9349034239176398,
+      "learning_rate": 1.6064777608520666e-05,
+      "loss": 0.2152,
+      "step": 6171
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.393936008162501,
+      "learning_rate": 1.606346802077741e-05,
+      "loss": 0.2057,
+      "step": 6172
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.204797087355292,
+      "learning_rate": 1.6062158268562237e-05,
+      "loss": 0.207,
+      "step": 6173
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1014356914347432,
+      "learning_rate": 1.6060848351910685e-05,
+      "loss": 0.2191,
+      "step": 6174
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9548581883876565,
+      "learning_rate": 1.6059538270858275e-05,
+      "loss": 0.1874,
+      "step": 6175
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1023120629875385,
+      "learning_rate": 1.6058228025440548e-05,
+      "loss": 0.2297,
+      "step": 6176
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9500603896039294,
+      "learning_rate": 1.6056917615693045e-05,
+      "loss": 0.2156,
+      "step": 6177
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0595746338846443,
+      "learning_rate": 1.605560704165131e-05,
+      "loss": 0.1836,
+      "step": 6178
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.601444715454401,
+      "learning_rate": 1.6054296303350886e-05,
+      "loss": 0.2015,
+      "step": 6179
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2146261540538896,
+      "learning_rate": 1.605298540082734e-05,
+      "loss": 0.2129,
+      "step": 6180
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.873388952037026,
+      "learning_rate": 1.605167433411622e-05,
+      "loss": 0.2025,
+      "step": 6181
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.726024719124285,
+      "learning_rate": 1.6050363103253093e-05,
+      "loss": 0.1764,
+      "step": 6182
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0938330909494385,
+      "learning_rate": 1.6049051708273526e-05,
+      "loss": 0.2067,
+      "step": 6183
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.079586704593921,
+      "learning_rate": 1.6047740149213085e-05,
+      "loss": 0.1896,
+      "step": 6184
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2714408408105322,
+      "learning_rate": 1.6046428426107354e-05,
+      "loss": 0.2357,
+      "step": 6185
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2726345863274997,
+      "learning_rate": 1.604511653899191e-05,
+      "loss": 0.2032,
+      "step": 6186
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1207757688502018,
+      "learning_rate": 1.604380448790234e-05,
+      "loss": 0.2256,
+      "step": 6187
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.093894661508314,
+      "learning_rate": 1.604249227287423e-05,
+      "loss": 0.2159,
+      "step": 6188
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.197879422109431,
+      "learning_rate": 1.6041179893943178e-05,
+      "loss": 0.2238,
+      "step": 6189
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.019478303665953,
+      "learning_rate": 1.6039867351144778e-05,
+      "loss": 0.19,
+      "step": 6190
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.892825698089129,
+      "learning_rate": 1.6038554644514634e-05,
+      "loss": 0.2223,
+      "step": 6191
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.018159470859247,
+      "learning_rate": 1.6037241774088355e-05,
+      "loss": 0.1961,
+      "step": 6192
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5166257860517596,
+      "learning_rate": 1.6035928739901555e-05,
+      "loss": 0.2292,
+      "step": 6193
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8754496142517274,
+      "learning_rate": 1.6034615541989845e-05,
+      "loss": 0.2046,
+      "step": 6194
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2530659916913465,
+      "learning_rate": 1.603330218038885e-05,
+      "loss": 0.2013,
+      "step": 6195
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2702145828391984,
+      "learning_rate": 1.6031988655134186e-05,
+      "loss": 0.1842,
+      "step": 6196
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0243935600013727,
+      "learning_rate": 1.6030674966261496e-05,
+      "loss": 0.2063,
+      "step": 6197
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0114038595335235,
+      "learning_rate": 1.60293611138064e-05,
+      "loss": 0.2125,
+      "step": 6198
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9342042489144418,
+      "learning_rate": 1.6028047097804548e-05,
+      "loss": 0.1981,
+      "step": 6199
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0782026950345764,
+      "learning_rate": 1.6026732918291577e-05,
+      "loss": 0.171,
+      "step": 6200
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8232405471508174,
+      "learning_rate": 1.6025418575303135e-05,
+      "loss": 0.2016,
+      "step": 6201
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8404034639112508,
+      "learning_rate": 1.6024104068874877e-05,
+      "loss": 0.1885,
+      "step": 6202
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9614631039331818,
+      "learning_rate": 1.6022789399042454e-05,
+      "loss": 0.2043,
+      "step": 6203
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6367109191337368,
+      "learning_rate": 1.6021474565841527e-05,
+      "loss": 0.2171,
+      "step": 6204
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.109163775941658,
+      "learning_rate": 1.602015956930776e-05,
+      "loss": 0.1892,
+      "step": 6205
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9729785517016886,
+      "learning_rate": 1.601884440947683e-05,
+      "loss": 0.201,
+      "step": 6206
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9578508782413219,
+      "learning_rate": 1.60175290863844e-05,
+      "loss": 0.1853,
+      "step": 6207
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2746326590873274,
+      "learning_rate": 1.601621360006616e-05,
+      "loss": 0.2266,
+      "step": 6208
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8786848186524946,
+      "learning_rate": 1.601489795055778e-05,
+      "loss": 0.213,
+      "step": 6209
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9956355351416327,
+      "learning_rate": 1.6013582137894957e-05,
+      "loss": 0.1952,
+      "step": 6210
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8905803437082221,
+      "learning_rate": 1.6012266162113378e-05,
+      "loss": 0.1862,
+      "step": 6211
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.276417244394672,
+      "learning_rate": 1.601095002324874e-05,
+      "loss": 0.2076,
+      "step": 6212
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0567990624078263,
+      "learning_rate": 1.6009633721336745e-05,
+      "loss": 0.2151,
+      "step": 6213
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.103289688097003,
+      "learning_rate": 1.6008317256413092e-05,
+      "loss": 0.2407,
+      "step": 6214
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0858152536697987,
+      "learning_rate": 1.6007000628513498e-05,
+      "loss": 0.176,
+      "step": 6215
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.024115694386055,
+      "learning_rate": 1.600568383767367e-05,
+      "loss": 0.1804,
+      "step": 6216
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7915259534803418,
+      "learning_rate": 1.600436688392933e-05,
+      "loss": 0.2151,
+      "step": 6217
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9311081543873564,
+      "learning_rate": 1.6003049767316196e-05,
+      "loss": 0.2032,
+      "step": 6218
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0398814378944885,
+      "learning_rate": 1.6001732487870002e-05,
+      "loss": 0.2193,
+      "step": 6219
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7722538840376368,
+      "learning_rate": 1.6000415045626474e-05,
+      "loss": 0.1869,
+      "step": 6220
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7243802802542608,
+      "learning_rate": 1.599909744062135e-05,
+      "loss": 0.1802,
+      "step": 6221
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9532863679444862,
+      "learning_rate": 1.5997779672890367e-05,
+      "loss": 0.1937,
+      "step": 6222
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7103301266208392,
+      "learning_rate": 1.5996461742469273e-05,
+      "loss": 0.1711,
+      "step": 6223
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9093473889476353,
+      "learning_rate": 1.5995143649393814e-05,
+      "loss": 0.191,
+      "step": 6224
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9615416909069292,
+      "learning_rate": 1.5993825393699746e-05,
+      "loss": 0.2005,
+      "step": 6225
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9817798027230897,
+      "learning_rate": 1.5992506975422826e-05,
+      "loss": 0.2155,
+      "step": 6226
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0178129605623831,
+      "learning_rate": 1.5991188394598817e-05,
+      "loss": 0.2074,
+      "step": 6227
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9295853846888892,
+      "learning_rate": 1.598986965126348e-05,
+      "loss": 0.1846,
+      "step": 6228
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4018557858414185,
+      "learning_rate": 1.5988550745452593e-05,
+      "loss": 0.1982,
+      "step": 6229
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9804221302826391,
+      "learning_rate": 1.598723167720193e-05,
+      "loss": 0.2109,
+      "step": 6230
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7358355239165495,
+      "learning_rate": 1.598591244654727e-05,
+      "loss": 0.1874,
+      "step": 6231
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9962790646475322,
+      "learning_rate": 1.5984593053524395e-05,
+      "loss": 0.1918,
+      "step": 6232
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.035123937100931,
+      "learning_rate": 1.5983273498169095e-05,
+      "loss": 0.1729,
+      "step": 6233
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7762721872372287,
+      "learning_rate": 1.5981953780517166e-05,
+      "loss": 0.2003,
+      "step": 6234
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.922755801851708,
+      "learning_rate": 1.59806339006044e-05,
+      "loss": 0.1896,
+      "step": 6235
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3481005397591015,
+      "learning_rate": 1.59793138584666e-05,
+      "loss": 0.1951,
+      "step": 6236
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7834068245640107,
+      "learning_rate": 1.597799365413958e-05,
+      "loss": 0.1814,
+      "step": 6237
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1028864518367505,
+      "learning_rate": 1.597667328765914e-05,
+      "loss": 0.2181,
+      "step": 6238
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9488861711209522,
+      "learning_rate": 1.5975352759061103e-05,
+      "loss": 0.2215,
+      "step": 6239
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3137009303969078,
+      "learning_rate": 1.597403206838128e-05,
+      "loss": 0.1913,
+      "step": 6240
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0603030370685838,
+      "learning_rate": 1.5972711215655504e-05,
+      "loss": 0.2342,
+      "step": 6241
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.005721988712769,
+      "learning_rate": 1.5971390200919597e-05,
+      "loss": 0.1838,
+      "step": 6242
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9540413565559143,
+      "learning_rate": 1.5970069024209394e-05,
+      "loss": 0.2288,
+      "step": 6243
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.056861645845795,
+      "learning_rate": 1.596874768556073e-05,
+      "loss": 0.2376,
+      "step": 6244
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8608096734735928,
+      "learning_rate": 1.5967426185009448e-05,
+      "loss": 0.1994,
+      "step": 6245
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9993820475578914,
+      "learning_rate": 1.5966104522591397e-05,
+      "loss": 0.2048,
+      "step": 6246
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.052166738593178,
+      "learning_rate": 1.5964782698342423e-05,
+      "loss": 0.2037,
+      "step": 6247
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4623813432862196,
+      "learning_rate": 1.596346071229838e-05,
+      "loss": 0.2014,
+      "step": 6248
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4257348201093032,
+      "learning_rate": 1.5962138564495128e-05,
+      "loss": 0.2111,
+      "step": 6249
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8422300324358466,
+      "learning_rate": 1.596081625496853e-05,
+      "loss": 0.2061,
+      "step": 6250
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8232571074717809,
+      "learning_rate": 1.5959493783754456e-05,
+      "loss": 0.2089,
+      "step": 6251
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7018511761221633,
+      "learning_rate": 1.595817115088878e-05,
+      "loss": 0.1917,
+      "step": 6252
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9579800641887645,
+      "learning_rate": 1.5956848356407365e-05,
+      "loss": 0.2103,
+      "step": 6253
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.010412464498688,
+      "learning_rate": 1.595552540034611e-05,
+      "loss": 0.2362,
+      "step": 6254
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8006817815454528,
+      "learning_rate": 1.595420228274089e-05,
+      "loss": 0.1742,
+      "step": 6255
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9325631107213238,
+      "learning_rate": 1.59528790036276e-05,
+      "loss": 0.1993,
+      "step": 6256
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.926514554437551,
+      "learning_rate": 1.5951555563042128e-05,
+      "loss": 0.2306,
+      "step": 6257
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7324924450988743,
+      "learning_rate": 1.5950231961020373e-05,
+      "loss": 0.214,
+      "step": 6258
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0247474294843106,
+      "learning_rate": 1.5948908197598242e-05,
+      "loss": 0.1911,
+      "step": 6259
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1524455317274755,
+      "learning_rate": 1.594758427281164e-05,
+      "loss": 0.2054,
+      "step": 6260
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3710386575529927,
+      "learning_rate": 1.5946260186696477e-05,
+      "loss": 0.236,
+      "step": 6261
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.177146195681321,
+      "learning_rate": 1.5944935939288675e-05,
+      "loss": 0.1931,
+      "step": 6262
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0671464671711801,
+      "learning_rate": 1.5943611530624147e-05,
+      "loss": 0.2121,
+      "step": 6263
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9365291413240289,
+      "learning_rate": 1.594228696073882e-05,
+      "loss": 0.201,
+      "step": 6264
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9882062230623252,
+      "learning_rate": 1.5940962229668625e-05,
+      "loss": 0.2131,
+      "step": 6265
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0502799091230823,
+      "learning_rate": 1.5939637337449493e-05,
+      "loss": 0.2051,
+      "step": 6266
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.86354473763294,
+      "learning_rate": 1.5938312284117367e-05,
+      "loss": 0.1884,
+      "step": 6267
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3278074739442671,
+      "learning_rate": 1.593698706970818e-05,
+      "loss": 0.2101,
+      "step": 6268
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4480295763019384,
+      "learning_rate": 1.593566169425788e-05,
+      "loss": 0.2023,
+      "step": 6269
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2014894821170918,
+      "learning_rate": 1.5934336157802427e-05,
+      "loss": 0.2066,
+      "step": 6270
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.47506376769695,
+      "learning_rate": 1.593301046037777e-05,
+      "loss": 0.2199,
+      "step": 6271
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1430677599552572,
+      "learning_rate": 1.5931684602019866e-05,
+      "loss": 0.2121,
+      "step": 6272
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.856863370411823,
+      "learning_rate": 1.5930358582764686e-05,
+      "loss": 0.2003,
+      "step": 6273
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8649431457227784,
+      "learning_rate": 1.5929032402648194e-05,
+      "loss": 0.1998,
+      "step": 6274
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8653885888780428,
+      "learning_rate": 1.5927706061706363e-05,
+      "loss": 0.2049,
+      "step": 6275
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8381520198334308,
+      "learning_rate": 1.592637955997517e-05,
+      "loss": 0.1964,
+      "step": 6276
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1500336757509857,
+      "learning_rate": 1.5925052897490595e-05,
+      "loss": 0.1976,
+      "step": 6277
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0400983335658338,
+      "learning_rate": 1.5923726074288627e-05,
+      "loss": 0.2028,
+      "step": 6278
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7803330730926793,
+      "learning_rate": 1.5922399090405255e-05,
+      "loss": 0.1951,
+      "step": 6279
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2882797730496849,
+      "learning_rate": 1.592107194587647e-05,
+      "loss": 0.2098,
+      "step": 6280
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.949332819796093,
+      "learning_rate": 1.591974464073828e-05,
+      "loss": 0.1939,
+      "step": 6281
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8669114548502495,
+      "learning_rate": 1.591841717502668e-05,
+      "loss": 0.2039,
+      "step": 6282
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8760982488429814,
+      "learning_rate": 1.5917089548777678e-05,
+      "loss": 0.1997,
+      "step": 6283
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.425809164271548,
+      "learning_rate": 1.591576176202729e-05,
+      "loss": 0.2227,
+      "step": 6284
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0285643242186238,
+      "learning_rate": 1.5914433814811524e-05,
+      "loss": 0.2049,
+      "step": 6285
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9789355523220069,
+      "learning_rate": 1.591310570716641e-05,
+      "loss": 0.2057,
+      "step": 6286
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9412502930882807,
+      "learning_rate": 1.5911777439127975e-05,
+      "loss": 0.1977,
+      "step": 6287
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.021924695120764,
+      "learning_rate": 1.591044901073224e-05,
+      "loss": 0.2136,
+      "step": 6288
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0255290361753848,
+      "learning_rate": 1.590912042201524e-05,
+      "loss": 0.196,
+      "step": 6289
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9600707351927388,
+      "learning_rate": 1.5907791673013016e-05,
+      "loss": 0.1881,
+      "step": 6290
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2638585822881137,
+      "learning_rate": 1.5906462763761606e-05,
+      "loss": 0.1805,
+      "step": 6291
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0002196629413556,
+      "learning_rate": 1.5905133694297065e-05,
+      "loss": 0.2008,
+      "step": 6292
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0106544911549504,
+      "learning_rate": 1.5903804464655437e-05,
+      "loss": 0.2092,
+      "step": 6293
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.314786481203474,
+      "learning_rate": 1.590247507487278e-05,
+      "loss": 0.2297,
+      "step": 6294
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.112341901630231,
+      "learning_rate": 1.5901145524985155e-05,
+      "loss": 0.1985,
+      "step": 6295
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1707892567465332,
+      "learning_rate": 1.5899815815028618e-05,
+      "loss": 0.2032,
+      "step": 6296
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0298589833768144,
+      "learning_rate": 1.589848594503925e-05,
+      "loss": 0.2026,
+      "step": 6297
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1082543126402602,
+      "learning_rate": 1.5897155915053113e-05,
+      "loss": 0.2433,
+      "step": 6298
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6329212771296873,
+      "learning_rate": 1.5895825725106295e-05,
+      "loss": 0.1874,
+      "step": 6299
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1679904879141962,
+      "learning_rate": 1.5894495375234865e-05,
+      "loss": 0.2348,
+      "step": 6300
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7321960858669198,
+      "learning_rate": 1.589316486547492e-05,
+      "loss": 0.1893,
+      "step": 6301
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8250329944711491,
+      "learning_rate": 1.589183419586254e-05,
+      "loss": 0.2105,
+      "step": 6302
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8952598550183528,
+      "learning_rate": 1.5890503366433827e-05,
+      "loss": 0.217,
+      "step": 6303
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.930619450747476,
+      "learning_rate": 1.5889172377224877e-05,
+      "loss": 0.208,
+      "step": 6304
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.805703441367176,
+      "learning_rate": 1.5887841228271794e-05,
+      "loss": 0.1803,
+      "step": 6305
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1449471204861255,
+      "learning_rate": 1.5886509919610683e-05,
+      "loss": 0.2274,
+      "step": 6306
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0149187593723974,
+      "learning_rate": 1.588517845127766e-05,
+      "loss": 0.2062,
+      "step": 6307
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.517302821901913,
+      "learning_rate": 1.5883846823308843e-05,
+      "loss": 0.1797,
+      "step": 6308
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.81017351323962,
+      "learning_rate": 1.5882515035740345e-05,
+      "loss": 0.2159,
+      "step": 6309
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8318977084874352,
+      "learning_rate": 1.5881183088608293e-05,
+      "loss": 0.2093,
+      "step": 6310
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.681851519727326,
+      "learning_rate": 1.587985098194882e-05,
+      "loss": 0.1841,
+      "step": 6311
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8934676155716955,
+      "learning_rate": 1.5878518715798053e-05,
+      "loss": 0.1861,
+      "step": 6312
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9779274252743558,
+      "learning_rate": 1.587718629019214e-05,
+      "loss": 0.213,
+      "step": 6313
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9371299745261433,
+      "learning_rate": 1.5875853705167214e-05,
+      "loss": 0.193,
+      "step": 6314
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8759217719403662,
+      "learning_rate": 1.5874520960759423e-05,
+      "loss": 0.1968,
+      "step": 6315
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1231242328686373,
+      "learning_rate": 1.5873188057004924e-05,
+      "loss": 0.1896,
+      "step": 6316
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0473174398740164,
+      "learning_rate": 1.5871854993939862e-05,
+      "loss": 0.2339,
+      "step": 6317
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1763710546308124,
+      "learning_rate": 1.5870521771600402e-05,
+      "loss": 0.2014,
+      "step": 6318
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0078407380626875,
+      "learning_rate": 1.5869188390022708e-05,
+      "loss": 0.2139,
+      "step": 6319
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1132752101430765,
+      "learning_rate": 1.586785484924295e-05,
+      "loss": 0.213,
+      "step": 6320
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1250644268142815,
+      "learning_rate": 1.5866521149297294e-05,
+      "loss": 0.2078,
+      "step": 6321
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0358756499365598,
+      "learning_rate": 1.586518729022192e-05,
+      "loss": 0.2045,
+      "step": 6322
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4276322383882027,
+      "learning_rate": 1.5863853272053017e-05,
+      "loss": 0.2078,
+      "step": 6323
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1332020736682589,
+      "learning_rate": 1.5862519094826753e-05,
+      "loss": 0.1952,
+      "step": 6324
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3461541846759482,
+      "learning_rate": 1.5861184758579332e-05,
+      "loss": 0.1962,
+      "step": 6325
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9436324996098256,
+      "learning_rate": 1.5859850263346945e-05,
+      "loss": 0.1897,
+      "step": 6326
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.08030280388521,
+      "learning_rate": 1.5858515609165786e-05,
+      "loss": 0.2137,
+      "step": 6327
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2932692522911984,
+      "learning_rate": 1.5857180796072064e-05,
+      "loss": 0.1772,
+      "step": 6328
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0491984582849074,
+      "learning_rate": 1.5855845824101978e-05,
+      "loss": 0.1888,
+      "step": 6329
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.023037106386847,
+      "learning_rate": 1.5854510693291744e-05,
+      "loss": 0.1993,
+      "step": 6330
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8958489251993921,
+      "learning_rate": 1.5853175403677577e-05,
+      "loss": 0.216,
+      "step": 6331
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9890468500643292,
+      "learning_rate": 1.5851839955295697e-05,
+      "loss": 0.2089,
+      "step": 6332
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1823366094267416,
+      "learning_rate": 1.5850504348182327e-05,
+      "loss": 0.2098,
+      "step": 6333
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.427086580965463,
+      "learning_rate": 1.5849168582373698e-05,
+      "loss": 0.2417,
+      "step": 6334
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9748347836779382,
+      "learning_rate": 1.584783265790604e-05,
+      "loss": 0.2211,
+      "step": 6335
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8419382988560267,
+      "learning_rate": 1.584649657481559e-05,
+      "loss": 0.2043,
+      "step": 6336
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0180216149981338,
+      "learning_rate": 1.5845160333138593e-05,
+      "loss": 0.2057,
+      "step": 6337
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.438738728038019,
+      "learning_rate": 1.584382393291129e-05,
+      "loss": 0.2175,
+      "step": 6338
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8929441096966545,
+      "learning_rate": 1.584248737416993e-05,
+      "loss": 0.1956,
+      "step": 6339
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4119312182239425,
+      "learning_rate": 1.584115065695077e-05,
+      "loss": 0.2038,
+      "step": 6340
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9668934126268923,
+      "learning_rate": 1.583981378129007e-05,
+      "loss": 0.2106,
+      "step": 6341
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.194773506228996,
+      "learning_rate": 1.5838476747224096e-05,
+      "loss": 0.2029,
+      "step": 6342
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4189946860982916,
+      "learning_rate": 1.5837139554789106e-05,
+      "loss": 0.1852,
+      "step": 6343
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7200228788960104,
+      "learning_rate": 1.5835802204021374e-05,
+      "loss": 0.1945,
+      "step": 6344
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.049067356353654,
+      "learning_rate": 1.583446469495718e-05,
+      "loss": 0.2132,
+      "step": 6345
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8484716174067365,
+      "learning_rate": 1.5833127027632803e-05,
+      "loss": 0.1767,
+      "step": 6346
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2644673501039674,
+      "learning_rate": 1.5831789202084524e-05,
+      "loss": 0.196,
+      "step": 6347
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0563048846551186,
+      "learning_rate": 1.5830451218348634e-05,
+      "loss": 0.1971,
+      "step": 6348
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8555301237648472,
+      "learning_rate": 1.5829113076461426e-05,
+      "loss": 0.1986,
+      "step": 6349
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0299193138241485,
+      "learning_rate": 1.5827774776459195e-05,
+      "loss": 0.1942,
+      "step": 6350
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4125485894893952,
+      "learning_rate": 1.5826436318378248e-05,
+      "loss": 0.1968,
+      "step": 6351
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9133171869368052,
+      "learning_rate": 1.5825097702254885e-05,
+      "loss": 0.2082,
+      "step": 6352
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.330384605741934,
+      "learning_rate": 1.5823758928125416e-05,
+      "loss": 0.2016,
+      "step": 6353
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4065688789992765,
+      "learning_rate": 1.5822419996026156e-05,
+      "loss": 0.1982,
+      "step": 6354
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0543744183417312,
+      "learning_rate": 1.5821080905993425e-05,
+      "loss": 0.2171,
+      "step": 6355
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.010192278983979,
+      "learning_rate": 1.5819741658063548e-05,
+      "loss": 0.209,
+      "step": 6356
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1998783585968111,
+      "learning_rate": 1.5818402252272852e-05,
+      "loss": 0.2112,
+      "step": 6357
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3962202709631413,
+      "learning_rate": 1.581706268865766e-05,
+      "loss": 0.1958,
+      "step": 6358
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.154206370575021,
+      "learning_rate": 1.5815722967254317e-05,
+      "loss": 0.2085,
+      "step": 6359
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9722024390165005,
+      "learning_rate": 1.5814383088099164e-05,
+      "loss": 0.2087,
+      "step": 6360
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9929739020758829,
+      "learning_rate": 1.5813043051228536e-05,
+      "loss": 0.2216,
+      "step": 6361
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.016156213865893,
+      "learning_rate": 1.581170285667879e-05,
+      "loss": 0.1903,
+      "step": 6362
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5700458159966972,
+      "learning_rate": 1.5810362504486274e-05,
+      "loss": 0.2197,
+      "step": 6363
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1682437605649398,
+      "learning_rate": 1.5809021994687346e-05,
+      "loss": 0.1977,
+      "step": 6364
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2753786330700614,
+      "learning_rate": 1.5807681327318372e-05,
+      "loss": 0.2343,
+      "step": 6365
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.130208997823953,
+      "learning_rate": 1.580634050241571e-05,
+      "loss": 0.1902,
+      "step": 6366
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1777548703995049,
+      "learning_rate": 1.5804999520015735e-05,
+      "loss": 0.1963,
+      "step": 6367
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9107676559244275,
+      "learning_rate": 1.5803658380154822e-05,
+      "loss": 0.1923,
+      "step": 6368
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0898334399860248,
+      "learning_rate": 1.5802317082869346e-05,
+      "loss": 0.1831,
+      "step": 6369
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3729466617948922,
+      "learning_rate": 1.5800975628195692e-05,
+      "loss": 0.1964,
+      "step": 6370
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9014294103357848,
+      "learning_rate": 1.5799634016170245e-05,
+      "loss": 0.1943,
+      "step": 6371
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0305709033064328,
+      "learning_rate": 1.57982922468294e-05,
+      "loss": 0.1945,
+      "step": 6372
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.955380919591037,
+      "learning_rate": 1.5796950320209548e-05,
+      "loss": 0.2081,
+      "step": 6373
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9686561104117304,
+      "learning_rate": 1.5795608236347092e-05,
+      "loss": 0.1725,
+      "step": 6374
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9817649288066675,
+      "learning_rate": 1.5794265995278438e-05,
+      "loss": 0.2223,
+      "step": 6375
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9605643115340182,
+      "learning_rate": 1.5792923597039988e-05,
+      "loss": 0.1965,
+      "step": 6376
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9827142089808804,
+      "learning_rate": 1.579158104166816e-05,
+      "loss": 0.2116,
+      "step": 6377
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9111504054520023,
+      "learning_rate": 1.579023832919937e-05,
+      "loss": 0.1855,
+      "step": 6378
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2261708996905023,
+      "learning_rate": 1.5788895459670036e-05,
+      "loss": 0.2072,
+      "step": 6379
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2240954515580862,
+      "learning_rate": 1.5787552433116586e-05,
+      "loss": 0.2059,
+      "step": 6380
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9802222648434512,
+      "learning_rate": 1.5786209249575452e-05,
+      "loss": 0.2064,
+      "step": 6381
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4678277710491507,
+      "learning_rate": 1.5784865909083064e-05,
+      "loss": 0.2009,
+      "step": 6382
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0693945382145387,
+      "learning_rate": 1.5783522411675857e-05,
+      "loss": 0.2025,
+      "step": 6383
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5394552546625184,
+      "learning_rate": 1.5782178757390284e-05,
+      "loss": 0.1903,
+      "step": 6384
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0743984781457123,
+      "learning_rate": 1.5780834946262782e-05,
+      "loss": 0.2184,
+      "step": 6385
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.585305608844427,
+      "learning_rate": 1.5779490978329806e-05,
+      "loss": 0.2141,
+      "step": 6386
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9386347211553212,
+      "learning_rate": 1.5778146853627813e-05,
+      "loss": 0.1865,
+      "step": 6387
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0530458992126261,
+      "learning_rate": 1.577680257219326e-05,
+      "loss": 0.1897,
+      "step": 6388
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2499711672577036,
+      "learning_rate": 1.577545813406261e-05,
+      "loss": 0.2078,
+      "step": 6389
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.127326725810852,
+      "learning_rate": 1.5774113539272332e-05,
+      "loss": 0.2238,
+      "step": 6390
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3929687223303266,
+      "learning_rate": 1.57727687878589e-05,
+      "loss": 0.1949,
+      "step": 6391
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9969779103098971,
+      "learning_rate": 1.5771423879858783e-05,
+      "loss": 0.1882,
+      "step": 6392
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.41306573541083,
+      "learning_rate": 1.5770078815308473e-05,
+      "loss": 0.1869,
+      "step": 6393
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.838933203170488,
+      "learning_rate": 1.5768733594244445e-05,
+      "loss": 0.2058,
+      "step": 6394
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0659843745048814,
+      "learning_rate": 1.5767388216703196e-05,
+      "loss": 0.2124,
+      "step": 6395
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.152946095323606,
+      "learning_rate": 1.5766042682721213e-05,
+      "loss": 0.2095,
+      "step": 6396
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0228491280513095,
+      "learning_rate": 1.5764696992335e-05,
+      "loss": 0.2322,
+      "step": 6397
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3934301752230933,
+      "learning_rate": 1.5763351145581052e-05,
+      "loss": 0.2031,
+      "step": 6398
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0083370881264966,
+      "learning_rate": 1.576200514249588e-05,
+      "loss": 0.217,
+      "step": 6399
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4430508995785853,
+      "learning_rate": 1.576065898311599e-05,
+      "loss": 0.1991,
+      "step": 6400
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6340672012097428,
+      "learning_rate": 1.5759312667477907e-05,
+      "loss": 0.1983,
+      "step": 6401
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6808778569379283,
+      "learning_rate": 1.575796619561814e-05,
+      "loss": 0.1861,
+      "step": 6402
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2748513907509549,
+      "learning_rate": 1.5756619567573213e-05,
+      "loss": 0.2164,
+      "step": 6403
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.936128174627036,
+      "learning_rate": 1.5755272783379656e-05,
+      "loss": 0.1953,
+      "step": 6404
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2888785381934886,
+      "learning_rate": 1.5753925843073997e-05,
+      "loss": 0.2246,
+      "step": 6405
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.078888056580051,
+      "learning_rate": 1.5752578746692776e-05,
+      "loss": 0.1997,
+      "step": 6406
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9884413373743257,
+      "learning_rate": 1.5751231494272535e-05,
+      "loss": 0.1972,
+      "step": 6407
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.882073537819897,
+      "learning_rate": 1.574988408584981e-05,
+      "loss": 0.2124,
+      "step": 6408
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8856297252603791,
+      "learning_rate": 1.574853652146116e-05,
+      "loss": 0.2084,
+      "step": 6409
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9096131648444687,
+      "learning_rate": 1.5747188801143126e-05,
+      "loss": 0.2104,
+      "step": 6410
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1483305633397578,
+      "learning_rate": 1.574584092493227e-05,
+      "loss": 0.2229,
+      "step": 6411
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2905589927471228,
+      "learning_rate": 1.574449289286516e-05,
+      "loss": 0.1895,
+      "step": 6412
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.055202980082456,
+      "learning_rate": 1.5743144704978358e-05,
+      "loss": 0.2162,
+      "step": 6413
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.571357528956249,
+      "learning_rate": 1.5741796361308424e-05,
+      "loss": 0.2185,
+      "step": 6414
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.27518127262942,
+      "learning_rate": 1.5740447861891946e-05,
+      "loss": 0.2178,
+      "step": 6415
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1005877561884827,
+      "learning_rate": 1.573909920676549e-05,
+      "loss": 0.2353,
+      "step": 6416
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8713251975400061,
+      "learning_rate": 1.5737750395965646e-05,
+      "loss": 0.1908,
+      "step": 6417
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2112967064517417,
+      "learning_rate": 1.5736401429529e-05,
+      "loss": 0.1913,
+      "step": 6418
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9991676495662414,
+      "learning_rate": 1.573505230749214e-05,
+      "loss": 0.1855,
+      "step": 6419
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8813185706705601,
+      "learning_rate": 1.573370302989166e-05,
+      "loss": 0.211,
+      "step": 6420
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2830617593144382,
+      "learning_rate": 1.5732353596764168e-05,
+      "loss": 0.19,
+      "step": 6421
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3583092644558936,
+      "learning_rate": 1.5731004008146255e-05,
+      "loss": 0.1827,
+      "step": 6422
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0762212789850942,
+      "learning_rate": 1.5729654264074536e-05,
+      "loss": 0.1934,
+      "step": 6423
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2099248654148724,
+      "learning_rate": 1.572830436458562e-05,
+      "loss": 0.1947,
+      "step": 6424
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.592646821803759,
+      "learning_rate": 1.5726954309716128e-05,
+      "loss": 0.1849,
+      "step": 6425
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0505585309130414,
+      "learning_rate": 1.5725604099502673e-05,
+      "loss": 0.2032,
+      "step": 6426
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5026472365739723,
+      "learning_rate": 1.572425373398189e-05,
+      "loss": 0.1994,
+      "step": 6427
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8287932239533964,
+      "learning_rate": 1.5722903213190393e-05,
+      "loss": 0.1654,
+      "step": 6428
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0151895468230154,
+      "learning_rate": 1.572155253716483e-05,
+      "loss": 0.1894,
+      "step": 6429
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.813161979459579,
+      "learning_rate": 1.5720201705941827e-05,
+      "loss": 0.2153,
+      "step": 6430
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4623926629210309,
+      "learning_rate": 1.571885071955803e-05,
+      "loss": 0.1959,
+      "step": 6431
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.197393529509465,
+      "learning_rate": 1.5717499578050083e-05,
+      "loss": 0.2121,
+      "step": 6432
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8507227862656661,
+      "learning_rate": 1.571614828145464e-05,
+      "loss": 0.1791,
+      "step": 6433
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9143042226597085,
+      "learning_rate": 1.5714796829808352e-05,
+      "loss": 0.2176,
+      "step": 6434
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0246240592941134,
+      "learning_rate": 1.5713445223147876e-05,
+      "loss": 0.2225,
+      "step": 6435
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2215280314440582,
+      "learning_rate": 1.5712093461509878e-05,
+      "loss": 0.2066,
+      "step": 6436
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.947008583860387,
+      "learning_rate": 1.5710741544931017e-05,
+      "loss": 0.2278,
+      "step": 6437
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1323169093199634,
+      "learning_rate": 1.5709389473447974e-05,
+      "loss": 0.2134,
+      "step": 6438
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8026527983512476,
+      "learning_rate": 1.5708037247097418e-05,
+      "loss": 0.2012,
+      "step": 6439
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1969855479924654,
+      "learning_rate": 1.5706684865916025e-05,
+      "loss": 0.2282,
+      "step": 6440
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9403857301981011,
+      "learning_rate": 1.570533232994049e-05,
+      "loss": 0.1967,
+      "step": 6441
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.305053539772037,
+      "learning_rate": 1.570397963920749e-05,
+      "loss": 0.1965,
+      "step": 6442
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1363241248573346,
+      "learning_rate": 1.5702626793753717e-05,
+      "loss": 0.2044,
+      "step": 6443
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8737256816832991,
+      "learning_rate": 1.5701273793615876e-05,
+      "loss": 0.2048,
+      "step": 6444
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6215538187724758,
+      "learning_rate": 1.5699920638830656e-05,
+      "loss": 0.218,
+      "step": 6445
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0878235244057186,
+      "learning_rate": 1.5698567329434768e-05,
+      "loss": 0.222,
+      "step": 6446
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0513041461655643,
+      "learning_rate": 1.569721386546492e-05,
+      "loss": 0.2166,
+      "step": 6447
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9622985569927587,
+      "learning_rate": 1.5695860246957826e-05,
+      "loss": 0.1979,
+      "step": 6448
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.106124143025721,
+      "learning_rate": 1.5694506473950198e-05,
+      "loss": 0.2146,
+      "step": 6449
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9022371205892392,
+      "learning_rate": 1.5693152546478762e-05,
+      "loss": 0.1904,
+      "step": 6450
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8059048062766315,
+      "learning_rate": 1.569179846458024e-05,
+      "loss": 0.1933,
+      "step": 6451
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8776693830331572,
+      "learning_rate": 1.5690444228291366e-05,
+      "loss": 0.2228,
+      "step": 6452
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0309478566891377,
+      "learning_rate": 1.568908983764887e-05,
+      "loss": 0.1922,
+      "step": 6453
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.050309678890914,
+      "learning_rate": 1.568773529268949e-05,
+      "loss": 0.2045,
+      "step": 6454
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0229060671888315,
+      "learning_rate": 1.5686380593449966e-05,
+      "loss": 0.2047,
+      "step": 6455
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.001874802951346,
+      "learning_rate": 1.5685025739967056e-05,
+      "loss": 0.229,
+      "step": 6456
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9513827785300544,
+      "learning_rate": 1.568367073227749e-05,
+      "loss": 0.1857,
+      "step": 6457
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8963176883806334,
+      "learning_rate": 1.5682315570418043e-05,
+      "loss": 0.1848,
+      "step": 6458
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9336871285878903,
+      "learning_rate": 1.5680960254425467e-05,
+      "loss": 0.2199,
+      "step": 6459
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9301503405854965,
+      "learning_rate": 1.5679604784336516e-05,
+      "loss": 0.1947,
+      "step": 6460
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8430696671678639,
+      "learning_rate": 1.567824916018797e-05,
+      "loss": 0.2281,
+      "step": 6461
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0779499877725116,
+      "learning_rate": 1.567689338201659e-05,
+      "loss": 0.2232,
+      "step": 6462
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9323416339337731,
+      "learning_rate": 1.5675537449859158e-05,
+      "loss": 0.2095,
+      "step": 6463
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1019950763615698,
+      "learning_rate": 1.5674181363752452e-05,
+      "loss": 0.1894,
+      "step": 6464
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2348306315579887,
+      "learning_rate": 1.5672825123733257e-05,
+      "loss": 0.1921,
+      "step": 6465
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0231332496881906,
+      "learning_rate": 1.567146872983836e-05,
+      "loss": 0.1972,
+      "step": 6466
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.220781301600539,
+      "learning_rate": 1.5670112182104552e-05,
+      "loss": 0.1978,
+      "step": 6467
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1006520137959355,
+      "learning_rate": 1.566875548056863e-05,
+      "loss": 0.2557,
+      "step": 6468
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0133967400662975,
+      "learning_rate": 1.5667398625267402e-05,
+      "loss": 0.1999,
+      "step": 6469
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9847310910093158,
+      "learning_rate": 1.566604161623766e-05,
+      "loss": 0.1787,
+      "step": 6470
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3193309942941929,
+      "learning_rate": 1.5664684453516218e-05,
+      "loss": 0.2028,
+      "step": 6471
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0438169210754187,
+      "learning_rate": 1.5663327137139893e-05,
+      "loss": 0.1959,
+      "step": 6472
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9198281649456738,
+      "learning_rate": 1.56619696671455e-05,
+      "loss": 0.2137,
+      "step": 6473
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3123289776496023,
+      "learning_rate": 1.5660612043569864e-05,
+      "loss": 0.2134,
+      "step": 6474
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9629029433242554,
+      "learning_rate": 1.56592542664498e-05,
+      "loss": 0.1986,
+      "step": 6475
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0251036113371328,
+      "learning_rate": 1.5657896335822147e-05,
+      "loss": 0.1771,
+      "step": 6476
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8869554671856225,
+      "learning_rate": 1.5656538251723734e-05,
+      "loss": 0.1902,
+      "step": 6477
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3756739947044674,
+      "learning_rate": 1.5655180014191404e-05,
+      "loss": 0.2119,
+      "step": 6478
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9633315984036738,
+      "learning_rate": 1.5653821623261998e-05,
+      "loss": 0.2,
+      "step": 6479
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.251828938015827,
+      "learning_rate": 1.565246307897236e-05,
+      "loss": 0.203,
+      "step": 6480
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.14068294685859,
+      "learning_rate": 1.565110438135934e-05,
+      "loss": 0.191,
+      "step": 6481
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5452138562508575,
+      "learning_rate": 1.5649745530459794e-05,
+      "loss": 0.1995,
+      "step": 6482
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0536420677133722,
+      "learning_rate": 1.5648386526310582e-05,
+      "loss": 0.1993,
+      "step": 6483
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1356216475216754,
+      "learning_rate": 1.564702736894857e-05,
+      "loss": 0.2186,
+      "step": 6484
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8532781011175651,
+      "learning_rate": 1.5645668058410617e-05,
+      "loss": 0.2082,
+      "step": 6485
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8455231181313126,
+      "learning_rate": 1.56443085947336e-05,
+      "loss": 0.1931,
+      "step": 6486
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2629784905997012,
+      "learning_rate": 1.5642948977954395e-05,
+      "loss": 0.2528,
+      "step": 6487
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2047458771536588,
+      "learning_rate": 1.564158920810988e-05,
+      "loss": 0.2098,
+      "step": 6488
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.154183894969296,
+      "learning_rate": 1.5640229285236938e-05,
+      "loss": 0.2142,
+      "step": 6489
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4439578060813603,
+      "learning_rate": 1.563886920937246e-05,
+      "loss": 0.1939,
+      "step": 6490
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1269334348934161,
+      "learning_rate": 1.5637508980553335e-05,
+      "loss": 0.202,
+      "step": 6491
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8474971196663091,
+      "learning_rate": 1.563614859881646e-05,
+      "loss": 0.1944,
+      "step": 6492
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.775445743159406,
+      "learning_rate": 1.5634788064198736e-05,
+      "loss": 0.1838,
+      "step": 6493
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8829878157374699,
+      "learning_rate": 1.5633427376737072e-05,
+      "loss": 0.2099,
+      "step": 6494
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.388770427746395,
+      "learning_rate": 1.5632066536468367e-05,
+      "loss": 0.1908,
+      "step": 6495
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0099467639292705,
+      "learning_rate": 1.5630705543429542e-05,
+      "loss": 0.191,
+      "step": 6496
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9404162366537719,
+      "learning_rate": 1.5629344397657506e-05,
+      "loss": 0.194,
+      "step": 6497
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9446683305039197,
+      "learning_rate": 1.562798309918919e-05,
+      "loss": 0.1796,
+      "step": 6498
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9309802969044433,
+      "learning_rate": 1.5626621648061514e-05,
+      "loss": 0.202,
+      "step": 6499
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7207205399944414,
+      "learning_rate": 1.5625260044311405e-05,
+      "loss": 0.2094,
+      "step": 6500
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8054028028686318,
+      "learning_rate": 1.5623898287975806e-05,
+      "loss": 0.204,
+      "step": 6501
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4060494249650275,
+      "learning_rate": 1.562253637909164e-05,
+      "loss": 0.194,
+      "step": 6502
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8105266496711662,
+      "learning_rate": 1.5621174317695862e-05,
+      "loss": 0.2025,
+      "step": 6503
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8944536557109091,
+      "learning_rate": 1.561981210382541e-05,
+      "loss": 0.2159,
+      "step": 6504
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7844477751552985,
+      "learning_rate": 1.5618449737517242e-05,
+      "loss": 0.2015,
+      "step": 6505
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8709928078499433,
+      "learning_rate": 1.5617087218808307e-05,
+      "loss": 0.1814,
+      "step": 6506
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3657649463539294,
+      "learning_rate": 1.5615724547735562e-05,
+      "loss": 0.2111,
+      "step": 6507
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3173488373145754,
+      "learning_rate": 1.561436172433597e-05,
+      "loss": 0.1981,
+      "step": 6508
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.078990856080944,
+      "learning_rate": 1.56129987486465e-05,
+      "loss": 0.2151,
+      "step": 6509
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8717188418025383,
+      "learning_rate": 1.5611635620704128e-05,
+      "loss": 0.1983,
+      "step": 6510
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8053767630345547,
+      "learning_rate": 1.5610272340545814e-05,
+      "loss": 0.1947,
+      "step": 6511
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.842561014891099,
+      "learning_rate": 1.560890890820855e-05,
+      "loss": 0.1883,
+      "step": 6512
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1991401747191601,
+      "learning_rate": 1.5607545323729313e-05,
+      "loss": 0.1756,
+      "step": 6513
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.059056198298021,
+      "learning_rate": 1.5606181587145097e-05,
+      "loss": 0.2366,
+      "step": 6514
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7460856086533205,
+      "learning_rate": 1.5604817698492886e-05,
+      "loss": 0.2369,
+      "step": 6515
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0681687279055996,
+      "learning_rate": 1.560345365780968e-05,
+      "loss": 0.2114,
+      "step": 6516
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.240767782666872,
+      "learning_rate": 1.5602089465132474e-05,
+      "loss": 0.197,
+      "step": 6517
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6283348417848393,
+      "learning_rate": 1.5600725120498273e-05,
+      "loss": 0.1867,
+      "step": 6518
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8753049390892731,
+      "learning_rate": 1.5599360623944092e-05,
+      "loss": 0.198,
+      "step": 6519
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8253790374894722,
+      "learning_rate": 1.5597995975506936e-05,
+      "loss": 0.2095,
+      "step": 6520
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.679152258436176,
+      "learning_rate": 1.5596631175223823e-05,
+      "loss": 0.2221,
+      "step": 6521
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.363989379907467,
+      "learning_rate": 1.559526622313177e-05,
+      "loss": 0.1862,
+      "step": 6522
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.99462758090583,
+      "learning_rate": 1.559390111926781e-05,
+      "loss": 0.2021,
+      "step": 6523
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.105152738508639,
+      "learning_rate": 1.559253586366896e-05,
+      "loss": 0.2166,
+      "step": 6524
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8798114773050834,
+      "learning_rate": 1.5591170456372264e-05,
+      "loss": 0.2005,
+      "step": 6525
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2415787362049853,
+      "learning_rate": 1.5589804897414757e-05,
+      "loss": 0.2013,
+      "step": 6526
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1061386725225641,
+      "learning_rate": 1.5588439186833467e-05,
+      "loss": 0.2422,
+      "step": 6527
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7700214560139843,
+      "learning_rate": 1.5587073324665457e-05,
+      "loss": 0.197,
+      "step": 6528
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3002606779597166,
+      "learning_rate": 1.558570731094776e-05,
+      "loss": 0.2082,
+      "step": 6529
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.120351330109009,
+      "learning_rate": 1.558434114571744e-05,
+      "loss": 0.202,
+      "step": 6530
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4396268572009103,
+      "learning_rate": 1.558297482901155e-05,
+      "loss": 0.2189,
+      "step": 6531
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0460960319227817,
+      "learning_rate": 1.5581608360867154e-05,
+      "loss": 0.217,
+      "step": 6532
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0874166588094762,
+      "learning_rate": 1.5580241741321317e-05,
+      "loss": 0.2019,
+      "step": 6533
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1894921553008606,
+      "learning_rate": 1.5578874970411105e-05,
+      "loss": 0.1977,
+      "step": 6534
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9151050298336578,
+      "learning_rate": 1.5577508048173596e-05,
+      "loss": 0.1988,
+      "step": 6535
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8028004085361633,
+      "learning_rate": 1.5576140974645868e-05,
+      "loss": 0.2192,
+      "step": 6536
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0884779279303818,
+      "learning_rate": 1.5574773749865e-05,
+      "loss": 0.2029,
+      "step": 6537
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8143331503528025,
+      "learning_rate": 1.5573406373868077e-05,
+      "loss": 0.2073,
+      "step": 6538
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0414387222930812,
+      "learning_rate": 1.5572038846692193e-05,
+      "loss": 0.1941,
+      "step": 6539
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4755430457912222,
+      "learning_rate": 1.557067116837444e-05,
+      "loss": 0.1945,
+      "step": 6540
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.063393639956194,
+      "learning_rate": 1.5569303338951914e-05,
+      "loss": 0.2216,
+      "step": 6541
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.354543793682284,
+      "learning_rate": 1.5567935358461724e-05,
+      "loss": 0.22,
+      "step": 6542
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1240307573927577,
+      "learning_rate": 1.5566567226940974e-05,
+      "loss": 0.2162,
+      "step": 6543
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8625214215267929,
+      "learning_rate": 1.556519894442677e-05,
+      "loss": 0.207,
+      "step": 6544
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9190977071427817,
+      "learning_rate": 1.5563830510956234e-05,
+      "loss": 0.1947,
+      "step": 6545
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7671693867274025,
+      "learning_rate": 1.556246192656648e-05,
+      "loss": 0.1928,
+      "step": 6546
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9212945113465504,
+      "learning_rate": 1.556109319129463e-05,
+      "loss": 0.2034,
+      "step": 6547
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7883832295181457,
+      "learning_rate": 1.5559724305177814e-05,
+      "loss": 0.217,
+      "step": 6548
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5995930916405159,
+      "learning_rate": 1.5558355268253166e-05,
+      "loss": 0.1975,
+      "step": 6549
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8519660979050935,
+      "learning_rate": 1.555698608055781e-05,
+      "loss": 0.195,
+      "step": 6550
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1385807258455753,
+      "learning_rate": 1.5555616742128897e-05,
+      "loss": 0.1953,
+      "step": 6551
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8376325527210272,
+      "learning_rate": 1.5554247253003567e-05,
+      "loss": 0.1755,
+      "step": 6552
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.790951262306196,
+      "learning_rate": 1.5552877613218964e-05,
+      "loss": 0.1992,
+      "step": 6553
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0898402865456223,
+      "learning_rate": 1.555150782281224e-05,
+      "loss": 0.1971,
+      "step": 6554
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0067994486256289,
+      "learning_rate": 1.555013788182056e-05,
+      "loss": 0.2048,
+      "step": 6555
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.205884701347545,
+      "learning_rate": 1.554876779028107e-05,
+      "loss": 0.2196,
+      "step": 6556
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.322602602161279,
+      "learning_rate": 1.5547397548230943e-05,
+      "loss": 0.2039,
+      "step": 6557
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7720006927722426,
+      "learning_rate": 1.554602715570735e-05,
+      "loss": 0.186,
+      "step": 6558
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0980304504859055,
+      "learning_rate": 1.554465661274745e-05,
+      "loss": 0.2072,
+      "step": 6559
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0093278518246684,
+      "learning_rate": 1.5543285919388426e-05,
+      "loss": 0.2197,
+      "step": 6560
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2019674945560757,
+      "learning_rate": 1.554191507566746e-05,
+      "loss": 0.1965,
+      "step": 6561
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9670683442046764,
+      "learning_rate": 1.5540544081621736e-05,
+      "loss": 0.2288,
+      "step": 6562
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.949605601109131,
+      "learning_rate": 1.5539172937288437e-05,
+      "loss": 0.1993,
+      "step": 6563
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9442795776353967,
+      "learning_rate": 1.5537801642704763e-05,
+      "loss": 0.2031,
+      "step": 6564
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0005769850834774,
+      "learning_rate": 1.5536430197907904e-05,
+      "loss": 0.2164,
+      "step": 6565
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9143813378018446,
+      "learning_rate": 1.5535058602935065e-05,
+      "loss": 0.1804,
+      "step": 6566
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8805621406735882,
+      "learning_rate": 1.5533686857823447e-05,
+      "loss": 0.2052,
+      "step": 6567
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9730532725900689,
+      "learning_rate": 1.5532314962610263e-05,
+      "loss": 0.2081,
+      "step": 6568
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1364802182104736,
+      "learning_rate": 1.553094291733272e-05,
+      "loss": 0.2015,
+      "step": 6569
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.95822533473699,
+      "learning_rate": 1.552957072202804e-05,
+      "loss": 0.2093,
+      "step": 6570
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3920492379804228,
+      "learning_rate": 1.5528198376733444e-05,
+      "loss": 0.1986,
+      "step": 6571
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8671365676659566,
+      "learning_rate": 1.552682588148615e-05,
+      "loss": 0.2204,
+      "step": 6572
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8180428446337018,
+      "learning_rate": 1.5525453236323396e-05,
+      "loss": 0.1759,
+      "step": 6573
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0814408848789787,
+      "learning_rate": 1.5524080441282408e-05,
+      "loss": 0.1984,
+      "step": 6574
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8779769686721743,
+      "learning_rate": 1.5522707496400425e-05,
+      "loss": 0.1806,
+      "step": 6575
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9724116612781548,
+      "learning_rate": 1.5521334401714692e-05,
+      "loss": 0.1971,
+      "step": 6576
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.916346755900272,
+      "learning_rate": 1.551996115726245e-05,
+      "loss": 0.1832,
+      "step": 6577
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2003326335004167,
+      "learning_rate": 1.5518587763080956e-05,
+      "loss": 0.2176,
+      "step": 6578
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1705118254184124,
+      "learning_rate": 1.551721421920745e-05,
+      "loss": 0.194,
+      "step": 6579
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0446029535231542,
+      "learning_rate": 1.55158405256792e-05,
+      "loss": 0.2208,
+      "step": 6580
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0562633998936752,
+      "learning_rate": 1.551446668253346e-05,
+      "loss": 0.2092,
+      "step": 6581
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8909198049009172,
+      "learning_rate": 1.5513092689807505e-05,
+      "loss": 0.2235,
+      "step": 6582
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1344678807289725,
+      "learning_rate": 1.5511718547538596e-05,
+      "loss": 0.1994,
+      "step": 6583
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1833341979307461,
+      "learning_rate": 1.551034425576401e-05,
+      "loss": 0.1932,
+      "step": 6584
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9294885641880941,
+      "learning_rate": 1.5508969814521026e-05,
+      "loss": 0.2624,
+      "step": 6585
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9262316728429739,
+      "learning_rate": 1.550759522384693e-05,
+      "loss": 0.2206,
+      "step": 6586
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.793394516959776,
+      "learning_rate": 1.5506220483778994e-05,
+      "loss": 0.1744,
+      "step": 6587
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0311135703428027,
+      "learning_rate": 1.550484559435452e-05,
+      "loss": 0.2464,
+      "step": 6588
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1685635601265862,
+      "learning_rate": 1.5503470555610797e-05,
+      "loss": 0.2192,
+      "step": 6589
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2425139947073716,
+      "learning_rate": 1.5502095367585124e-05,
+      "loss": 0.2053,
+      "step": 6590
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1825964827556712,
+      "learning_rate": 1.5500720030314805e-05,
+      "loss": 0.2195,
+      "step": 6591
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0157324689975242,
+      "learning_rate": 1.5499344543837144e-05,
+      "loss": 0.2074,
+      "step": 6592
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1504316423024223,
+      "learning_rate": 1.549796890818945e-05,
+      "loss": 0.2221,
+      "step": 6593
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1223861108775626,
+      "learning_rate": 1.5496593123409042e-05,
+      "loss": 0.1861,
+      "step": 6594
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5017109310443328,
+      "learning_rate": 1.549521718953323e-05,
+      "loss": 0.1979,
+      "step": 6595
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9587162726326114,
+      "learning_rate": 1.549384110659935e-05,
+      "loss": 0.2071,
+      "step": 6596
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4159051199900652,
+      "learning_rate": 1.5492464874644713e-05,
+      "loss": 0.2177,
+      "step": 6597
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.468995929268174,
+      "learning_rate": 1.5491088493706657e-05,
+      "loss": 0.2205,
+      "step": 6598
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1113322064440028,
+      "learning_rate": 1.548971196382252e-05,
+      "loss": 0.1878,
+      "step": 6599
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1024645553263197,
+      "learning_rate": 1.548833528502963e-05,
+      "loss": 0.2177,
+      "step": 6600
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9146618497170281,
+      "learning_rate": 1.5486958457365338e-05,
+      "loss": 0.1983,
+      "step": 6601
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.004880063563026,
+      "learning_rate": 1.5485581480866985e-05,
+      "loss": 0.1895,
+      "step": 6602
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8620626063742314,
+      "learning_rate": 1.5484204355571927e-05,
+      "loss": 0.1999,
+      "step": 6603
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6230596374204962,
+      "learning_rate": 1.5482827081517516e-05,
+      "loss": 0.2073,
+      "step": 6604
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8772018169236586,
+      "learning_rate": 1.5481449658741112e-05,
+      "loss": 0.1892,
+      "step": 6605
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9995458502864405,
+      "learning_rate": 1.5480072087280075e-05,
+      "loss": 0.1928,
+      "step": 6606
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1101682764237155,
+      "learning_rate": 1.5478694367171772e-05,
+      "loss": 0.1975,
+      "step": 6607
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1474296693817685,
+      "learning_rate": 1.547731649845358e-05,
+      "loss": 0.2093,
+      "step": 6608
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0708314496289437,
+      "learning_rate": 1.5475938481162862e-05,
+      "loss": 0.215,
+      "step": 6609
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4647452705106907,
+      "learning_rate": 1.5474560315337007e-05,
+      "loss": 0.2148,
+      "step": 6610
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0394882698957975,
+      "learning_rate": 1.5473182001013394e-05,
+      "loss": 0.18,
+      "step": 6611
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.177655440718952,
+      "learning_rate": 1.547180353822941e-05,
+      "loss": 0.204,
+      "step": 6612
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6098661456952317,
+      "learning_rate": 1.5470424927022442e-05,
+      "loss": 0.2157,
+      "step": 6613
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4013179662489288,
+      "learning_rate": 1.5469046167429895e-05,
+      "loss": 0.1768,
+      "step": 6614
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8876557891047067,
+      "learning_rate": 1.5467667259489157e-05,
+      "loss": 0.1968,
+      "step": 6615
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7364338213122036,
+      "learning_rate": 1.546628820323764e-05,
+      "loss": 0.1951,
+      "step": 6616
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8233505032421118,
+      "learning_rate": 1.5464908998712743e-05,
+      "loss": 0.2177,
+      "step": 6617
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7988196118879016,
+      "learning_rate": 1.5463529645951884e-05,
+      "loss": 0.1774,
+      "step": 6618
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7863804154493699,
+      "learning_rate": 1.5462150144992473e-05,
+      "loss": 0.2375,
+      "step": 6619
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1876939499505215,
+      "learning_rate": 1.546077049587193e-05,
+      "loss": 0.1965,
+      "step": 6620
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7738270979735855,
+      "learning_rate": 1.545939069862768e-05,
+      "loss": 0.2106,
+      "step": 6621
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8918711817915994,
+      "learning_rate": 1.545801075329715e-05,
+      "loss": 0.1929,
+      "step": 6622
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9377869351034032,
+      "learning_rate": 1.5456630659917768e-05,
+      "loss": 0.1894,
+      "step": 6623
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8214995234734142,
+      "learning_rate": 1.5455250418526976e-05,
+      "loss": 0.2025,
+      "step": 6624
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9899732290557761,
+      "learning_rate": 1.5453870029162202e-05,
+      "loss": 0.1985,
+      "step": 6625
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0791010808495334,
+      "learning_rate": 1.5452489491860897e-05,
+      "loss": 0.2171,
+      "step": 6626
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8035978835832673,
+      "learning_rate": 1.5451108806660508e-05,
+      "loss": 0.1964,
+      "step": 6627
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8693180246456279,
+      "learning_rate": 1.5449727973598487e-05,
+      "loss": 0.1976,
+      "step": 6628
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8965073835730478,
+      "learning_rate": 1.544834699271228e-05,
+      "loss": 0.2127,
+      "step": 6629
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1466026109131418,
+      "learning_rate": 1.5446965864039357e-05,
+      "loss": 0.1774,
+      "step": 6630
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9670915922502352,
+      "learning_rate": 1.544558458761718e-05,
+      "loss": 0.2116,
+      "step": 6631
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3603515418339192,
+      "learning_rate": 1.5444203163483212e-05,
+      "loss": 0.2025,
+      "step": 6632
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9222393157039742,
+      "learning_rate": 1.544282159167492e-05,
+      "loss": 0.2049,
+      "step": 6633
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9098251873140276,
+      "learning_rate": 1.5441439872229793e-05,
+      "loss": 0.1982,
+      "step": 6634
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0500483855202372,
+      "learning_rate": 1.5440058005185295e-05,
+      "loss": 0.219,
+      "step": 6635
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9415410885951533,
+      "learning_rate": 1.5438675990578923e-05,
+      "loss": 0.1917,
+      "step": 6636
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9378779126221987,
+      "learning_rate": 1.5437293828448153e-05,
+      "loss": 0.2228,
+      "step": 6637
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9004931303198556,
+      "learning_rate": 1.5435911518830485e-05,
+      "loss": 0.2029,
+      "step": 6638
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.782645150359459,
+      "learning_rate": 1.5434529061763405e-05,
+      "loss": 0.1996,
+      "step": 6639
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.004324869305781,
+      "learning_rate": 1.543314645728442e-05,
+      "loss": 0.2044,
+      "step": 6640
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6039110715845524,
+      "learning_rate": 1.543176370543103e-05,
+      "loss": 0.2098,
+      "step": 6641
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2148324779559043,
+      "learning_rate": 1.5430380806240744e-05,
+      "loss": 0.2267,
+      "step": 6642
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9685722004536197,
+      "learning_rate": 1.5428997759751073e-05,
+      "loss": 0.1986,
+      "step": 6643
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4287506123498237,
+      "learning_rate": 1.5427614565999527e-05,
+      "loss": 0.1698,
+      "step": 6644
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2560403408390468,
+      "learning_rate": 1.542623122502363e-05,
+      "loss": 0.2371,
+      "step": 6645
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.493029626176336,
+      "learning_rate": 1.5424847736860907e-05,
+      "loss": 0.2216,
+      "step": 6646
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.502436700329338,
+      "learning_rate": 1.5423464101548883e-05,
+      "loss": 0.217,
+      "step": 6647
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7559959460499155,
+      "learning_rate": 1.5422080319125085e-05,
+      "loss": 0.2062,
+      "step": 6648
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0631140913339923,
+      "learning_rate": 1.5420696389627057e-05,
+      "loss": 0.2386,
+      "step": 6649
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.224313506950377,
+      "learning_rate": 1.5419312313092328e-05,
+      "loss": 0.1964,
+      "step": 6650
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3134515484754445,
+      "learning_rate": 1.541792808955845e-05,
+      "loss": 0.1774,
+      "step": 6651
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9334696811383187,
+      "learning_rate": 1.5416543719062967e-05,
+      "loss": 0.1961,
+      "step": 6652
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.177362982763627,
+      "learning_rate": 1.541515920164343e-05,
+      "loss": 0.2205,
+      "step": 6653
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2813984057128454,
+      "learning_rate": 1.541377453733739e-05,
+      "loss": 0.1877,
+      "step": 6654
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8656007817415847,
+      "learning_rate": 1.541238972618241e-05,
+      "loss": 0.1998,
+      "step": 6655
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0367298136556913,
+      "learning_rate": 1.541100476821606e-05,
+      "loss": 0.2136,
+      "step": 6656
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9975140810289721,
+      "learning_rate": 1.5409619663475894e-05,
+      "loss": 0.2154,
+      "step": 6657
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2458834798559473,
+      "learning_rate": 1.540823441199949e-05,
+      "loss": 0.193,
+      "step": 6658
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5142136985614558,
+      "learning_rate": 1.540684901382442e-05,
+      "loss": 0.2157,
+      "step": 6659
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8924500791959433,
+      "learning_rate": 1.540546346898827e-05,
+      "loss": 0.2003,
+      "step": 6660
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9653697735047038,
+      "learning_rate": 1.5404077777528613e-05,
+      "loss": 0.2131,
+      "step": 6661
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8638110891762141,
+      "learning_rate": 1.5402691939483046e-05,
+      "loss": 0.1747,
+      "step": 6662
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9144960253692133,
+      "learning_rate": 1.540130595488915e-05,
+      "loss": 0.2166,
+      "step": 6663
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.735671035910116,
+      "learning_rate": 1.539991982378453e-05,
+      "loss": 0.1921,
+      "step": 6664
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2657061752905912,
+      "learning_rate": 1.539853354620678e-05,
+      "loss": 0.2117,
+      "step": 6665
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.283712998444193,
+      "learning_rate": 1.53971471221935e-05,
+      "loss": 0.179,
+      "step": 6666
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8989145391276644,
+      "learning_rate": 1.53957605517823e-05,
+      "loss": 0.1903,
+      "step": 6667
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0048519214602194,
+      "learning_rate": 1.539437383501079e-05,
+      "loss": 0.2057,
+      "step": 6668
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0132983115014504,
+      "learning_rate": 1.5392986971916583e-05,
+      "loss": 0.2205,
+      "step": 6669
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4896675971356934,
+      "learning_rate": 1.53915999625373e-05,
+      "loss": 0.1955,
+      "step": 6670
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5033619249115908,
+      "learning_rate": 1.539021280691057e-05,
+      "loss": 0.2164,
+      "step": 6671
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4593355835166457,
+      "learning_rate": 1.5388825505074006e-05,
+      "loss": 0.2004,
+      "step": 6672
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.148201487459412,
+      "learning_rate": 1.538743805706525e-05,
+      "loss": 0.1927,
+      "step": 6673
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6333105220118933,
+      "learning_rate": 1.538605046292193e-05,
+      "loss": 0.2123,
+      "step": 6674
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9890080278145551,
+      "learning_rate": 1.5384662722681688e-05,
+      "loss": 0.1891,
+      "step": 6675
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9610666592755823,
+      "learning_rate": 1.5383274836382163e-05,
+      "loss": 0.1825,
+      "step": 6676
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0378885292526392,
+      "learning_rate": 1.5381886804061005e-05,
+      "loss": 0.2171,
+      "step": 6677
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1277837007123201,
+      "learning_rate": 1.5380498625755867e-05,
+      "loss": 0.2176,
+      "step": 6678
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0877352146653516,
+      "learning_rate": 1.5379110301504397e-05,
+      "loss": 0.2081,
+      "step": 6679
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2947756337756804,
+      "learning_rate": 1.5377721831344258e-05,
+      "loss": 0.2119,
+      "step": 6680
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.117584689887842,
+      "learning_rate": 1.5376333215313106e-05,
+      "loss": 0.2149,
+      "step": 6681
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2078699493400893,
+      "learning_rate": 1.5374944453448617e-05,
+      "loss": 0.2057,
+      "step": 6682
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9548572667866763,
+      "learning_rate": 1.5373555545788456e-05,
+      "loss": 0.2049,
+      "step": 6683
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3772461895252333,
+      "learning_rate": 1.5372166492370297e-05,
+      "loss": 0.1893,
+      "step": 6684
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9630206874145618,
+      "learning_rate": 1.5370777293231814e-05,
+      "loss": 0.1958,
+      "step": 6685
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3183565853223211,
+      "learning_rate": 1.5369387948410695e-05,
+      "loss": 0.2091,
+      "step": 6686
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0363802125546604,
+      "learning_rate": 1.536799845794463e-05,
+      "loss": 0.241,
+      "step": 6687
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8656078936422779,
+      "learning_rate": 1.53666088218713e-05,
+      "loss": 0.1964,
+      "step": 6688
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9737685003439084,
+      "learning_rate": 1.5365219040228402e-05,
+      "loss": 0.1881,
+      "step": 6689
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0616642657522841,
+      "learning_rate": 1.5363829113053633e-05,
+      "loss": 0.237,
+      "step": 6690
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8708419272926546,
+      "learning_rate": 1.53624390403847e-05,
+      "loss": 0.1886,
+      "step": 6691
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1335578938991093,
+      "learning_rate": 1.5361048822259302e-05,
+      "loss": 0.2157,
+      "step": 6692
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.326790601511181,
+      "learning_rate": 1.5359658458715158e-05,
+      "loss": 0.228,
+      "step": 6693
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9248727896429133,
+      "learning_rate": 1.5358267949789968e-05,
+      "loss": 0.2035,
+      "step": 6694
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2076276869173062,
+      "learning_rate": 1.535687729552146e-05,
+      "loss": 0.2129,
+      "step": 6695
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8942627148982174,
+      "learning_rate": 1.5355486495947353e-05,
+      "loss": 0.2178,
+      "step": 6696
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8807289461435734,
+      "learning_rate": 1.5354095551105374e-05,
+      "loss": 0.2192,
+      "step": 6697
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.842682012974875,
+      "learning_rate": 1.5352704461033247e-05,
+      "loss": 0.203,
+      "step": 6698
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9057246496827717,
+      "learning_rate": 1.5351313225768713e-05,
+      "loss": 0.1882,
+      "step": 6699
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1617478367242047,
+      "learning_rate": 1.5349921845349504e-05,
+      "loss": 0.2036,
+      "step": 6700
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8422370104317792,
+      "learning_rate": 1.5348530319813365e-05,
+      "loss": 0.2045,
+      "step": 6701
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.172859237684915,
+      "learning_rate": 1.5347138649198036e-05,
+      "loss": 0.1777,
+      "step": 6702
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.851183073220529,
+      "learning_rate": 1.5345746833541268e-05,
+      "loss": 0.1951,
+      "step": 6703
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3893811227479087,
+      "learning_rate": 1.5344354872880817e-05,
+      "loss": 0.1876,
+      "step": 6704
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2961449575191275,
+      "learning_rate": 1.534296276725444e-05,
+      "loss": 0.2221,
+      "step": 6705
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.471827649921739,
+      "learning_rate": 1.5341570516699893e-05,
+      "loss": 0.1939,
+      "step": 6706
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8421797836882626,
+      "learning_rate": 1.5340178121254944e-05,
+      "loss": 0.2059,
+      "step": 6707
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.968441843112463,
+      "learning_rate": 1.5338785580957366e-05,
+      "loss": 0.2059,
+      "step": 6708
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1233535791260196,
+      "learning_rate": 1.5337392895844923e-05,
+      "loss": 0.2172,
+      "step": 6709
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8078361898244383,
+      "learning_rate": 1.53360000659554e-05,
+      "loss": 0.1763,
+      "step": 6710
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7165845447840065,
+      "learning_rate": 1.533460709132657e-05,
+      "loss": 0.1835,
+      "step": 6711
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.5322930348415293,
+      "learning_rate": 1.5333213971996223e-05,
+      "loss": 0.2092,
+      "step": 6712
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9859837992060969,
+      "learning_rate": 1.5331820708002148e-05,
+      "loss": 0.2159,
+      "step": 6713
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9879252355628823,
+      "learning_rate": 1.533042729938213e-05,
+      "loss": 0.1894,
+      "step": 6714
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8812018427849039,
+      "learning_rate": 1.5329033746173975e-05,
+      "loss": 0.1833,
+      "step": 6715
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4601846109173515,
+      "learning_rate": 1.5327640048415476e-05,
+      "loss": 0.2237,
+      "step": 6716
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0581061376121739,
+      "learning_rate": 1.5326246206144443e-05,
+      "loss": 0.2092,
+      "step": 6717
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7203814696096753,
+      "learning_rate": 1.532485221939868e-05,
+      "loss": 0.1884,
+      "step": 6718
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2641468880148858,
+      "learning_rate": 1.5323458088216e-05,
+      "loss": 0.2215,
+      "step": 6719
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9041915947266804,
+      "learning_rate": 1.5322063812634213e-05,
+      "loss": 0.182,
+      "step": 6720
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9312729062386421,
+      "learning_rate": 1.532066939269115e-05,
+      "loss": 0.2172,
+      "step": 6721
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7275672748749746,
+      "learning_rate": 1.531927482842463e-05,
+      "loss": 0.1762,
+      "step": 6722
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9269472309163541,
+      "learning_rate": 1.531788011987248e-05,
+      "loss": 0.206,
+      "step": 6723
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1568597220771661,
+      "learning_rate": 1.5316485267072528e-05,
+      "loss": 0.1887,
+      "step": 6724
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9338998705897923,
+      "learning_rate": 1.5315090270062612e-05,
+      "loss": 0.203,
+      "step": 6725
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2715192485484121,
+      "learning_rate": 1.5313695128880578e-05,
+      "loss": 0.2145,
+      "step": 6726
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0559807576449627,
+      "learning_rate": 1.531229984356426e-05,
+      "loss": 0.1945,
+      "step": 6727
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1443412285252021,
+      "learning_rate": 1.5310904414151505e-05,
+      "loss": 0.2212,
+      "step": 6728
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2483375607136242,
+      "learning_rate": 1.530950884068017e-05,
+      "loss": 0.2071,
+      "step": 6729
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1469468803995153,
+      "learning_rate": 1.530811312318811e-05,
+      "loss": 0.2045,
+      "step": 6730
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2036637391446452,
+      "learning_rate": 1.530671726171318e-05,
+      "loss": 0.1866,
+      "step": 6731
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.198487986397559,
+      "learning_rate": 1.530532125629325e-05,
+      "loss": 0.2133,
+      "step": 6732
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7976346078056735,
+      "learning_rate": 1.5303925106966176e-05,
+      "loss": 0.2204,
+      "step": 6733
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.205968836433249,
+      "learning_rate": 1.5302528813769832e-05,
+      "loss": 0.2076,
+      "step": 6734
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9959892711878979,
+      "learning_rate": 1.5301132376742097e-05,
+      "loss": 0.2149,
+      "step": 6735
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0439639919923365,
+      "learning_rate": 1.5299735795920852e-05,
+      "loss": 0.2182,
+      "step": 6736
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3157354916778414,
+      "learning_rate": 1.5298339071343965e-05,
+      "loss": 0.1834,
+      "step": 6737
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.008918421266012,
+      "learning_rate": 1.5296942203049336e-05,
+      "loss": 0.2016,
+      "step": 6738
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0213466637424429,
+      "learning_rate": 1.5295545191074854e-05,
+      "loss": 0.1827,
+      "step": 6739
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.393285649452404,
+      "learning_rate": 1.5294148035458406e-05,
+      "loss": 0.1925,
+      "step": 6740
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.611635542327183,
+      "learning_rate": 1.529275073623789e-05,
+      "loss": 0.2131,
+      "step": 6741
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9380453950042238,
+      "learning_rate": 1.5291353293451216e-05,
+      "loss": 0.2109,
+      "step": 6742
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.780304518112081,
+      "learning_rate": 1.5289955707136282e-05,
+      "loss": 0.21,
+      "step": 6743
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.276791397406679,
+      "learning_rate": 1.5288557977331006e-05,
+      "loss": 0.1979,
+      "step": 6744
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0230013122394754,
+      "learning_rate": 1.528716010407329e-05,
+      "loss": 0.2092,
+      "step": 6745
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.911278338876281,
+      "learning_rate": 1.528576208740106e-05,
+      "loss": 0.1953,
+      "step": 6746
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8718440559042655,
+      "learning_rate": 1.5284363927352234e-05,
+      "loss": 0.2164,
+      "step": 6747
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.960141783585586,
+      "learning_rate": 1.528296562396474e-05,
+      "loss": 0.2246,
+      "step": 6748
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0057641215159536,
+      "learning_rate": 1.5281567177276504e-05,
+      "loss": 0.2114,
+      "step": 6749
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1115177064216062,
+      "learning_rate": 1.5280168587325462e-05,
+      "loss": 0.2166,
+      "step": 6750
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9984220338545741,
+      "learning_rate": 1.5278769854149544e-05,
+      "loss": 0.2036,
+      "step": 6751
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0728686079952425,
+      "learning_rate": 1.5277370977786698e-05,
+      "loss": 0.2143,
+      "step": 6752
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.00311623585292,
+      "learning_rate": 1.527597195827487e-05,
+      "loss": 0.2022,
+      "step": 6753
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4416304137365434,
+      "learning_rate": 1.5274572795652e-05,
+      "loss": 0.1861,
+      "step": 6754
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.353646006666074,
+      "learning_rate": 1.5273173489956045e-05,
+      "loss": 0.1966,
+      "step": 6755
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6279188709189063,
+      "learning_rate": 1.5271774041224965e-05,
+      "loss": 0.2124,
+      "step": 6756
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2112152569319565,
+      "learning_rate": 1.5270374449496713e-05,
+      "loss": 0.1893,
+      "step": 6757
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9026300065868202,
+      "learning_rate": 1.526897471480926e-05,
+      "loss": 0.2145,
+      "step": 6758
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4535905698686216,
+      "learning_rate": 1.5267574837200567e-05,
+      "loss": 0.1969,
+      "step": 6759
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9155814290404279,
+      "learning_rate": 1.5266174816708607e-05,
+      "loss": 0.2077,
+      "step": 6760
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4571808090129743,
+      "learning_rate": 1.526477465337136e-05,
+      "loss": 0.2047,
+      "step": 6761
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.700143986136528,
+      "learning_rate": 1.5263374347226804e-05,
+      "loss": 0.207,
+      "step": 6762
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3011385282165446,
+      "learning_rate": 1.526197389831292e-05,
+      "loss": 0.1692,
+      "step": 6763
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0324443289219334,
+      "learning_rate": 1.52605733066677e-05,
+      "loss": 0.2005,
+      "step": 6764
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9902437055657942,
+      "learning_rate": 1.5259172572329132e-05,
+      "loss": 0.181,
+      "step": 6765
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4277802318478219,
+      "learning_rate": 1.5257771695335207e-05,
+      "loss": 0.1872,
+      "step": 6766
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9587925594158275,
+      "learning_rate": 1.5256370675723928e-05,
+      "loss": 0.1822,
+      "step": 6767
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.169017797107394,
+      "learning_rate": 1.52549695135333e-05,
+      "loss": 0.2135,
+      "step": 6768
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0043873111957573,
+      "learning_rate": 1.5253568208801324e-05,
+      "loss": 0.2032,
+      "step": 6769
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9227197727775629,
+      "learning_rate": 1.5252166761566018e-05,
+      "loss": 0.2208,
+      "step": 6770
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9284620558439093,
+      "learning_rate": 1.5250765171865391e-05,
+      "loss": 0.2122,
+      "step": 6771
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2444535096292564,
+      "learning_rate": 1.5249363439737458e-05,
+      "loss": 0.2359,
+      "step": 6772
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8917021534524678,
+      "learning_rate": 1.5247961565220251e-05,
+      "loss": 0.1886,
+      "step": 6773
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0070824235122227,
+      "learning_rate": 1.5246559548351786e-05,
+      "loss": 0.1946,
+      "step": 6774
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0143443272664339,
+      "learning_rate": 1.5245157389170099e-05,
+      "loss": 0.2004,
+      "step": 6775
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8925425331684878,
+      "learning_rate": 1.5243755087713221e-05,
+      "loss": 0.1978,
+      "step": 6776
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9067333501121372,
+      "learning_rate": 1.5242352644019188e-05,
+      "loss": 0.1804,
+      "step": 6777
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8315046848626775,
+      "learning_rate": 1.5240950058126047e-05,
+      "loss": 0.2057,
+      "step": 6778
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2926472057668967,
+      "learning_rate": 1.5239547330071838e-05,
+      "loss": 0.1967,
+      "step": 6779
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4130144050799631,
+      "learning_rate": 1.5238144459894612e-05,
+      "loss": 0.2164,
+      "step": 6780
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5404583219652528,
+      "learning_rate": 1.523674144763242e-05,
+      "loss": 0.2328,
+      "step": 6781
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4509180543256275,
+      "learning_rate": 1.5235338293323322e-05,
+      "loss": 0.1865,
+      "step": 6782
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9155452259245301,
+      "learning_rate": 1.5233934997005377e-05,
+      "loss": 0.1868,
+      "step": 6783
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9785165899029814,
+      "learning_rate": 1.523253155871665e-05,
+      "loss": 0.2161,
+      "step": 6784
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9681027179097117,
+      "learning_rate": 1.5231127978495208e-05,
+      "loss": 0.1961,
+      "step": 6785
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.291598955342103,
+      "learning_rate": 1.5229724256379124e-05,
+      "loss": 0.1936,
+      "step": 6786
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2179833211574773,
+      "learning_rate": 1.5228320392406476e-05,
+      "loss": 0.1859,
+      "step": 6787
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.132302330300989,
+      "learning_rate": 1.522691638661534e-05,
+      "loss": 0.1998,
+      "step": 6788
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9243465153919224,
+      "learning_rate": 1.5225512239043805e-05,
+      "loss": 0.1908,
+      "step": 6789
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.794549770398676,
+      "learning_rate": 1.5224107949729952e-05,
+      "loss": 0.1878,
+      "step": 6790
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1940722120482383,
+      "learning_rate": 1.5222703518711876e-05,
+      "loss": 0.1877,
+      "step": 6791
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.006766552572996,
+      "learning_rate": 1.5221298946027674e-05,
+      "loss": 0.1982,
+      "step": 6792
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2122944014485775,
+      "learning_rate": 1.5219894231715443e-05,
+      "loss": 0.1661,
+      "step": 6793
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7706146021851047,
+      "learning_rate": 1.521848937581328e-05,
+      "loss": 0.1752,
+      "step": 6794
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8239227910933536,
+      "learning_rate": 1.5217084378359306e-05,
+      "loss": 0.1822,
+      "step": 6795
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8731121198592614,
+      "learning_rate": 1.5215679239391621e-05,
+      "loss": 0.1922,
+      "step": 6796
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.218467420504002,
+      "learning_rate": 1.5214273958948343e-05,
+      "loss": 0.198,
+      "step": 6797
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.232086371764998,
+      "learning_rate": 1.5212868537067587e-05,
+      "loss": 0.1799,
+      "step": 6798
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8349998328027023,
+      "learning_rate": 1.5211462973787478e-05,
+      "loss": 0.1858,
+      "step": 6799
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9479616491015617,
+      "learning_rate": 1.5210057269146141e-05,
+      "loss": 0.2169,
+      "step": 6800
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.021839551116278,
+      "learning_rate": 1.5208651423181709e-05,
+      "loss": 0.2123,
+      "step": 6801
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7427252710048137,
+      "learning_rate": 1.5207245435932312e-05,
+      "loss": 0.2014,
+      "step": 6802
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3946540105732994,
+      "learning_rate": 1.5205839307436088e-05,
+      "loss": 0.206,
+      "step": 6803
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9192566427112906,
+      "learning_rate": 1.5204433037731177e-05,
+      "loss": 0.1898,
+      "step": 6804
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0661683152184878,
+      "learning_rate": 1.5203026626855728e-05,
+      "loss": 0.217,
+      "step": 6805
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9281868878900686,
+      "learning_rate": 1.5201620074847888e-05,
+      "loss": 0.2235,
+      "step": 6806
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8851710721864421,
+      "learning_rate": 1.5200213381745807e-05,
+      "loss": 0.2138,
+      "step": 6807
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8892054097059979,
+      "learning_rate": 1.5198806547587648e-05,
+      "loss": 0.2257,
+      "step": 6808
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2053124528725399,
+      "learning_rate": 1.5197399572411566e-05,
+      "loss": 0.1975,
+      "step": 6809
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.960929308363589,
+      "learning_rate": 1.5195992456255728e-05,
+      "loss": 0.1986,
+      "step": 6810
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0636077023307362,
+      "learning_rate": 1.51945851991583e-05,
+      "loss": 0.2068,
+      "step": 6811
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.867963274482051,
+      "learning_rate": 1.5193177801157456e-05,
+      "loss": 0.1769,
+      "step": 6812
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8875259655818101,
+      "learning_rate": 1.5191770262291367e-05,
+      "loss": 0.1952,
+      "step": 6813
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9265142993267459,
+      "learning_rate": 1.5190362582598223e-05,
+      "loss": 0.1973,
+      "step": 6814
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0713731845568697,
+      "learning_rate": 1.5188954762116197e-05,
+      "loss": 0.1891,
+      "step": 6815
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.170054639379894,
+      "learning_rate": 1.518754680088348e-05,
+      "loss": 0.1806,
+      "step": 6816
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5358593286525253,
+      "learning_rate": 1.5186138698938262e-05,
+      "loss": 0.2304,
+      "step": 6817
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2267745756341035,
+      "learning_rate": 1.5184730456318742e-05,
+      "loss": 0.21,
+      "step": 6818
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2369692943548545,
+      "learning_rate": 1.5183322073063113e-05,
+      "loss": 0.1932,
+      "step": 6819
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9842943782482564,
+      "learning_rate": 1.5181913549209582e-05,
+      "loss": 0.199,
+      "step": 6820
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1127812635949084,
+      "learning_rate": 1.5180504884796352e-05,
+      "loss": 0.1981,
+      "step": 6821
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.15296947156494,
+      "learning_rate": 1.5179096079861633e-05,
+      "loss": 0.1961,
+      "step": 6822
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3997778340175386,
+      "learning_rate": 1.5177687134443644e-05,
+      "loss": 0.2149,
+      "step": 6823
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1547947538700556,
+      "learning_rate": 1.51762780485806e-05,
+      "loss": 0.2027,
+      "step": 6824
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2744762270890364,
+      "learning_rate": 1.5174868822310715e-05,
+      "loss": 0.1796,
+      "step": 6825
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8809148208865116,
+      "learning_rate": 1.5173459455672225e-05,
+      "loss": 0.2006,
+      "step": 6826
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7797320005770062,
+      "learning_rate": 1.5172049948703356e-05,
+      "loss": 0.2017,
+      "step": 6827
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.074741508538611,
+      "learning_rate": 1.5170640301442339e-05,
+      "loss": 0.2007,
+      "step": 6828
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1055816149226974,
+      "learning_rate": 1.516923051392741e-05,
+      "loss": 0.2054,
+      "step": 6829
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8006123473485938,
+      "learning_rate": 1.516782058619681e-05,
+      "loss": 0.2051,
+      "step": 6830
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9676537160614846,
+      "learning_rate": 1.516641051828879e-05,
+      "loss": 0.177,
+      "step": 6831
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2592218417012093,
+      "learning_rate": 1.5165000310241592e-05,
+      "loss": 0.1988,
+      "step": 6832
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7618946631132122,
+      "learning_rate": 1.5163589962093466e-05,
+      "loss": 0.1668,
+      "step": 6833
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8695663377266801,
+      "learning_rate": 1.5162179473882668e-05,
+      "loss": 0.1817,
+      "step": 6834
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1313227587735268,
+      "learning_rate": 1.5160768845647464e-05,
+      "loss": 0.2208,
+      "step": 6835
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7788299794995026,
+      "learning_rate": 1.5159358077426114e-05,
+      "loss": 0.2113,
+      "step": 6836
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0772132694371461,
+      "learning_rate": 1.5157947169256886e-05,
+      "loss": 0.2016,
+      "step": 6837
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8650455266282737,
+      "learning_rate": 1.515653612117805e-05,
+      "loss": 0.1963,
+      "step": 6838
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0680136956495005,
+      "learning_rate": 1.5155124933227876e-05,
+      "loss": 0.2127,
+      "step": 6839
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0971031970802028,
+      "learning_rate": 1.515371360544465e-05,
+      "loss": 0.1729,
+      "step": 6840
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9096219601191085,
+      "learning_rate": 1.5152302137866653e-05,
+      "loss": 0.1937,
+      "step": 6841
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9816498176529843,
+      "learning_rate": 1.5150890530532165e-05,
+      "loss": 0.1817,
+      "step": 6842
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3215012680366147,
+      "learning_rate": 1.5149478783479484e-05,
+      "loss": 0.2153,
+      "step": 6843
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.101868774315477,
+      "learning_rate": 1.51480668967469e-05,
+      "loss": 0.2143,
+      "step": 6844
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8953111804081765,
+      "learning_rate": 1.514665487037271e-05,
+      "loss": 0.2136,
+      "step": 6845
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9155024355336064,
+      "learning_rate": 1.5145242704395215e-05,
+      "loss": 0.1709,
+      "step": 6846
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.194699702974594,
+      "learning_rate": 1.5143830398852722e-05,
+      "loss": 0.2143,
+      "step": 6847
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.815983227890357,
+      "learning_rate": 1.5142417953783536e-05,
+      "loss": 0.1805,
+      "step": 6848
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3899500734068588,
+      "learning_rate": 1.5141005369225976e-05,
+      "loss": 0.1977,
+      "step": 6849
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8772900169020614,
+      "learning_rate": 1.5139592645218355e-05,
+      "loss": 0.2173,
+      "step": 6850
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.3715237882830085,
+      "learning_rate": 1.5138179781798994e-05,
+      "loss": 0.2088,
+      "step": 6851
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9728600232206785,
+      "learning_rate": 1.513676677900621e-05,
+      "loss": 0.2141,
+      "step": 6852
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2467143391094206,
+      "learning_rate": 1.5135353636878343e-05,
+      "loss": 0.2127,
+      "step": 6853
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4410997273671753,
+      "learning_rate": 1.5133940355453717e-05,
+      "loss": 0.1994,
+      "step": 6854
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0130725754368426,
+      "learning_rate": 1.513252693477067e-05,
+      "loss": 0.191,
+      "step": 6855
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9178135147372484,
+      "learning_rate": 1.5131113374867537e-05,
+      "loss": 0.157,
+      "step": 6856
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0776349600749444,
+      "learning_rate": 1.5129699675782666e-05,
+      "loss": 0.2282,
+      "step": 6857
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2034064612598838,
+      "learning_rate": 1.5128285837554404e-05,
+      "loss": 0.2008,
+      "step": 6858
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0843112113793854,
+      "learning_rate": 1.5126871860221098e-05,
+      "loss": 0.1881,
+      "step": 6859
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0250256906050836,
+      "learning_rate": 1.5125457743821098e-05,
+      "loss": 0.2129,
+      "step": 6860
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9537784444716515,
+      "learning_rate": 1.5124043488392772e-05,
+      "loss": 0.2287,
+      "step": 6861
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0963638776333486,
+      "learning_rate": 1.5122629093974476e-05,
+      "loss": 0.1926,
+      "step": 6862
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.18682479755038,
+      "learning_rate": 1.5121214560604579e-05,
+      "loss": 0.2174,
+      "step": 6863
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8760608982889208,
+      "learning_rate": 1.5119799888321444e-05,
+      "loss": 0.2168,
+      "step": 6864
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0627961084019029,
+      "learning_rate": 1.5118385077163446e-05,
+      "loss": 0.1941,
+      "step": 6865
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6957928296348652,
+      "learning_rate": 1.5116970127168969e-05,
+      "loss": 0.2307,
+      "step": 6866
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0456603258940167,
+      "learning_rate": 1.5115555038376386e-05,
+      "loss": 0.2038,
+      "step": 6867
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0644078326243325,
+      "learning_rate": 1.5114139810824084e-05,
+      "loss": 0.2085,
+      "step": 6868
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9346639998075899,
+      "learning_rate": 1.5112724444550449e-05,
+      "loss": 0.1976,
+      "step": 6869
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9871990707491872,
+      "learning_rate": 1.5111308939593876e-05,
+      "loss": 0.2325,
+      "step": 6870
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.862600105336854,
+      "learning_rate": 1.510989329599276e-05,
+      "loss": 0.2257,
+      "step": 6871
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8081876993422035,
+      "learning_rate": 1.51084775137855e-05,
+      "loss": 0.1908,
+      "step": 6872
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1136152808647861,
+      "learning_rate": 1.5107061593010497e-05,
+      "loss": 0.1845,
+      "step": 6873
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9049398819784213,
+      "learning_rate": 1.5105645533706161e-05,
+      "loss": 0.2169,
+      "step": 6874
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8437673319992209,
+      "learning_rate": 1.5104229335910901e-05,
+      "loss": 0.1851,
+      "step": 6875
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9911256372429206,
+      "learning_rate": 1.5102812999663136e-05,
+      "loss": 0.204,
+      "step": 6876
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.059383024074309,
+      "learning_rate": 1.5101396525001275e-05,
+      "loss": 0.2161,
+      "step": 6877
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.807238408215533,
+      "learning_rate": 1.5099979911963747e-05,
+      "loss": 0.1751,
+      "step": 6878
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0651713275527612,
+      "learning_rate": 1.5098563160588975e-05,
+      "loss": 0.1957,
+      "step": 6879
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1544476706394258,
+      "learning_rate": 1.5097146270915391e-05,
+      "loss": 0.2063,
+      "step": 6880
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0130726774046317,
+      "learning_rate": 1.5095729242981426e-05,
+      "loss": 0.2123,
+      "step": 6881
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.946893609225108,
+      "learning_rate": 1.5094312076825514e-05,
+      "loss": 0.1984,
+      "step": 6882
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9464456103251006,
+      "learning_rate": 1.5092894772486104e-05,
+      "loss": 0.2102,
+      "step": 6883
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0216789633156547,
+      "learning_rate": 1.5091477330001634e-05,
+      "loss": 0.1997,
+      "step": 6884
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6466262185119827,
+      "learning_rate": 1.5090059749410553e-05,
+      "loss": 0.2085,
+      "step": 6885
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9729954755515294,
+      "learning_rate": 1.5088642030751314e-05,
+      "loss": 0.171,
+      "step": 6886
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.385590133366258,
+      "learning_rate": 1.5087224174062371e-05,
+      "loss": 0.1922,
+      "step": 6887
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8916447836837915,
+      "learning_rate": 1.5085806179382188e-05,
+      "loss": 0.2069,
+      "step": 6888
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2889193077338137,
+      "learning_rate": 1.5084388046749224e-05,
+      "loss": 0.2201,
+      "step": 6889
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.921388178955955,
+      "learning_rate": 1.5082969776201948e-05,
+      "loss": 0.2002,
+      "step": 6890
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2911495515489098,
+      "learning_rate": 1.5081551367778828e-05,
+      "loss": 0.2063,
+      "step": 6891
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.54823405008985,
+      "learning_rate": 1.508013282151834e-05,
+      "loss": 0.186,
+      "step": 6892
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8821794175637585,
+      "learning_rate": 1.5078714137458965e-05,
+      "loss": 0.2036,
+      "step": 6893
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5243194241716584,
+      "learning_rate": 1.5077295315639183e-05,
+      "loss": 0.2197,
+      "step": 6894
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7558299954036014,
+      "learning_rate": 1.5075876356097472e-05,
+      "loss": 0.1974,
+      "step": 6895
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0411962229686005,
+      "learning_rate": 1.5074457258872332e-05,
+      "loss": 0.1895,
+      "step": 6896
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1459677441955394,
+      "learning_rate": 1.5073038024002254e-05,
+      "loss": 0.2032,
+      "step": 6897
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3905183101820433,
+      "learning_rate": 1.5071618651525733e-05,
+      "loss": 0.2074,
+      "step": 6898
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9654236834489405,
+      "learning_rate": 1.5070199141481267e-05,
+      "loss": 0.21,
+      "step": 6899
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0754205007601063,
+      "learning_rate": 1.5068779493907364e-05,
+      "loss": 0.2037,
+      "step": 6900
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1811109530453496,
+      "learning_rate": 1.5067359708842531e-05,
+      "loss": 0.2191,
+      "step": 6901
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8896704583059244,
+      "learning_rate": 1.506593978632528e-05,
+      "loss": 0.1918,
+      "step": 6902
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8190830743448189,
+      "learning_rate": 1.5064519726394127e-05,
+      "loss": 0.1848,
+      "step": 6903
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4839731613633682,
+      "learning_rate": 1.5063099529087588e-05,
+      "loss": 0.1942,
+      "step": 6904
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8611499038512758,
+      "learning_rate": 1.506167919444419e-05,
+      "loss": 0.1833,
+      "step": 6905
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9749771586556178,
+      "learning_rate": 1.5060258722502457e-05,
+      "loss": 0.2128,
+      "step": 6906
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1688444786842203,
+      "learning_rate": 1.5058838113300922e-05,
+      "loss": 0.1839,
+      "step": 6907
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9682085388119024,
+      "learning_rate": 1.5057417366878117e-05,
+      "loss": 0.201,
+      "step": 6908
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0901203502465975,
+      "learning_rate": 1.505599648327258e-05,
+      "loss": 0.2072,
+      "step": 6909
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.404939365745371,
+      "learning_rate": 1.505457546252285e-05,
+      "loss": 0.226,
+      "step": 6910
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0672437834253885,
+      "learning_rate": 1.5053154304667481e-05,
+      "loss": 0.2089,
+      "step": 6911
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0637403739652012,
+      "learning_rate": 1.5051733009745013e-05,
+      "loss": 0.1869,
+      "step": 6912
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9076819960560645,
+      "learning_rate": 1.5050311577794002e-05,
+      "loss": 0.1877,
+      "step": 6913
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9717309390704598,
+      "learning_rate": 1.5048890008853004e-05,
+      "loss": 0.1993,
+      "step": 6914
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9407606112256421,
+      "learning_rate": 1.5047468302960577e-05,
+      "loss": 0.2085,
+      "step": 6915
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4738042456951326,
+      "learning_rate": 1.504604646015529e-05,
+      "loss": 0.184,
+      "step": 6916
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7211102814223973,
+      "learning_rate": 1.5044624480475704e-05,
+      "loss": 0.2234,
+      "step": 6917
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4920048618936634,
+      "learning_rate": 1.50432023639604e-05,
+      "loss": 0.1944,
+      "step": 6918
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0084709392210685,
+      "learning_rate": 1.5041780110647945e-05,
+      "loss": 0.1899,
+      "step": 6919
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1115772913320738,
+      "learning_rate": 1.5040357720576917e-05,
+      "loss": 0.2065,
+      "step": 6920
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0794035077770285,
+      "learning_rate": 1.5038935193785904e-05,
+      "loss": 0.2026,
+      "step": 6921
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9209720954523564,
+      "learning_rate": 1.5037512530313487e-05,
+      "loss": 0.1812,
+      "step": 6922
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3927461318168444,
+      "learning_rate": 1.503608973019826e-05,
+      "loss": 0.2163,
+      "step": 6923
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8818625228439795,
+      "learning_rate": 1.5034666793478814e-05,
+      "loss": 0.2027,
+      "step": 6924
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9774795678754011,
+      "learning_rate": 1.5033243720193746e-05,
+      "loss": 0.1986,
+      "step": 6925
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1421701507240238,
+      "learning_rate": 1.5031820510381661e-05,
+      "loss": 0.2108,
+      "step": 6926
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9000965134357147,
+      "learning_rate": 1.5030397164081157e-05,
+      "loss": 0.2099,
+      "step": 6927
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.095911521854206,
+      "learning_rate": 1.502897368133085e-05,
+      "loss": 0.2009,
+      "step": 6928
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8489297047692981,
+      "learning_rate": 1.5027550062169343e-05,
+      "loss": 0.1884,
+      "step": 6929
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.997414952129682,
+      "learning_rate": 1.5026126306635256e-05,
+      "loss": 0.1995,
+      "step": 6930
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9949856638154007,
+      "learning_rate": 1.5024702414767212e-05,
+      "loss": 0.2306,
+      "step": 6931
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8183068565989067,
+      "learning_rate": 1.5023278386603832e-05,
+      "loss": 0.1763,
+      "step": 6932
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1024967773167957,
+      "learning_rate": 1.502185422218374e-05,
+      "loss": 0.2041,
+      "step": 6933
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0647604029272146,
+      "learning_rate": 1.5020429921545572e-05,
+      "loss": 0.2104,
+      "step": 6934
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2222934743683858,
+      "learning_rate": 1.5019005484727953e-05,
+      "loss": 0.1928,
+      "step": 6935
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.949740467637576,
+      "learning_rate": 1.501758091176953e-05,
+      "loss": 0.2148,
+      "step": 6936
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.057136051282396,
+      "learning_rate": 1.5016156202708942e-05,
+      "loss": 0.2267,
+      "step": 6937
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9248210993970946,
+      "learning_rate": 1.5014731357584835e-05,
+      "loss": 0.1782,
+      "step": 6938
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0259772541994394,
+      "learning_rate": 1.5013306376435852e-05,
+      "loss": 0.2216,
+      "step": 6939
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6482033868110175,
+      "learning_rate": 1.5011881259300654e-05,
+      "loss": 0.2146,
+      "step": 6940
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.137004966715899,
+      "learning_rate": 1.5010456006217892e-05,
+      "loss": 0.2031,
+      "step": 6941
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8897749780607621,
+      "learning_rate": 1.5009030617226227e-05,
+      "loss": 0.2103,
+      "step": 6942
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7726206917351606,
+      "learning_rate": 1.5007605092364329e-05,
+      "loss": 0.2039,
+      "step": 6943
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9287323221846299,
+      "learning_rate": 1.5006179431670853e-05,
+      "loss": 0.2021,
+      "step": 6944
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8194781241438943,
+      "learning_rate": 1.5004753635184482e-05,
+      "loss": 0.2027,
+      "step": 6945
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3093313015385672,
+      "learning_rate": 1.5003327702943886e-05,
+      "loss": 0.2096,
+      "step": 6946
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.751471795194662,
+      "learning_rate": 1.5001901634987741e-05,
+      "loss": 0.1795,
+      "step": 6947
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8658145897043582,
+      "learning_rate": 1.500047543135473e-05,
+      "loss": 0.198,
+      "step": 6948
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.927812722124989,
+      "learning_rate": 1.4999049092083546e-05,
+      "loss": 0.204,
+      "step": 6949
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0196694768308388,
+      "learning_rate": 1.499762261721287e-05,
+      "loss": 0.2097,
+      "step": 6950
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1506071629214236,
+      "learning_rate": 1.4996196006781398e-05,
+      "loss": 0.2234,
+      "step": 6951
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0757424215079343,
+      "learning_rate": 1.4994769260827825e-05,
+      "loss": 0.2187,
+      "step": 6952
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8124410268278714,
+      "learning_rate": 1.4993342379390859e-05,
+      "loss": 0.2122,
+      "step": 6953
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.055419148942473,
+      "learning_rate": 1.4991915362509196e-05,
+      "loss": 0.1975,
+      "step": 6954
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.074342125307164,
+      "learning_rate": 1.4990488210221545e-05,
+      "loss": 0.2257,
+      "step": 6955
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8775555168046059,
+      "learning_rate": 1.4989060922566623e-05,
+      "loss": 0.1871,
+      "step": 6956
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9828925549110333,
+      "learning_rate": 1.4987633499583138e-05,
+      "loss": 0.2055,
+      "step": 6957
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0764743184030248,
+      "learning_rate": 1.4986205941309818e-05,
+      "loss": 0.205,
+      "step": 6958
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9363216899302678,
+      "learning_rate": 1.4984778247785375e-05,
+      "loss": 0.1804,
+      "step": 6959
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.931636292642853,
+      "learning_rate": 1.4983350419048544e-05,
+      "loss": 0.1797,
+      "step": 6960
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9904177526333235,
+      "learning_rate": 1.498192245513805e-05,
+      "loss": 0.1902,
+      "step": 6961
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8392932536433085,
+      "learning_rate": 1.4980494356092626e-05,
+      "loss": 0.206,
+      "step": 6962
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8715178627695,
+      "learning_rate": 1.4979066121951014e-05,
+      "loss": 0.1972,
+      "step": 6963
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1678631470250729,
+      "learning_rate": 1.4977637752751953e-05,
+      "loss": 0.1891,
+      "step": 6964
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7647725339845265,
+      "learning_rate": 1.4976209248534183e-05,
+      "loss": 0.1959,
+      "step": 6965
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8787375165725418,
+      "learning_rate": 1.4974780609336459e-05,
+      "loss": 0.2146,
+      "step": 6966
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8037882146692238,
+      "learning_rate": 1.497335183519753e-05,
+      "loss": 0.2131,
+      "step": 6967
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9437660746025482,
+      "learning_rate": 1.497192292615615e-05,
+      "loss": 0.1672,
+      "step": 6968
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9548608764647659,
+      "learning_rate": 1.497049388225108e-05,
+      "loss": 0.1965,
+      "step": 6969
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8126396317616935,
+      "learning_rate": 1.4969064703521082e-05,
+      "loss": 0.206,
+      "step": 6970
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.600045784968216,
+      "learning_rate": 1.4967635390004924e-05,
+      "loss": 0.181,
+      "step": 6971
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1969410873168724,
+      "learning_rate": 1.496620594174138e-05,
+      "loss": 0.2295,
+      "step": 6972
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9415730347137686,
+      "learning_rate": 1.4964776358769213e-05,
+      "loss": 0.2113,
+      "step": 6973
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.163755494326272,
+      "learning_rate": 1.496334664112721e-05,
+      "loss": 0.2001,
+      "step": 6974
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9033674842959668,
+      "learning_rate": 1.4961916788854147e-05,
+      "loss": 0.1906,
+      "step": 6975
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1287017676044906,
+      "learning_rate": 1.4960486801988811e-05,
+      "loss": 0.2106,
+      "step": 6976
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1671485012836642,
+      "learning_rate": 1.4959056680569992e-05,
+      "loss": 0.1838,
+      "step": 6977
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9909927570451686,
+      "learning_rate": 1.4957626424636482e-05,
+      "loss": 0.2005,
+      "step": 6978
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7024089312556172,
+      "learning_rate": 1.495619603422707e-05,
+      "loss": 0.1863,
+      "step": 6979
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8736425378884803,
+      "learning_rate": 1.4954765509380565e-05,
+      "loss": 0.2134,
+      "step": 6980
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0248494978958178,
+      "learning_rate": 1.4953334850135765e-05,
+      "loss": 0.2203,
+      "step": 6981
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1433657512824882,
+      "learning_rate": 1.495190405653148e-05,
+      "loss": 0.2075,
+      "step": 6982
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.797062012180071,
+      "learning_rate": 1.4950473128606513e-05,
+      "loss": 0.1951,
+      "step": 6983
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9021502798448437,
+      "learning_rate": 1.4949042066399684e-05,
+      "loss": 0.1921,
+      "step": 6984
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9354344111856875,
+      "learning_rate": 1.4947610869949811e-05,
+      "loss": 0.2152,
+      "step": 6985
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2873601297598602,
+      "learning_rate": 1.4946179539295713e-05,
+      "loss": 0.2066,
+      "step": 6986
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1742473061174559,
+      "learning_rate": 1.4944748074476211e-05,
+      "loss": 0.2242,
+      "step": 6987
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.027615230081389,
+      "learning_rate": 1.4943316475530145e-05,
+      "loss": 0.2241,
+      "step": 6988
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4938750326455092,
+      "learning_rate": 1.4941884742496338e-05,
+      "loss": 0.2058,
+      "step": 6989
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8638539100195232,
+      "learning_rate": 1.4940452875413627e-05,
+      "loss": 0.2076,
+      "step": 6990
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8909119223100532,
+      "learning_rate": 1.4939020874320856e-05,
+      "loss": 0.2382,
+      "step": 6991
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.959846464244343,
+      "learning_rate": 1.4937588739256861e-05,
+      "loss": 0.1874,
+      "step": 6992
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8957640924216222,
+      "learning_rate": 1.4936156470260494e-05,
+      "loss": 0.1897,
+      "step": 6993
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9760286467099434,
+      "learning_rate": 1.4934724067370604e-05,
+      "loss": 0.1896,
+      "step": 6994
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7759132698260575,
+      "learning_rate": 1.4933291530626047e-05,
+      "loss": 0.1905,
+      "step": 6995
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6179256438101102,
+      "learning_rate": 1.4931858860065676e-05,
+      "loss": 0.1889,
+      "step": 6996
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1805021494268257,
+      "learning_rate": 1.4930426055728354e-05,
+      "loss": 0.2042,
+      "step": 6997
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9349921230667492,
+      "learning_rate": 1.492899311765295e-05,
+      "loss": 0.1933,
+      "step": 6998
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0884404176575875,
+      "learning_rate": 1.4927560045878328e-05,
+      "loss": 0.1874,
+      "step": 6999
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.371867437927246,
+      "learning_rate": 1.492612684044336e-05,
+      "loss": 0.1859,
+      "step": 7000
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5014463510510556,
+      "learning_rate": 1.4924693501386925e-05,
+      "loss": 0.1972,
+      "step": 7001
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.658847598307763,
+      "learning_rate": 1.4923260028747899e-05,
+      "loss": 0.1931,
+      "step": 7002
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9026298128366345,
+      "learning_rate": 1.4921826422565169e-05,
+      "loss": 0.2159,
+      "step": 7003
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.740979167213659,
+      "learning_rate": 1.4920392682877618e-05,
+      "loss": 0.1958,
+      "step": 7004
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8531910200230725,
+      "learning_rate": 1.4918958809724135e-05,
+      "loss": 0.1846,
+      "step": 7005
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7906279464839503,
+      "learning_rate": 1.491752480314362e-05,
+      "loss": 0.2073,
+      "step": 7006
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9873303650652501,
+      "learning_rate": 1.4916090663174966e-05,
+      "loss": 0.1949,
+      "step": 7007
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.881165750191825,
+      "learning_rate": 1.4914656389857076e-05,
+      "loss": 0.2105,
+      "step": 7008
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9239039519027061,
+      "learning_rate": 1.4913221983228851e-05,
+      "loss": 0.2083,
+      "step": 7009
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0745052245332594,
+      "learning_rate": 1.4911787443329204e-05,
+      "loss": 0.2027,
+      "step": 7010
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8717707200753981,
+      "learning_rate": 1.4910352770197044e-05,
+      "loss": 0.2013,
+      "step": 7011
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8507639295424179,
+      "learning_rate": 1.4908917963871292e-05,
+      "loss": 0.2,
+      "step": 7012
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7524155811833033,
+      "learning_rate": 1.4907483024390859e-05,
+      "loss": 0.1818,
+      "step": 7013
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8231612141524862,
+      "learning_rate": 1.4906047951794671e-05,
+      "loss": 0.1897,
+      "step": 7014
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0364202333602426,
+      "learning_rate": 1.4904612746121657e-05,
+      "loss": 0.2018,
+      "step": 7015
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3215366699557016,
+      "learning_rate": 1.4903177407410749e-05,
+      "loss": 0.1915,
+      "step": 7016
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8798560451011856,
+      "learning_rate": 1.4901741935700873e-05,
+      "loss": 0.2068,
+      "step": 7017
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8262393299901581,
+      "learning_rate": 1.4900306331030967e-05,
+      "loss": 0.2209,
+      "step": 7018
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.817376552435723,
+      "learning_rate": 1.489887059343998e-05,
+      "loss": 0.2057,
+      "step": 7019
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2862260016588265,
+      "learning_rate": 1.4897434722966851e-05,
+      "loss": 0.2076,
+      "step": 7020
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8367966460811618,
+      "learning_rate": 1.4895998719650526e-05,
+      "loss": 0.2173,
+      "step": 7021
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6914550113491021,
+      "learning_rate": 1.4894562583529961e-05,
+      "loss": 0.197,
+      "step": 7022
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0438177686296022,
+      "learning_rate": 1.4893126314644106e-05,
+      "loss": 0.1845,
+      "step": 7023
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8025077624120668,
+      "learning_rate": 1.4891689913031928e-05,
+      "loss": 0.1915,
+      "step": 7024
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9815403441598226,
+      "learning_rate": 1.4890253378732385e-05,
+      "loss": 0.2024,
+      "step": 7025
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4646887196561633,
+      "learning_rate": 1.488881671178444e-05,
+      "loss": 0.2244,
+      "step": 7026
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8324670358442616,
+      "learning_rate": 1.4887379912227064e-05,
+      "loss": 0.2185,
+      "step": 7027
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.870578956635888,
+      "learning_rate": 1.4885942980099236e-05,
+      "loss": 0.2036,
+      "step": 7028
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3454427996276077,
+      "learning_rate": 1.488450591543993e-05,
+      "loss": 0.2091,
+      "step": 7029
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1991823657329022,
+      "learning_rate": 1.4883068718288121e-05,
+      "loss": 0.2294,
+      "step": 7030
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8558186090970756,
+      "learning_rate": 1.4881631388682797e-05,
+      "loss": 0.1883,
+      "step": 7031
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0901259681525386,
+      "learning_rate": 1.4880193926662948e-05,
+      "loss": 0.1977,
+      "step": 7032
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.000147735629142,
+      "learning_rate": 1.4878756332267563e-05,
+      "loss": 0.1937,
+      "step": 7033
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9307866687602493,
+      "learning_rate": 1.4877318605535638e-05,
+      "loss": 0.2057,
+      "step": 7034
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8674026920795987,
+      "learning_rate": 1.4875880746506169e-05,
+      "loss": 0.1858,
+      "step": 7035
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.066773643311902,
+      "learning_rate": 1.4874442755218156e-05,
+      "loss": 0.1948,
+      "step": 7036
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9281596268519601,
+      "learning_rate": 1.4873004631710613e-05,
+      "loss": 0.2231,
+      "step": 7037
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8973859838471487,
+      "learning_rate": 1.4871566376022542e-05,
+      "loss": 0.1794,
+      "step": 7038
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9926979536022584,
+      "learning_rate": 1.4870127988192957e-05,
+      "loss": 0.1973,
+      "step": 7039
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1506318008619696,
+      "learning_rate": 1.4868689468260876e-05,
+      "loss": 0.2115,
+      "step": 7040
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0734519175053021,
+      "learning_rate": 1.4867250816265318e-05,
+      "loss": 0.182,
+      "step": 7041
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.229578365298459,
+      "learning_rate": 1.4865812032245308e-05,
+      "loss": 0.1948,
+      "step": 7042
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9429939490887894,
+      "learning_rate": 1.4864373116239872e-05,
+      "loss": 0.2185,
+      "step": 7043
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1696130548769823,
+      "learning_rate": 1.486293406828804e-05,
+      "loss": 0.2244,
+      "step": 7044
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.382434322965532,
+      "learning_rate": 1.4861494888428845e-05,
+      "loss": 0.2003,
+      "step": 7045
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0071367527080957,
+      "learning_rate": 1.4860055576701327e-05,
+      "loss": 0.1995,
+      "step": 7046
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1758286360602521,
+      "learning_rate": 1.4858616133144527e-05,
+      "loss": 0.1906,
+      "step": 7047
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0913830281505044,
+      "learning_rate": 1.4857176557797493e-05,
+      "loss": 0.2076,
+      "step": 7048
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8585908518502955,
+      "learning_rate": 1.4855736850699268e-05,
+      "loss": 0.2014,
+      "step": 7049
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6561467985014158,
+      "learning_rate": 1.4854297011888905e-05,
+      "loss": 0.2059,
+      "step": 7050
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4893530541324937,
+      "learning_rate": 1.4852857041405466e-05,
+      "loss": 0.1821,
+      "step": 7051
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1291541762800112,
+      "learning_rate": 1.4851416939288003e-05,
+      "loss": 0.2142,
+      "step": 7052
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.889361943725798,
+      "learning_rate": 1.484997670557558e-05,
+      "loss": 0.1824,
+      "step": 7053
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8613533659766596,
+      "learning_rate": 1.4848536340307267e-05,
+      "loss": 0.2051,
+      "step": 7054
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3496139019332325,
+      "learning_rate": 1.4847095843522133e-05,
+      "loss": 0.2061,
+      "step": 7055
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8141814539341413,
+      "learning_rate": 1.4845655215259249e-05,
+      "loss": 0.1955,
+      "step": 7056
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7389541225733909,
+      "learning_rate": 1.4844214455557693e-05,
+      "loss": 0.1944,
+      "step": 7057
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.100671299271619,
+      "learning_rate": 1.4842773564456545e-05,
+      "loss": 0.2101,
+      "step": 7058
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1313577250220692,
+      "learning_rate": 1.4841332541994893e-05,
+      "loss": 0.2043,
+      "step": 7059
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7963336071239322,
+      "learning_rate": 1.4839891388211822e-05,
+      "loss": 0.2073,
+      "step": 7060
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9057394292716163,
+      "learning_rate": 1.4838450103146424e-05,
+      "loss": 0.2432,
+      "step": 7061
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.270303282906903,
+      "learning_rate": 1.4837008686837791e-05,
+      "loss": 0.2007,
+      "step": 7062
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8417121898936355,
+      "learning_rate": 1.4835567139325026e-05,
+      "loss": 0.2002,
+      "step": 7063
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8494513029708324,
+      "learning_rate": 1.4834125460647231e-05,
+      "loss": 0.1912,
+      "step": 7064
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1110096798803188,
+      "learning_rate": 1.483268365084351e-05,
+      "loss": 0.1917,
+      "step": 7065
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2050849762105256,
+      "learning_rate": 1.4831241709952969e-05,
+      "loss": 0.2051,
+      "step": 7066
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1217378012786416,
+      "learning_rate": 1.4829799638014724e-05,
+      "loss": 0.2045,
+      "step": 7067
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.118223071720989,
+      "learning_rate": 1.4828357435067895e-05,
+      "loss": 0.1838,
+      "step": 7068
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5884446354343718,
+      "learning_rate": 1.4826915101151595e-05,
+      "loss": 0.2154,
+      "step": 7069
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9053567421261107,
+      "learning_rate": 1.4825472636304952e-05,
+      "loss": 0.2375,
+      "step": 7070
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8367797883521595,
+      "learning_rate": 1.4824030040567086e-05,
+      "loss": 0.1871,
+      "step": 7071
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7442778069105056,
+      "learning_rate": 1.4822587313977137e-05,
+      "loss": 0.2197,
+      "step": 7072
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.46195344130919,
+      "learning_rate": 1.4821144456574235e-05,
+      "loss": 0.1843,
+      "step": 7073
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.871892278282872,
+      "learning_rate": 1.4819701468397516e-05,
+      "loss": 0.1998,
+      "step": 7074
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0590408819917725,
+      "learning_rate": 1.4818258349486121e-05,
+      "loss": 0.2003,
+      "step": 7075
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9773154856611629,
+      "learning_rate": 1.4816815099879199e-05,
+      "loss": 0.2079,
+      "step": 7076
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9642359721391925,
+      "learning_rate": 1.4815371719615895e-05,
+      "loss": 0.2001,
+      "step": 7077
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8066183899205053,
+      "learning_rate": 1.481392820873536e-05,
+      "loss": 0.1761,
+      "step": 7078
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0007716990939366,
+      "learning_rate": 1.481248456727675e-05,
+      "loss": 0.209,
+      "step": 7079
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8113832340307315,
+      "learning_rate": 1.4811040795279223e-05,
+      "loss": 0.1979,
+      "step": 7080
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0519165147720977,
+      "learning_rate": 1.4809596892781946e-05,
+      "loss": 0.2089,
+      "step": 7081
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8878676928357122,
+      "learning_rate": 1.480815285982408e-05,
+      "loss": 0.2352,
+      "step": 7082
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8365083305411455,
+      "learning_rate": 1.4806708696444796e-05,
+      "loss": 0.2044,
+      "step": 7083
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.128972835355977,
+      "learning_rate": 1.4805264402683268e-05,
+      "loss": 0.1973,
+      "step": 7084
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2933985370978633,
+      "learning_rate": 1.480381997857867e-05,
+      "loss": 0.2061,
+      "step": 7085
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1958606496129567,
+      "learning_rate": 1.4802375424170187e-05,
+      "loss": 0.2116,
+      "step": 7086
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9234574265240589,
+      "learning_rate": 1.4800930739497e-05,
+      "loss": 0.185,
+      "step": 7087
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9777658505483192,
+      "learning_rate": 1.4799485924598292e-05,
+      "loss": 0.2006,
+      "step": 7088
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8632732697046354,
+      "learning_rate": 1.4798040979513258e-05,
+      "loss": 0.1881,
+      "step": 7089
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0198595161320083,
+      "learning_rate": 1.479659590428109e-05,
+      "loss": 0.2127,
+      "step": 7090
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9215793081250065,
+      "learning_rate": 1.479515069894099e-05,
+      "loss": 0.1954,
+      "step": 7091
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9916694114763142,
+      "learning_rate": 1.4793705363532156e-05,
+      "loss": 0.1978,
+      "step": 7092
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.592385505546323,
+      "learning_rate": 1.4792259898093791e-05,
+      "loss": 0.2388,
+      "step": 7093
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.012073087478319,
+      "learning_rate": 1.4790814302665107e-05,
+      "loss": 0.1814,
+      "step": 7094
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.90806762365769,
+      "learning_rate": 1.4789368577285314e-05,
+      "loss": 0.2186,
+      "step": 7095
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8809231723802302,
+      "learning_rate": 1.478792272199363e-05,
+      "loss": 0.1912,
+      "step": 7096
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8225179914084455,
+      "learning_rate": 1.4786476736829267e-05,
+      "loss": 0.1968,
+      "step": 7097
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8506939817328405,
+      "learning_rate": 1.4785030621831458e-05,
+      "loss": 0.1949,
+      "step": 7098
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9884889449489223,
+      "learning_rate": 1.478358437703942e-05,
+      "loss": 0.2254,
+      "step": 7099
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.4120500445281,
+      "learning_rate": 1.4782138002492385e-05,
+      "loss": 0.2025,
+      "step": 7100
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6147739500037153,
+      "learning_rate": 1.4780691498229588e-05,
+      "loss": 0.1985,
+      "step": 7101
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9097814287498975,
+      "learning_rate": 1.4779244864290264e-05,
+      "loss": 0.2045,
+      "step": 7102
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8348803855026005,
+      "learning_rate": 1.4777798100713655e-05,
+      "loss": 0.1769,
+      "step": 7103
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1900370878471898,
+      "learning_rate": 1.4776351207538999e-05,
+      "loss": 0.1719,
+      "step": 7104
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2091092328565474,
+      "learning_rate": 1.477490418480555e-05,
+      "loss": 0.1898,
+      "step": 7105
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5838612601289892,
+      "learning_rate": 1.4773457032552551e-05,
+      "loss": 0.224,
+      "step": 7106
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.945232541245546,
+      "learning_rate": 1.4772009750819262e-05,
+      "loss": 0.2107,
+      "step": 7107
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9322743971599436,
+      "learning_rate": 1.4770562339644943e-05,
+      "loss": 0.235,
+      "step": 7108
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1461959180016132,
+      "learning_rate": 1.4769114799068847e-05,
+      "loss": 0.2051,
+      "step": 7109
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.636216515301017,
+      "learning_rate": 1.4767667129130243e-05,
+      "loss": 0.221,
+      "step": 7110
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1442129928861824,
+      "learning_rate": 1.4766219329868399e-05,
+      "loss": 0.2033,
+      "step": 7111
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4749380356676163,
+      "learning_rate": 1.4764771401322588e-05,
+      "loss": 0.1876,
+      "step": 7112
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.982928862953428,
+      "learning_rate": 1.4763323343532083e-05,
+      "loss": 0.2089,
+      "step": 7113
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0080157854821021,
+      "learning_rate": 1.4761875156536163e-05,
+      "loss": 0.2059,
+      "step": 7114
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9390907781578757,
+      "learning_rate": 1.476042684037411e-05,
+      "loss": 0.1792,
+      "step": 7115
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.299468734945118,
+      "learning_rate": 1.4758978395085208e-05,
+      "loss": 0.2316,
+      "step": 7116
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9578609983641396,
+      "learning_rate": 1.4757529820708754e-05,
+      "loss": 0.1964,
+      "step": 7117
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8416583906906582,
+      "learning_rate": 1.475608111728403e-05,
+      "loss": 0.1789,
+      "step": 7118
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9457542076436503,
+      "learning_rate": 1.4754632284850338e-05,
+      "loss": 0.1833,
+      "step": 7119
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9545702827475283,
+      "learning_rate": 1.4753183323446978e-05,
+      "loss": 0.2101,
+      "step": 7120
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.088974962874592,
+      "learning_rate": 1.4751734233113253e-05,
+      "loss": 0.2124,
+      "step": 7121
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9677299958915025,
+      "learning_rate": 1.4750285013888466e-05,
+      "loss": 0.1812,
+      "step": 7122
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9383390825244804,
+      "learning_rate": 1.474883566581193e-05,
+      "loss": 0.2187,
+      "step": 7123
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.049802052114001,
+      "learning_rate": 1.474738618892296e-05,
+      "loss": 0.1903,
+      "step": 7124
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2918584280404934,
+      "learning_rate": 1.474593658326087e-05,
+      "loss": 0.2075,
+      "step": 7125
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2190451902433268,
+      "learning_rate": 1.4744486848864982e-05,
+      "loss": 0.1998,
+      "step": 7126
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2539579091445217,
+      "learning_rate": 1.4743036985774621e-05,
+      "loss": 0.1809,
+      "step": 7127
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8069144816251373,
+      "learning_rate": 1.4741586994029113e-05,
+      "loss": 0.1841,
+      "step": 7128
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8739653517671883,
+      "learning_rate": 1.474013687366779e-05,
+      "loss": 0.1951,
+      "step": 7129
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1976902309195656,
+      "learning_rate": 1.4738686624729987e-05,
+      "loss": 0.2206,
+      "step": 7130
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1363599177482622,
+      "learning_rate": 1.4737236247255045e-05,
+      "loss": 0.1928,
+      "step": 7131
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0141472201673492,
+      "learning_rate": 1.4735785741282298e-05,
+      "loss": 0.17,
+      "step": 7132
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8962760373086162,
+      "learning_rate": 1.4734335106851095e-05,
+      "loss": 0.1788,
+      "step": 7133
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0975108161546672,
+      "learning_rate": 1.4732884344000787e-05,
+      "loss": 0.1932,
+      "step": 7134
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1304931506411495,
+      "learning_rate": 1.4731433452770723e-05,
+      "loss": 0.2232,
+      "step": 7135
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9684346870572593,
+      "learning_rate": 1.4729982433200261e-05,
+      "loss": 0.2124,
+      "step": 7136
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5665638986844876,
+      "learning_rate": 1.4728531285328753e-05,
+      "loss": 0.2219,
+      "step": 7137
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9854542831804197,
+      "learning_rate": 1.4727080009195573e-05,
+      "loss": 0.217,
+      "step": 7138
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0710552426739544,
+      "learning_rate": 1.4725628604840078e-05,
+      "loss": 0.1855,
+      "step": 7139
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2343554376103933,
+      "learning_rate": 1.4724177072301642e-05,
+      "loss": 0.2069,
+      "step": 7140
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9514401751415055,
+      "learning_rate": 1.4722725411619634e-05,
+      "loss": 0.2106,
+      "step": 7141
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.030571956714932,
+      "learning_rate": 1.4721273622833432e-05,
+      "loss": 0.2027,
+      "step": 7142
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.506294984385844,
+      "learning_rate": 1.4719821705982417e-05,
+      "loss": 0.199,
+      "step": 7143
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.258818755205757,
+      "learning_rate": 1.4718369661105973e-05,
+      "loss": 0.2058,
+      "step": 7144
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7594735060195563,
+      "learning_rate": 1.4716917488243485e-05,
+      "loss": 0.1839,
+      "step": 7145
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3661682393111336,
+      "learning_rate": 1.4715465187434342e-05,
+      "loss": 0.1978,
+      "step": 7146
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1330714487070268,
+      "learning_rate": 1.4714012758717941e-05,
+      "loss": 0.1889,
+      "step": 7147
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1189407523486754,
+      "learning_rate": 1.4712560202133679e-05,
+      "loss": 0.2008,
+      "step": 7148
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8707259683031279,
+      "learning_rate": 1.4711107517720953e-05,
+      "loss": 0.1833,
+      "step": 7149
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.05214004080036,
+      "learning_rate": 1.4709654705519168e-05,
+      "loss": 0.1987,
+      "step": 7150
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0682777254050864,
+      "learning_rate": 1.4708201765567736e-05,
+      "loss": 0.1798,
+      "step": 7151
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1839164935435411,
+      "learning_rate": 1.4706748697906065e-05,
+      "loss": 0.201,
+      "step": 7152
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1184231051281768,
+      "learning_rate": 1.4705295502573571e-05,
+      "loss": 0.2206,
+      "step": 7153
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8979461790109566,
+      "learning_rate": 1.4703842179609668e-05,
+      "loss": 0.2269,
+      "step": 7154
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9264038467836124,
+      "learning_rate": 1.470238872905378e-05,
+      "loss": 0.1965,
+      "step": 7155
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8079661378752513,
+      "learning_rate": 1.4700935150945334e-05,
+      "loss": 0.1951,
+      "step": 7156
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1541144569815773,
+      "learning_rate": 1.4699481445323757e-05,
+      "loss": 0.2105,
+      "step": 7157
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1023098748718994,
+      "learning_rate": 1.4698027612228478e-05,
+      "loss": 0.2134,
+      "step": 7158
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9886838350178068,
+      "learning_rate": 1.4696573651698937e-05,
+      "loss": 0.2033,
+      "step": 7159
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6959608514363341,
+      "learning_rate": 1.4695119563774568e-05,
+      "loss": 0.1985,
+      "step": 7160
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 43.03118587420249,
+      "learning_rate": 1.4693665348494819e-05,
+      "loss": 0.1765,
+      "step": 7161
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0170267614935629,
+      "learning_rate": 1.469221100589913e-05,
+      "loss": 0.182,
+      "step": 7162
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7638325075184338,
+      "learning_rate": 1.4690756536026952e-05,
+      "loss": 0.2022,
+      "step": 7163
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7929466280136379,
+      "learning_rate": 1.4689301938917737e-05,
+      "loss": 0.1787,
+      "step": 7164
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9428273508130095,
+      "learning_rate": 1.4687847214610944e-05,
+      "loss": 0.1829,
+      "step": 7165
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7559421200980139,
+      "learning_rate": 1.4686392363146032e-05,
+      "loss": 0.2028,
+      "step": 7166
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9443670684991925,
+      "learning_rate": 1.468493738456246e-05,
+      "loss": 0.1837,
+      "step": 7167
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.187792744550687,
+      "learning_rate": 1.4683482278899696e-05,
+      "loss": 0.2238,
+      "step": 7168
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1056454767564257,
+      "learning_rate": 1.4682027046197214e-05,
+      "loss": 0.2076,
+      "step": 7169
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8500674701813452,
+      "learning_rate": 1.4680571686494483e-05,
+      "loss": 0.2032,
+      "step": 7170
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1124369387671862,
+      "learning_rate": 1.4679116199830978e-05,
+      "loss": 0.1817,
+      "step": 7171
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0194354860544301,
+      "learning_rate": 1.4677660586246183e-05,
+      "loss": 0.1875,
+      "step": 7172
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8772144588619847,
+      "learning_rate": 1.4676204845779585e-05,
+      "loss": 0.1882,
+      "step": 7173
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.169638257110541,
+      "learning_rate": 1.4674748978470663e-05,
+      "loss": 0.195,
+      "step": 7174
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0328090611192982,
+      "learning_rate": 1.4673292984358911e-05,
+      "loss": 0.1922,
+      "step": 7175
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8585369331116066,
+      "learning_rate": 1.4671836863483819e-05,
+      "loss": 0.1723,
+      "step": 7176
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9757881472890122,
+      "learning_rate": 1.4670380615884891e-05,
+      "loss": 0.2182,
+      "step": 7177
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1618572034347405,
+      "learning_rate": 1.4668924241601627e-05,
+      "loss": 0.1893,
+      "step": 7178
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8359477577026576,
+      "learning_rate": 1.4667467740673528e-05,
+      "loss": 0.1752,
+      "step": 7179
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9593541875986683,
+      "learning_rate": 1.4666011113140103e-05,
+      "loss": 0.2085,
+      "step": 7180
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9373030226818541,
+      "learning_rate": 1.4664554359040862e-05,
+      "loss": 0.2049,
+      "step": 7181
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.046339481942067,
+      "learning_rate": 1.4663097478415322e-05,
+      "loss": 0.2115,
+      "step": 7182
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.16317388913559,
+      "learning_rate": 1.4661640471302998e-05,
+      "loss": 0.1874,
+      "step": 7183
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0277480391584366,
+      "learning_rate": 1.4660183337743414e-05,
+      "loss": 0.2128,
+      "step": 7184
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.18453723334063,
+      "learning_rate": 1.4658726077776093e-05,
+      "loss": 0.2066,
+      "step": 7185
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0996444774192848,
+      "learning_rate": 1.4657268691440564e-05,
+      "loss": 0.2115,
+      "step": 7186
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8886067711511231,
+      "learning_rate": 1.465581117877636e-05,
+      "loss": 0.178,
+      "step": 7187
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.113528221312272,
+      "learning_rate": 1.4654353539823014e-05,
+      "loss": 0.2113,
+      "step": 7188
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.355610201304584,
+      "learning_rate": 1.4652895774620066e-05,
+      "loss": 0.1913,
+      "step": 7189
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9662037049135782,
+      "learning_rate": 1.4651437883207056e-05,
+      "loss": 0.2023,
+      "step": 7190
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.791090091633652,
+      "learning_rate": 1.4649979865623531e-05,
+      "loss": 0.1935,
+      "step": 7191
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1973119684181808,
+      "learning_rate": 1.4648521721909042e-05,
+      "loss": 0.1807,
+      "step": 7192
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7438397230367796,
+      "learning_rate": 1.4647063452103135e-05,
+      "loss": 0.1938,
+      "step": 7193
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.017059128027557,
+      "learning_rate": 1.464560505624537e-05,
+      "loss": 0.1893,
+      "step": 7194
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1394480960376436,
+      "learning_rate": 1.4644146534375307e-05,
+      "loss": 0.1966,
+      "step": 7195
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.309035780184552,
+      "learning_rate": 1.4642687886532507e-05,
+      "loss": 0.2063,
+      "step": 7196
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.965556266585576,
+      "learning_rate": 1.4641229112756537e-05,
+      "loss": 0.2061,
+      "step": 7197
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9247343728821459,
+      "learning_rate": 1.4639770213086962e-05,
+      "loss": 0.2257,
+      "step": 7198
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2873322184303597,
+      "learning_rate": 1.463831118756336e-05,
+      "loss": 0.2219,
+      "step": 7199
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0503873999197977,
+      "learning_rate": 1.4636852036225304e-05,
+      "loss": 0.1835,
+      "step": 7200
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0441262383963452,
+      "learning_rate": 1.4635392759112374e-05,
+      "loss": 0.2196,
+      "step": 7201
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0886323583800523,
+      "learning_rate": 1.4633933356264156e-05,
+      "loss": 0.2264,
+      "step": 7202
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0740117152730773,
+      "learning_rate": 1.463247382772023e-05,
+      "loss": 0.2016,
+      "step": 7203
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0431229186759934,
+      "learning_rate": 1.4631014173520192e-05,
+      "loss": 0.2064,
+      "step": 7204
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2250889849637991,
+      "learning_rate": 1.4629554393703635e-05,
+      "loss": 0.1874,
+      "step": 7205
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8985830906857448,
+      "learning_rate": 1.4628094488310154e-05,
+      "loss": 0.1906,
+      "step": 7206
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4827841710717786,
+      "learning_rate": 1.4626634457379343e-05,
+      "loss": 0.1823,
+      "step": 7207
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3135064026868895,
+      "learning_rate": 1.4625174300950817e-05,
+      "loss": 0.2345,
+      "step": 7208
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3767976555094654,
+      "learning_rate": 1.4623714019064178e-05,
+      "loss": 0.2239,
+      "step": 7209
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0399025276719043,
+      "learning_rate": 1.462225361175903e-05,
+      "loss": 0.1845,
+      "step": 7210
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1365818325865469,
+      "learning_rate": 1.4620793079074991e-05,
+      "loss": 0.2074,
+      "step": 7211
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9650778862329886,
+      "learning_rate": 1.4619332421051682e-05,
+      "loss": 0.201,
+      "step": 7212
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2317824187188442,
+      "learning_rate": 1.4617871637728719e-05,
+      "loss": 0.181,
+      "step": 7213
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2293569861227824,
+      "learning_rate": 1.461641072914573e-05,
+      "loss": 0.1788,
+      "step": 7214
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8421056403657364,
+      "learning_rate": 1.4614949695342335e-05,
+      "loss": 0.1826,
+      "step": 7215
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7556245059003615,
+      "learning_rate": 1.461348853635817e-05,
+      "loss": 0.198,
+      "step": 7216
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8373865110950348,
+      "learning_rate": 1.4612027252232868e-05,
+      "loss": 0.1869,
+      "step": 7217
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4859909383682117,
+      "learning_rate": 1.4610565843006066e-05,
+      "loss": 0.2279,
+      "step": 7218
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9840411718276176,
+      "learning_rate": 1.4609104308717405e-05,
+      "loss": 0.2095,
+      "step": 7219
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0813751383332257,
+      "learning_rate": 1.4607642649406529e-05,
+      "loss": 0.1955,
+      "step": 7220
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7086729803488199,
+      "learning_rate": 1.4606180865113087e-05,
+      "loss": 0.1809,
+      "step": 7221
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0543148531001396,
+      "learning_rate": 1.460471895587673e-05,
+      "loss": 0.2351,
+      "step": 7222
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5590508133007468,
+      "learning_rate": 1.460325692173711e-05,
+      "loss": 0.1929,
+      "step": 7223
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8575470786430182,
+      "learning_rate": 1.4601794762733885e-05,
+      "loss": 0.2414,
+      "step": 7224
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0713855039589257,
+      "learning_rate": 1.4600332478906718e-05,
+      "loss": 0.191,
+      "step": 7225
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.050489799066373,
+      "learning_rate": 1.4598870070295274e-05,
+      "loss": 0.1957,
+      "step": 7226
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1241824559481621,
+      "learning_rate": 1.4597407536939221e-05,
+      "loss": 0.2044,
+      "step": 7227
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9398946171930076,
+      "learning_rate": 1.4595944878878226e-05,
+      "loss": 0.1788,
+      "step": 7228
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6923161481460482,
+      "learning_rate": 1.4594482096151965e-05,
+      "loss": 0.1869,
+      "step": 7229
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8537422935352744,
+      "learning_rate": 1.459301918880012e-05,
+      "loss": 0.1784,
+      "step": 7230
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.251240074898674,
+      "learning_rate": 1.459155615686237e-05,
+      "loss": 0.2119,
+      "step": 7231
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.04468001598618,
+      "learning_rate": 1.45900930003784e-05,
+      "loss": 0.2061,
+      "step": 7232
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9481835045186339,
+      "learning_rate": 1.4588629719387895e-05,
+      "loss": 0.1806,
+      "step": 7233
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5470941140754693,
+      "learning_rate": 1.4587166313930551e-05,
+      "loss": 0.1995,
+      "step": 7234
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8879328653548417,
+      "learning_rate": 1.4585702784046065e-05,
+      "loss": 0.2003,
+      "step": 7235
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9206448905493988,
+      "learning_rate": 1.4584239129774131e-05,
+      "loss": 0.1888,
+      "step": 7236
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7014245751046675,
+      "learning_rate": 1.458277535115445e-05,
+      "loss": 0.1778,
+      "step": 7237
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0059919959825143,
+      "learning_rate": 1.458131144822673e-05,
+      "loss": 0.2028,
+      "step": 7238
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0593151433111,
+      "learning_rate": 1.4579847421030677e-05,
+      "loss": 0.2188,
+      "step": 7239
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4445554920908814,
+      "learning_rate": 1.4578383269606004e-05,
+      "loss": 0.1781,
+      "step": 7240
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.774630230057031,
+      "learning_rate": 1.4576918993992429e-05,
+      "loss": 0.1994,
+      "step": 7241
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.711497024864439,
+      "learning_rate": 1.4575454594229666e-05,
+      "loss": 0.2623,
+      "step": 7242
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8612693862757297,
+      "learning_rate": 1.4573990070357437e-05,
+      "loss": 0.1703,
+      "step": 7243
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0058030680566827,
+      "learning_rate": 1.4572525422415471e-05,
+      "loss": 0.2049,
+      "step": 7244
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.913388143288835,
+      "learning_rate": 1.4571060650443494e-05,
+      "loss": 0.197,
+      "step": 7245
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4343812327113545,
+      "learning_rate": 1.4569595754481238e-05,
+      "loss": 0.1961,
+      "step": 7246
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2043942073468759,
+      "learning_rate": 1.456813073456844e-05,
+      "loss": 0.1734,
+      "step": 7247
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0221013480787766,
+      "learning_rate": 1.4566665590744838e-05,
+      "loss": 0.1844,
+      "step": 7248
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.427492512904666,
+      "learning_rate": 1.4565200323050173e-05,
+      "loss": 0.1967,
+      "step": 7249
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8977229493298299,
+      "learning_rate": 1.4563734931524191e-05,
+      "loss": 0.197,
+      "step": 7250
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8534301150679654,
+      "learning_rate": 1.4562269416206642e-05,
+      "loss": 0.1963,
+      "step": 7251
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7761348402379218,
+      "learning_rate": 1.4560803777137279e-05,
+      "loss": 0.195,
+      "step": 7252
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1073974709757195,
+      "learning_rate": 1.4559338014355853e-05,
+      "loss": 0.2045,
+      "step": 7253
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.488909966122852,
+      "learning_rate": 1.455787212790213e-05,
+      "loss": 0.2125,
+      "step": 7254
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8852613049754224,
+      "learning_rate": 1.4556406117815864e-05,
+      "loss": 0.1959,
+      "step": 7255
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8339723125115669,
+      "learning_rate": 1.4554939984136824e-05,
+      "loss": 0.1803,
+      "step": 7256
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9529550397471396,
+      "learning_rate": 1.4553473726904783e-05,
+      "loss": 0.223,
+      "step": 7257
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8851176434942727,
+      "learning_rate": 1.4552007346159509e-05,
+      "loss": 0.1759,
+      "step": 7258
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8007076173251486,
+      "learning_rate": 1.4550540841940778e-05,
+      "loss": 0.2015,
+      "step": 7259
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9714155323978105,
+      "learning_rate": 1.4549074214288368e-05,
+      "loss": 0.1795,
+      "step": 7260
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6302081131518966,
+      "learning_rate": 1.4547607463242068e-05,
+      "loss": 0.1857,
+      "step": 7261
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1714822203917645,
+      "learning_rate": 1.4546140588841656e-05,
+      "loss": 0.2201,
+      "step": 7262
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3467129135182294,
+      "learning_rate": 1.4544673591126924e-05,
+      "loss": 0.1882,
+      "step": 7263
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0569269794229985,
+      "learning_rate": 1.4543206470137663e-05,
+      "loss": 0.2448,
+      "step": 7264
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2157216517942167,
+      "learning_rate": 1.4541739225913669e-05,
+      "loss": 0.2044,
+      "step": 7265
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9933802128857785,
+      "learning_rate": 1.4540271858494746e-05,
+      "loss": 0.2042,
+      "step": 7266
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.933445012615465,
+      "learning_rate": 1.453880436792069e-05,
+      "loss": 0.2293,
+      "step": 7267
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7724878186469613,
+      "learning_rate": 1.4537336754231307e-05,
+      "loss": 0.1888,
+      "step": 7268
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4015165482196508,
+      "learning_rate": 1.453586901746641e-05,
+      "loss": 0.1998,
+      "step": 7269
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8703397411856586,
+      "learning_rate": 1.4534401157665812e-05,
+      "loss": 0.2205,
+      "step": 7270
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0065068650542348,
+      "learning_rate": 1.4532933174869323e-05,
+      "loss": 0.1992,
+      "step": 7271
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3699495130815498,
+      "learning_rate": 1.4531465069116771e-05,
+      "loss": 0.2028,
+      "step": 7272
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0793924999933862,
+      "learning_rate": 1.4529996840447966e-05,
+      "loss": 0.2349,
+      "step": 7273
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7948066903330879,
+      "learning_rate": 1.4528528488902745e-05,
+      "loss": 0.1809,
+      "step": 7274
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1284683175111199,
+      "learning_rate": 1.4527060014520932e-05,
+      "loss": 0.217,
+      "step": 7275
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5284300188468116,
+      "learning_rate": 1.452559141734236e-05,
+      "loss": 0.2062,
+      "step": 7276
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.781343543319652,
+      "learning_rate": 1.4524122697406866e-05,
+      "loss": 0.1962,
+      "step": 7277
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.295639290670492,
+      "learning_rate": 1.452265385475429e-05,
+      "loss": 0.212,
+      "step": 7278
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.740549444374038,
+      "learning_rate": 1.4521184889424472e-05,
+      "loss": 0.2036,
+      "step": 7279
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0011987948570245,
+      "learning_rate": 1.4519715801457256e-05,
+      "loss": 0.1967,
+      "step": 7280
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5341229496916164,
+      "learning_rate": 1.4518246590892493e-05,
+      "loss": 0.2037,
+      "step": 7281
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0225133863288018,
+      "learning_rate": 1.4516777257770039e-05,
+      "loss": 0.1988,
+      "step": 7282
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8600004079706661,
+      "learning_rate": 1.4515307802129746e-05,
+      "loss": 0.1886,
+      "step": 7283
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8580445077526663,
+      "learning_rate": 1.4513838224011474e-05,
+      "loss": 0.1931,
+      "step": 7284
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.277280462094834,
+      "learning_rate": 1.4512368523455085e-05,
+      "loss": 0.2025,
+      "step": 7285
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.00902764493686,
+      "learning_rate": 1.4510898700500442e-05,
+      "loss": 0.2112,
+      "step": 7286
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9961082386523968,
+      "learning_rate": 1.450942875518742e-05,
+      "loss": 0.1924,
+      "step": 7287
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0077175637689286,
+      "learning_rate": 1.4507958687555887e-05,
+      "loss": 0.1927,
+      "step": 7288
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.064534723234433,
+      "learning_rate": 1.4506488497645724e-05,
+      "loss": 0.1952,
+      "step": 7289
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8798686625193131,
+      "learning_rate": 1.4505018185496802e-05,
+      "loss": 0.2013,
+      "step": 7290
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9306270609813555,
+      "learning_rate": 1.4503547751149007e-05,
+      "loss": 0.2206,
+      "step": 7291
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1938661547775145,
+      "learning_rate": 1.4502077194642229e-05,
+      "loss": 0.2108,
+      "step": 7292
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0111927132329457,
+      "learning_rate": 1.4500606516016347e-05,
+      "loss": 0.1839,
+      "step": 7293
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2037406245910078,
+      "learning_rate": 1.4499135715311262e-05,
+      "loss": 0.212,
+      "step": 7294
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6647718656816444,
+      "learning_rate": 1.4497664792566865e-05,
+      "loss": 0.224,
+      "step": 7295
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9636072832329631,
+      "learning_rate": 1.4496193747823062e-05,
+      "loss": 0.2038,
+      "step": 7296
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8220322234345318,
+      "learning_rate": 1.4494722581119748e-05,
+      "loss": 0.1819,
+      "step": 7297
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1926677975279927,
+      "learning_rate": 1.4493251292496826e-05,
+      "loss": 0.1974,
+      "step": 7298
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2505678449152033,
+      "learning_rate": 1.4491779881994208e-05,
+      "loss": 0.2185,
+      "step": 7299
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.339102115123226,
+      "learning_rate": 1.4490308349651812e-05,
+      "loss": 0.1729,
+      "step": 7300
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3449780056843892,
+      "learning_rate": 1.4488836695509545e-05,
+      "loss": 0.1973,
+      "step": 7301
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0480233615897767,
+      "learning_rate": 1.4487364919607331e-05,
+      "loss": 0.2261,
+      "step": 7302
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1627938006870673,
+      "learning_rate": 1.4485893021985091e-05,
+      "loss": 0.1835,
+      "step": 7303
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9359446400184624,
+      "learning_rate": 1.4484421002682745e-05,
+      "loss": 0.2001,
+      "step": 7304
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7920756240041322,
+      "learning_rate": 1.4482948861740229e-05,
+      "loss": 0.2011,
+      "step": 7305
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.998299872861058,
+      "learning_rate": 1.4481476599197473e-05,
+      "loss": 0.2082,
+      "step": 7306
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9021048208921083,
+      "learning_rate": 1.448000421509441e-05,
+      "loss": 0.2173,
+      "step": 7307
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8386435457194779,
+      "learning_rate": 1.4478531709470978e-05,
+      "loss": 0.2075,
+      "step": 7308
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4749227890431893,
+      "learning_rate": 1.4477059082367122e-05,
+      "loss": 0.1917,
+      "step": 7309
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9811174959203954,
+      "learning_rate": 1.4475586333822787e-05,
+      "loss": 0.1978,
+      "step": 7310
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4198098792649345,
+      "learning_rate": 1.4474113463877917e-05,
+      "loss": 0.2122,
+      "step": 7311
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9695964071852331,
+      "learning_rate": 1.4472640472572468e-05,
+      "loss": 0.2082,
+      "step": 7312
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9142693265376828,
+      "learning_rate": 1.4471167359946394e-05,
+      "loss": 0.186,
+      "step": 7313
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0171477584067625,
+      "learning_rate": 1.4469694126039653e-05,
+      "loss": 0.1787,
+      "step": 7314
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.016606679527248,
+      "learning_rate": 1.4468220770892208e-05,
+      "loss": 0.1863,
+      "step": 7315
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.969808532614358,
+      "learning_rate": 1.4466747294544017e-05,
+      "loss": 0.2009,
+      "step": 7316
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.820909974683755,
+      "learning_rate": 1.4465273697035055e-05,
+      "loss": 0.2328,
+      "step": 7317
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.894778322942877,
+      "learning_rate": 1.4463799978405295e-05,
+      "loss": 0.2129,
+      "step": 7318
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4698423912176912,
+      "learning_rate": 1.4462326138694706e-05,
+      "loss": 0.2089,
+      "step": 7319
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8868102604174576,
+      "learning_rate": 1.446085217794327e-05,
+      "loss": 0.1753,
+      "step": 7320
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1381290790671303,
+      "learning_rate": 1.4459378096190966e-05,
+      "loss": 0.2157,
+      "step": 7321
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9477793579680197,
+      "learning_rate": 1.4457903893477779e-05,
+      "loss": 0.2068,
+      "step": 7322
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0369651684066326,
+      "learning_rate": 1.4456429569843698e-05,
+      "loss": 0.1826,
+      "step": 7323
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8862359245446289,
+      "learning_rate": 1.4454955125328711e-05,
+      "loss": 0.2096,
+      "step": 7324
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.971297388987907,
+      "learning_rate": 1.4453480559972817e-05,
+      "loss": 0.2092,
+      "step": 7325
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.621655362230059,
+      "learning_rate": 1.4452005873816009e-05,
+      "loss": 0.1852,
+      "step": 7326
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9555106190526966,
+      "learning_rate": 1.4450531066898293e-05,
+      "loss": 0.1971,
+      "step": 7327
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1937414917284117,
+      "learning_rate": 1.4449056139259667e-05,
+      "loss": 0.2201,
+      "step": 7328
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8844007950976225,
+      "learning_rate": 1.4447581090940144e-05,
+      "loss": 0.2079,
+      "step": 7329
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3966735678188713,
+      "learning_rate": 1.4446105921979731e-05,
+      "loss": 0.1648,
+      "step": 7330
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8794422204306628,
+      "learning_rate": 1.4444630632418449e-05,
+      "loss": 0.1935,
+      "step": 7331
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1053488098602837,
+      "learning_rate": 1.4443155222296305e-05,
+      "loss": 0.2236,
+      "step": 7332
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.165254773028472,
+      "learning_rate": 1.4441679691653327e-05,
+      "loss": 0.1892,
+      "step": 7333
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9169400364770195,
+      "learning_rate": 1.4440204040529536e-05,
+      "loss": 0.1849,
+      "step": 7334
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8791850419992676,
+      "learning_rate": 1.4438728268964956e-05,
+      "loss": 0.195,
+      "step": 7335
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8901616149988154,
+      "learning_rate": 1.4437252376999627e-05,
+      "loss": 0.2114,
+      "step": 7336
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0122521359050152,
+      "learning_rate": 1.4435776364673573e-05,
+      "loss": 0.195,
+      "step": 7337
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9392493572957487,
+      "learning_rate": 1.4434300232026837e-05,
+      "loss": 0.1802,
+      "step": 7338
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1768578635257316,
+      "learning_rate": 1.4432823979099453e-05,
+      "loss": 0.2006,
+      "step": 7339
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3795244554400317,
+      "learning_rate": 1.443134760593147e-05,
+      "loss": 0.1662,
+      "step": 7340
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.452143408625368,
+      "learning_rate": 1.4429871112562935e-05,
+      "loss": 0.197,
+      "step": 7341
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1602637874632054,
+      "learning_rate": 1.4428394499033893e-05,
+      "loss": 0.2098,
+      "step": 7342
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4394723268569924,
+      "learning_rate": 1.4426917765384398e-05,
+      "loss": 0.2087,
+      "step": 7343
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0905020016637779,
+      "learning_rate": 1.4425440911654514e-05,
+      "loss": 0.207,
+      "step": 7344
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5623935818671604,
+      "learning_rate": 1.4423963937884293e-05,
+      "loss": 0.2205,
+      "step": 7345
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9808215660267777,
+      "learning_rate": 1.44224868441138e-05,
+      "loss": 0.1919,
+      "step": 7346
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0041036859515802,
+      "learning_rate": 1.4421009630383103e-05,
+      "loss": 0.18,
+      "step": 7347
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1655556834955962,
+      "learning_rate": 1.4419532296732271e-05,
+      "loss": 0.1819,
+      "step": 7348
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.073691610451434,
+      "learning_rate": 1.4418054843201373e-05,
+      "loss": 0.1892,
+      "step": 7349
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9877252828329729,
+      "learning_rate": 1.441657726983049e-05,
+      "loss": 0.195,
+      "step": 7350
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0427460053130362,
+      "learning_rate": 1.4415099576659698e-05,
+      "loss": 0.1944,
+      "step": 7351
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9397173058898916,
+      "learning_rate": 1.4413621763729077e-05,
+      "loss": 0.1876,
+      "step": 7352
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1310771497695997,
+      "learning_rate": 1.4412143831078722e-05,
+      "loss": 0.1863,
+      "step": 7353
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.048926621899362,
+      "learning_rate": 1.4410665778748714e-05,
+      "loss": 0.2142,
+      "step": 7354
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9894402322478005,
+      "learning_rate": 1.4409187606779149e-05,
+      "loss": 0.1851,
+      "step": 7355
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9000909516825687,
+      "learning_rate": 1.4407709315210117e-05,
+      "loss": 0.1865,
+      "step": 7356
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6959787873830762,
+      "learning_rate": 1.4406230904081724e-05,
+      "loss": 0.194,
+      "step": 7357
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5086135648888905,
+      "learning_rate": 1.440475237343407e-05,
+      "loss": 0.2017,
+      "step": 7358
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5819638309039437,
+      "learning_rate": 1.4403273723307259e-05,
+      "loss": 0.1928,
+      "step": 7359
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6196587202106798,
+      "learning_rate": 1.4401794953741397e-05,
+      "loss": 0.1949,
+      "step": 7360
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0728889712252914,
+      "learning_rate": 1.4400316064776598e-05,
+      "loss": 0.2076,
+      "step": 7361
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9184986881056747,
+      "learning_rate": 1.4398837056452979e-05,
+      "loss": 0.1919,
+      "step": 7362
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8316174688989357,
+      "learning_rate": 1.4397357928810657e-05,
+      "loss": 0.1754,
+      "step": 7363
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7761411361598949,
+      "learning_rate": 1.4395878681889753e-05,
+      "loss": 0.2109,
+      "step": 7364
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2223321694261715,
+      "learning_rate": 1.4394399315730389e-05,
+      "loss": 0.2328,
+      "step": 7365
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9737362109650243,
+      "learning_rate": 1.4392919830372698e-05,
+      "loss": 0.2028,
+      "step": 7366
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9702903964399952,
+      "learning_rate": 1.4391440225856807e-05,
+      "loss": 0.1712,
+      "step": 7367
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7614903580983567,
+      "learning_rate": 1.4389960502222855e-05,
+      "loss": 0.2001,
+      "step": 7368
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9259848994936244,
+      "learning_rate": 1.438848065951097e-05,
+      "loss": 0.2102,
+      "step": 7369
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.48112954255448,
+      "learning_rate": 1.4387000697761305e-05,
+      "loss": 0.193,
+      "step": 7370
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8531334505763309,
+      "learning_rate": 1.4385520617013998e-05,
+      "loss": 0.1966,
+      "step": 7371
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.103475362674807,
+      "learning_rate": 1.4384040417309194e-05,
+      "loss": 0.1842,
+      "step": 7372
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.266035539089789,
+      "learning_rate": 1.4382560098687045e-05,
+      "loss": 0.2053,
+      "step": 7373
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0031143586674998,
+      "learning_rate": 1.4381079661187708e-05,
+      "loss": 0.2169,
+      "step": 7374
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9897718835102313,
+      "learning_rate": 1.4379599104851336e-05,
+      "loss": 0.2133,
+      "step": 7375
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.726912981667751,
+      "learning_rate": 1.4378118429718093e-05,
+      "loss": 0.249,
+      "step": 7376
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.84419251677018,
+      "learning_rate": 1.4376637635828142e-05,
+      "loss": 0.1989,
+      "step": 7377
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9721913264285105,
+      "learning_rate": 1.4375156723221642e-05,
+      "loss": 0.2163,
+      "step": 7378
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0242439576936002,
+      "learning_rate": 1.4373675691938773e-05,
+      "loss": 0.2078,
+      "step": 7379
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9130983342073931,
+      "learning_rate": 1.4372194542019705e-05,
+      "loss": 0.2066,
+      "step": 7380
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3429044399629129,
+      "learning_rate": 1.4370713273504611e-05,
+      "loss": 0.1946,
+      "step": 7381
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2895851595974883,
+      "learning_rate": 1.4369231886433672e-05,
+      "loss": 0.1807,
+      "step": 7382
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0427998515920929,
+      "learning_rate": 1.4367750380847073e-05,
+      "loss": 0.2046,
+      "step": 7383
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7801428574537339,
+      "learning_rate": 1.4366268756784998e-05,
+      "loss": 0.1846,
+      "step": 7384
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8523927910703909,
+      "learning_rate": 1.4364787014287636e-05,
+      "loss": 0.1846,
+      "step": 7385
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8105319148802839,
+      "learning_rate": 1.4363305153395179e-05,
+      "loss": 0.1764,
+      "step": 7386
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.856027812134765,
+      "learning_rate": 1.4361823174147822e-05,
+      "loss": 0.2052,
+      "step": 7387
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8132340069486842,
+      "learning_rate": 1.4360341076585769e-05,
+      "loss": 0.1954,
+      "step": 7388
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7088325614833517,
+      "learning_rate": 1.4358858860749213e-05,
+      "loss": 0.1911,
+      "step": 7389
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8891267655937457,
+      "learning_rate": 1.4357376526678368e-05,
+      "loss": 0.1839,
+      "step": 7390
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.866720734186058,
+      "learning_rate": 1.4355894074413436e-05,
+      "loss": 0.2069,
+      "step": 7391
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0826656456304604,
+      "learning_rate": 1.4354411503994634e-05,
+      "loss": 0.2116,
+      "step": 7392
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.064708131590027,
+      "learning_rate": 1.4352928815462175e-05,
+      "loss": 0.1811,
+      "step": 7393
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8589484404006575,
+      "learning_rate": 1.4351446008856274e-05,
+      "loss": 0.2032,
+      "step": 7394
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7668618139828199,
+      "learning_rate": 1.4349963084217154e-05,
+      "loss": 0.2154,
+      "step": 7395
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7697956911015835,
+      "learning_rate": 1.4348480041585037e-05,
+      "loss": 0.2101,
+      "step": 7396
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8783925945149618,
+      "learning_rate": 1.4346996881000157e-05,
+      "loss": 0.193,
+      "step": 7397
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.150747005100924,
+      "learning_rate": 1.434551360250274e-05,
+      "loss": 0.1899,
+      "step": 7398
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1226882074944422,
+      "learning_rate": 1.4344030206133022e-05,
+      "loss": 0.214,
+      "step": 7399
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8692486363623312,
+      "learning_rate": 1.4342546691931238e-05,
+      "loss": 0.2153,
+      "step": 7400
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4976160493577089,
+      "learning_rate": 1.4341063059937631e-05,
+      "loss": 0.2022,
+      "step": 7401
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1188129385862706,
+      "learning_rate": 1.4339579310192444e-05,
+      "loss": 0.2345,
+      "step": 7402
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0964171334407278,
+      "learning_rate": 1.4338095442735923e-05,
+      "loss": 0.1857,
+      "step": 7403
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8696443304840087,
+      "learning_rate": 1.4336611457608314e-05,
+      "loss": 0.2192,
+      "step": 7404
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1151258175613268,
+      "learning_rate": 1.4335127354849876e-05,
+      "loss": 0.1896,
+      "step": 7405
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8802160194360952,
+      "learning_rate": 1.4333643134500865e-05,
+      "loss": 0.2021,
+      "step": 7406
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8063957735402743,
+      "learning_rate": 1.433215879660154e-05,
+      "loss": 0.1948,
+      "step": 7407
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.913986916751216,
+      "learning_rate": 1.4330674341192163e-05,
+      "loss": 0.1951,
+      "step": 7408
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3042370977547812,
+      "learning_rate": 1.4329189768312997e-05,
+      "loss": 0.2044,
+      "step": 7409
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.18897831691075,
+      "learning_rate": 1.4327705078004317e-05,
+      "loss": 0.1884,
+      "step": 7410
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.80509128784862,
+      "learning_rate": 1.432622027030639e-05,
+      "loss": 0.1911,
+      "step": 7411
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1516519194253798,
+      "learning_rate": 1.4324735345259495e-05,
+      "loss": 0.1702,
+      "step": 7412
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8693490250888909,
+      "learning_rate": 1.432325030290391e-05,
+      "loss": 0.1905,
+      "step": 7413
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8508701142949681,
+      "learning_rate": 1.4321765143279916e-05,
+      "loss": 0.2042,
+      "step": 7414
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2434167042103295,
+      "learning_rate": 1.4320279866427798e-05,
+      "loss": 0.2234,
+      "step": 7415
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8106162299147703,
+      "learning_rate": 1.4318794472387845e-05,
+      "loss": 0.1855,
+      "step": 7416
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8147454401274455,
+      "learning_rate": 1.4317308961200347e-05,
+      "loss": 0.1982,
+      "step": 7417
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7589531970409737,
+      "learning_rate": 1.43158233329056e-05,
+      "loss": 0.1847,
+      "step": 7418
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7345539558339856,
+      "learning_rate": 1.4314337587543903e-05,
+      "loss": 0.1827,
+      "step": 7419
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8228984501748001,
+      "learning_rate": 1.4312851725155554e-05,
+      "loss": 0.1878,
+      "step": 7420
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7492280721147175,
+      "learning_rate": 1.431136574578086e-05,
+      "loss": 0.1696,
+      "step": 7421
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1691536457309282,
+      "learning_rate": 1.4309879649460123e-05,
+      "loss": 0.1944,
+      "step": 7422
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9063338491403092,
+      "learning_rate": 1.4308393436233658e-05,
+      "loss": 0.1912,
+      "step": 7423
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2322521307302403,
+      "learning_rate": 1.4306907106141781e-05,
+      "loss": 0.1909,
+      "step": 7424
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9291985895172377,
+      "learning_rate": 1.4305420659224802e-05,
+      "loss": 0.2008,
+      "step": 7425
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8540370909551723,
+      "learning_rate": 1.4303934095523046e-05,
+      "loss": 0.1881,
+      "step": 7426
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8181487956916195,
+      "learning_rate": 1.4302447415076835e-05,
+      "loss": 0.2017,
+      "step": 7427
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8370007506643188,
+      "learning_rate": 1.4300960617926495e-05,
+      "loss": 0.192,
+      "step": 7428
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6955940772303095,
+      "learning_rate": 1.4299473704112355e-05,
+      "loss": 0.2156,
+      "step": 7429
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9342504930900565,
+      "learning_rate": 1.4297986673674747e-05,
+      "loss": 0.2549,
+      "step": 7430
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9147917475970455,
+      "learning_rate": 1.4296499526654004e-05,
+      "loss": 0.1773,
+      "step": 7431
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8086678880363469,
+      "learning_rate": 1.4295012263090475e-05,
+      "loss": 0.1908,
+      "step": 7432
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.95526409863462,
+      "learning_rate": 1.4293524883024494e-05,
+      "loss": 0.1924,
+      "step": 7433
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9898514284623698,
+      "learning_rate": 1.4292037386496407e-05,
+      "loss": 0.1975,
+      "step": 7434
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.214407578688893,
+      "learning_rate": 1.4290549773546565e-05,
+      "loss": 0.2102,
+      "step": 7435
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7862349311310997,
+      "learning_rate": 1.4289062044215318e-05,
+      "loss": 0.1952,
+      "step": 7436
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.805684277750261,
+      "learning_rate": 1.428757419854302e-05,
+      "loss": 0.2077,
+      "step": 7437
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.52754589046096,
+      "learning_rate": 1.428608623657003e-05,
+      "loss": 0.2141,
+      "step": 7438
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.995220102864857,
+      "learning_rate": 1.4284598158336707e-05,
+      "loss": 0.2145,
+      "step": 7439
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9295949768900353,
+      "learning_rate": 1.428310996388342e-05,
+      "loss": 0.2259,
+      "step": 7440
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9462065479377103,
+      "learning_rate": 1.428162165325053e-05,
+      "loss": 0.2235,
+      "step": 7441
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2948090680155522,
+      "learning_rate": 1.4280133226478413e-05,
+      "loss": 0.1929,
+      "step": 7442
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4418156783883276,
+      "learning_rate": 1.4278644683607442e-05,
+      "loss": 0.1942,
+      "step": 7443
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8066543175705367,
+      "learning_rate": 1.4277156024677987e-05,
+      "loss": 0.1654,
+      "step": 7444
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.230279856593095,
+      "learning_rate": 1.4275667249730437e-05,
+      "loss": 0.2238,
+      "step": 7445
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1038156894093265,
+      "learning_rate": 1.427417835880517e-05,
+      "loss": 0.2287,
+      "step": 7446
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.218749043552778,
+      "learning_rate": 1.4272689351942577e-05,
+      "loss": 0.2005,
+      "step": 7447
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5229360109931684,
+      "learning_rate": 1.4271200229183043e-05,
+      "loss": 0.1836,
+      "step": 7448
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0477126003594448,
+      "learning_rate": 1.4269710990566958e-05,
+      "loss": 0.2061,
+      "step": 7449
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.009328444666866,
+      "learning_rate": 1.4268221636134722e-05,
+      "loss": 0.2418,
+      "step": 7450
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8603144680485548,
+      "learning_rate": 1.4266732165926735e-05,
+      "loss": 0.2214,
+      "step": 7451
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0262286597197119,
+      "learning_rate": 1.4265242579983398e-05,
+      "loss": 0.2136,
+      "step": 7452
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7806878375610516,
+      "learning_rate": 1.4263752878345112e-05,
+      "loss": 0.2169,
+      "step": 7453
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1225825958989664,
+      "learning_rate": 1.4262263061052291e-05,
+      "loss": 0.2022,
+      "step": 7454
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6108452092479106,
+      "learning_rate": 1.4260773128145341e-05,
+      "loss": 0.1909,
+      "step": 7455
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0173183086801756,
+      "learning_rate": 1.4259283079664683e-05,
+      "loss": 0.1895,
+      "step": 7456
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8439453311015034,
+      "learning_rate": 1.4257792915650728e-05,
+      "loss": 0.1973,
+      "step": 7457
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8723570887290385,
+      "learning_rate": 1.42563026361439e-05,
+      "loss": 0.2065,
+      "step": 7458
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8536901531933275,
+      "learning_rate": 1.4254812241184623e-05,
+      "loss": 0.2027,
+      "step": 7459
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3662775011425288,
+      "learning_rate": 1.4253321730813326e-05,
+      "loss": 0.2275,
+      "step": 7460
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.97442449862382,
+      "learning_rate": 1.4251831105070433e-05,
+      "loss": 0.2079,
+      "step": 7461
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.790465580849618,
+      "learning_rate": 1.4250340363996382e-05,
+      "loss": 0.2051,
+      "step": 7462
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.093983192420429,
+      "learning_rate": 1.4248849507631608e-05,
+      "loss": 0.212,
+      "step": 7463
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9556791044852044,
+      "learning_rate": 1.4247358536016554e-05,
+      "loss": 0.2081,
+      "step": 7464
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5223824451295664,
+      "learning_rate": 1.424586744919166e-05,
+      "loss": 0.1855,
+      "step": 7465
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8448789582666036,
+      "learning_rate": 1.4244376247197367e-05,
+      "loss": 0.2186,
+      "step": 7466
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3207283654711193,
+      "learning_rate": 1.424288493007413e-05,
+      "loss": 0.2033,
+      "step": 7467
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0601494536216605,
+      "learning_rate": 1.4241393497862403e-05,
+      "loss": 0.2069,
+      "step": 7468
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0956312757128686,
+      "learning_rate": 1.4239901950602635e-05,
+      "loss": 0.2081,
+      "step": 7469
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2581936092687398,
+      "learning_rate": 1.4238410288335289e-05,
+      "loss": 0.1978,
+      "step": 7470
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9771948394045026,
+      "learning_rate": 1.4236918511100824e-05,
+      "loss": 0.2054,
+      "step": 7471
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.166621037901537,
+      "learning_rate": 1.4235426618939704e-05,
+      "loss": 0.2156,
+      "step": 7472
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9622643795882899,
+      "learning_rate": 1.4233934611892399e-05,
+      "loss": 0.2164,
+      "step": 7473
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.004021903210207,
+      "learning_rate": 1.423244248999938e-05,
+      "loss": 0.1907,
+      "step": 7474
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0349109753246033,
+      "learning_rate": 1.4230950253301115e-05,
+      "loss": 0.1946,
+      "step": 7475
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.343381097620725,
+      "learning_rate": 1.4229457901838087e-05,
+      "loss": 0.2022,
+      "step": 7476
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.08645471980482,
+      "learning_rate": 1.4227965435650774e-05,
+      "loss": 0.1711,
+      "step": 7477
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.376681948697426,
+      "learning_rate": 1.4226472854779663e-05,
+      "loss": 0.1741,
+      "step": 7478
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8444694354214273,
+      "learning_rate": 1.4224980159265234e-05,
+      "loss": 0.1995,
+      "step": 7479
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.358410968788333,
+      "learning_rate": 1.422348734914798e-05,
+      "loss": 0.1953,
+      "step": 7480
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0823915381063276,
+      "learning_rate": 1.4221994424468395e-05,
+      "loss": 0.1951,
+      "step": 7481
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2856834550347218,
+      "learning_rate": 1.4220501385266971e-05,
+      "loss": 0.205,
+      "step": 7482
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.185673366244246,
+      "learning_rate": 1.4219008231584211e-05,
+      "loss": 0.205,
+      "step": 7483
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0602670135534036,
+      "learning_rate": 1.421751496346061e-05,
+      "loss": 0.1888,
+      "step": 7484
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1431011201509098,
+      "learning_rate": 1.4216021580936681e-05,
+      "loss": 0.201,
+      "step": 7485
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9083813915428518,
+      "learning_rate": 1.4214528084052925e-05,
+      "loss": 0.1979,
+      "step": 7486
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0573463288335716,
+      "learning_rate": 1.421303447284986e-05,
+      "loss": 0.1886,
+      "step": 7487
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9609320810574705,
+      "learning_rate": 1.4211540747367998e-05,
+      "loss": 0.1885,
+      "step": 7488
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9505170150612685,
+      "learning_rate": 1.4210046907647852e-05,
+      "loss": 0.2028,
+      "step": 7489
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1493563606221087,
+      "learning_rate": 1.4208552953729949e-05,
+      "loss": 0.2101,
+      "step": 7490
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9009289958930716,
+      "learning_rate": 1.4207058885654807e-05,
+      "loss": 0.2003,
+      "step": 7491
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0713540112615993,
+      "learning_rate": 1.4205564703462955e-05,
+      "loss": 0.1874,
+      "step": 7492
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0702896376829447,
+      "learning_rate": 1.4204070407194923e-05,
+      "loss": 0.1953,
+      "step": 7493
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.272473715175916,
+      "learning_rate": 1.4202575996891246e-05,
+      "loss": 0.2053,
+      "step": 7494
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.908677908811022,
+      "learning_rate": 1.4201081472592455e-05,
+      "loss": 0.2103,
+      "step": 7495
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.013295489272494,
+      "learning_rate": 1.4199586834339093e-05,
+      "loss": 0.2058,
+      "step": 7496
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0451142798524864,
+      "learning_rate": 1.41980920821717e-05,
+      "loss": 0.1923,
+      "step": 7497
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.086382225473405,
+      "learning_rate": 1.4196597216130823e-05,
+      "loss": 0.1879,
+      "step": 7498
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0869633424813556,
+      "learning_rate": 1.4195102236257011e-05,
+      "loss": 0.2132,
+      "step": 7499
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0450687978973898,
+      "learning_rate": 1.4193607142590812e-05,
+      "loss": 0.2081,
+      "step": 7500
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1569443333472407,
+      "learning_rate": 1.4192111935172781e-05,
+      "loss": 0.221,
+      "step": 7501
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0329301579206815,
+      "learning_rate": 1.419061661404348e-05,
+      "loss": 0.1791,
+      "step": 7502
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6107073747509522,
+      "learning_rate": 1.4189121179243466e-05,
+      "loss": 0.189,
+      "step": 7503
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9123903187240583,
+      "learning_rate": 1.4187625630813303e-05,
+      "loss": 0.1986,
+      "step": 7504
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9802098693497492,
+      "learning_rate": 1.4186129968793558e-05,
+      "loss": 0.1743,
+      "step": 7505
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0183773359868546,
+      "learning_rate": 1.41846341932248e-05,
+      "loss": 0.1797,
+      "step": 7506
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0375210887014656,
+      "learning_rate": 1.4183138304147605e-05,
+      "loss": 0.1949,
+      "step": 7507
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9658665195692431,
+      "learning_rate": 1.4181642301602547e-05,
+      "loss": 0.195,
+      "step": 7508
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3377407285594807,
+      "learning_rate": 1.4180146185630205e-05,
+      "loss": 0.2411,
+      "step": 7509
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2619605983747744,
+      "learning_rate": 1.4178649956271162e-05,
+      "loss": 0.1842,
+      "step": 7510
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.96968818167038,
+      "learning_rate": 1.4177153613566002e-05,
+      "loss": 0.191,
+      "step": 7511
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.992558708079268,
+      "learning_rate": 1.4175657157555316e-05,
+      "loss": 0.182,
+      "step": 7512
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4435494809485996,
+      "learning_rate": 1.4174160588279692e-05,
+      "loss": 0.2136,
+      "step": 7513
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0321323815625674,
+      "learning_rate": 1.4172663905779725e-05,
+      "loss": 0.198,
+      "step": 7514
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7564634893716371,
+      "learning_rate": 1.4171167110096017e-05,
+      "loss": 0.1783,
+      "step": 7515
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9330135482607076,
+      "learning_rate": 1.4169670201269164e-05,
+      "loss": 0.2053,
+      "step": 7516
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8472369283082383,
+      "learning_rate": 1.4168173179339772e-05,
+      "loss": 0.2047,
+      "step": 7517
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8801592059554315,
+      "learning_rate": 1.4166676044348448e-05,
+      "loss": 0.1955,
+      "step": 7518
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0766442544705805,
+      "learning_rate": 1.4165178796335797e-05,
+      "loss": 0.2025,
+      "step": 7519
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0174652520991234,
+      "learning_rate": 1.416368143534244e-05,
+      "loss": 0.23,
+      "step": 7520
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9339941915171122,
+      "learning_rate": 1.4162183961408987e-05,
+      "loss": 0.1897,
+      "step": 7521
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5105807693516058,
+      "learning_rate": 1.416068637457606e-05,
+      "loss": 0.1781,
+      "step": 7522
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.935675378996467,
+      "learning_rate": 1.4159188674884279e-05,
+      "loss": 0.2052,
+      "step": 7523
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1427958146169708,
+      "learning_rate": 1.4157690862374272e-05,
+      "loss": 0.2039,
+      "step": 7524
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.658788262602051,
+      "learning_rate": 1.4156192937086666e-05,
+      "loss": 0.2107,
+      "step": 7525
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8955541412189186,
+      "learning_rate": 1.4154694899062089e-05,
+      "loss": 0.1832,
+      "step": 7526
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8889059275984288,
+      "learning_rate": 1.4153196748341179e-05,
+      "loss": 0.2379,
+      "step": 7527
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0487778302161304,
+      "learning_rate": 1.4151698484964574e-05,
+      "loss": 0.2044,
+      "step": 7528
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8812194186270116,
+      "learning_rate": 1.4150200108972912e-05,
+      "loss": 0.2,
+      "step": 7529
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9193206661664284,
+      "learning_rate": 1.4148701620406838e-05,
+      "loss": 0.2098,
+      "step": 7530
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.993980966928357,
+      "learning_rate": 1.4147203019306997e-05,
+      "loss": 0.2048,
+      "step": 7531
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0750132796267666,
+      "learning_rate": 1.4145704305714038e-05,
+      "loss": 0.1998,
+      "step": 7532
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6834923657312354,
+      "learning_rate": 1.414420547966862e-05,
+      "loss": 0.1738,
+      "step": 7533
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7890938616586272,
+      "learning_rate": 1.4142706541211392e-05,
+      "loss": 0.1867,
+      "step": 7534
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2978997036990871,
+      "learning_rate": 1.4141207490383018e-05,
+      "loss": 0.1919,
+      "step": 7535
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8117457987197362,
+      "learning_rate": 1.4139708327224155e-05,
+      "loss": 0.1937,
+      "step": 7536
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9688092437532801,
+      "learning_rate": 1.4138209051775467e-05,
+      "loss": 0.2102,
+      "step": 7537
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0092214002394286,
+      "learning_rate": 1.4136709664077628e-05,
+      "loss": 0.2067,
+      "step": 7538
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2446739747004254,
+      "learning_rate": 1.4135210164171306e-05,
+      "loss": 0.2075,
+      "step": 7539
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.409950331815906,
+      "learning_rate": 1.4133710552097175e-05,
+      "loss": 0.1752,
+      "step": 7540
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0317598910736085,
+      "learning_rate": 1.4132210827895909e-05,
+      "loss": 0.2115,
+      "step": 7541
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.363377833518173,
+      "learning_rate": 1.4130710991608194e-05,
+      "loss": 0.2003,
+      "step": 7542
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0121446264043472,
+      "learning_rate": 1.4129211043274709e-05,
+      "loss": 0.2075,
+      "step": 7543
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8417773315163373,
+      "learning_rate": 1.412771098293614e-05,
+      "loss": 0.1946,
+      "step": 7544
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1855268084816621,
+      "learning_rate": 1.4126210810633176e-05,
+      "loss": 0.2152,
+      "step": 7545
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0237124307200967,
+      "learning_rate": 1.4124710526406516e-05,
+      "loss": 0.2443,
+      "step": 7546
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.404831522108842,
+      "learning_rate": 1.4123210130296845e-05,
+      "loss": 0.198,
+      "step": 7547
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9074529008026088,
+      "learning_rate": 1.412170962234487e-05,
+      "loss": 0.1943,
+      "step": 7548
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.51360435231819,
+      "learning_rate": 1.4120209002591286e-05,
+      "loss": 0.1871,
+      "step": 7549
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9224528569596793,
+      "learning_rate": 1.41187082710768e-05,
+      "loss": 0.1955,
+      "step": 7550
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8130885991407517,
+      "learning_rate": 1.4117207427842122e-05,
+      "loss": 0.1914,
+      "step": 7551
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0662624440238975,
+      "learning_rate": 1.4115706472927957e-05,
+      "loss": 0.1817,
+      "step": 7552
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8750154796571464,
+      "learning_rate": 1.4114205406375025e-05,
+      "loss": 0.2128,
+      "step": 7553
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0251582509921566,
+      "learning_rate": 1.4112704228224034e-05,
+      "loss": 0.1807,
+      "step": 7554
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8656304735082289,
+      "learning_rate": 1.4111202938515711e-05,
+      "loss": 0.1855,
+      "step": 7555
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9354978653471507,
+      "learning_rate": 1.4109701537290779e-05,
+      "loss": 0.2093,
+      "step": 7556
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9342037452734095,
+      "learning_rate": 1.4108200024589958e-05,
+      "loss": 0.1875,
+      "step": 7557
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9043537891505412,
+      "learning_rate": 1.410669840045398e-05,
+      "loss": 0.1928,
+      "step": 7558
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0488505805359303,
+      "learning_rate": 1.4105196664923577e-05,
+      "loss": 0.2078,
+      "step": 7559
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0669685487941905,
+      "learning_rate": 1.4103694818039483e-05,
+      "loss": 0.1933,
+      "step": 7560
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0754201019931622,
+      "learning_rate": 1.4102192859842432e-05,
+      "loss": 0.2111,
+      "step": 7561
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0524796337983948,
+      "learning_rate": 1.4100690790373168e-05,
+      "loss": 0.1853,
+      "step": 7562
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.023500297719706,
+      "learning_rate": 1.4099188609672436e-05,
+      "loss": 0.1698,
+      "step": 7563
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6937197254418842,
+      "learning_rate": 1.409768631778098e-05,
+      "loss": 0.1687,
+      "step": 7564
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9742004715333842,
+      "learning_rate": 1.4096183914739554e-05,
+      "loss": 0.1947,
+      "step": 7565
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.920578932705953,
+      "learning_rate": 1.4094681400588908e-05,
+      "loss": 0.1946,
+      "step": 7566
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3660496533422637,
+      "learning_rate": 1.4093178775369793e-05,
+      "loss": 0.2007,
+      "step": 7567
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8202394736784563,
+      "learning_rate": 1.4091676039122977e-05,
+      "loss": 0.2181,
+      "step": 7568
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8823324782020567,
+      "learning_rate": 1.4090173191889216e-05,
+      "loss": 0.2063,
+      "step": 7569
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6719778905366667,
+      "learning_rate": 1.4088670233709278e-05,
+      "loss": 0.185,
+      "step": 7570
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9286681884469269,
+      "learning_rate": 1.4087167164623927e-05,
+      "loss": 0.1838,
+      "step": 7571
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9178992253331234,
+      "learning_rate": 1.4085663984673936e-05,
+      "loss": 0.2011,
+      "step": 7572
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.867270369064487,
+      "learning_rate": 1.408416069390008e-05,
+      "loss": 0.1782,
+      "step": 7573
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6760773325114068,
+      "learning_rate": 1.4082657292343131e-05,
+      "loss": 0.2006,
+      "step": 7574
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9197042191937326,
+      "learning_rate": 1.4081153780043878e-05,
+      "loss": 0.1803,
+      "step": 7575
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7566894677669311,
+      "learning_rate": 1.4079650157043095e-05,
+      "loss": 0.2011,
+      "step": 7576
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9803458625908985,
+      "learning_rate": 1.4078146423381571e-05,
+      "loss": 0.2276,
+      "step": 7577
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8948601920458561,
+      "learning_rate": 1.4076642579100095e-05,
+      "loss": 0.1957,
+      "step": 7578
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8251175785770059,
+      "learning_rate": 1.407513862423946e-05,
+      "loss": 0.2001,
+      "step": 7579
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8159217586434515,
+      "learning_rate": 1.4073634558840454e-05,
+      "loss": 0.1887,
+      "step": 7580
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.712292700290352,
+      "learning_rate": 1.4072130382943888e-05,
+      "loss": 0.1882,
+      "step": 7581
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0144574994086635,
+      "learning_rate": 1.4070626096590552e-05,
+      "loss": 0.2054,
+      "step": 7582
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8214830372217938,
+      "learning_rate": 1.4069121699821253e-05,
+      "loss": 0.1965,
+      "step": 7583
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0635809609229483,
+      "learning_rate": 1.4067617192676799e-05,
+      "loss": 0.1768,
+      "step": 7584
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9959854507722363,
+      "learning_rate": 1.4066112575197996e-05,
+      "loss": 0.1972,
+      "step": 7585
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0418149972153863,
+      "learning_rate": 1.4064607847425664e-05,
+      "loss": 0.216,
+      "step": 7586
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9272614314221803,
+      "learning_rate": 1.4063103009400613e-05,
+      "loss": 0.2134,
+      "step": 7587
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8282259836531397,
+      "learning_rate": 1.4061598061163664e-05,
+      "loss": 0.2053,
+      "step": 7588
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.518892077046803,
+      "learning_rate": 1.4060093002755636e-05,
+      "loss": 0.21,
+      "step": 7589
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.872535797315508,
+      "learning_rate": 1.4058587834217356e-05,
+      "loss": 0.1814,
+      "step": 7590
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2812768967861146,
+      "learning_rate": 1.4057082555589654e-05,
+      "loss": 0.2118,
+      "step": 7591
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8252964740636697,
+      "learning_rate": 1.4055577166913358e-05,
+      "loss": 0.2015,
+      "step": 7592
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0118958760314445,
+      "learning_rate": 1.4054071668229304e-05,
+      "loss": 0.207,
+      "step": 7593
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8168769002372456,
+      "learning_rate": 1.4052566059578326e-05,
+      "loss": 0.1981,
+      "step": 7594
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9911446249748135,
+      "learning_rate": 1.4051060341001264e-05,
+      "loss": 0.2068,
+      "step": 7595
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8867221481455925,
+      "learning_rate": 1.4049554512538958e-05,
+      "loss": 0.1939,
+      "step": 7596
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8235464707735634,
+      "learning_rate": 1.4048048574232261e-05,
+      "loss": 0.1808,
+      "step": 7597
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8495631625786931,
+      "learning_rate": 1.4046542526122018e-05,
+      "loss": 0.2044,
+      "step": 7598
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0312588114857115,
+      "learning_rate": 1.4045036368249079e-05,
+      "loss": 0.1923,
+      "step": 7599
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2257480499644202,
+      "learning_rate": 1.4043530100654301e-05,
+      "loss": 0.2095,
+      "step": 7600
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9347192728461339,
+      "learning_rate": 1.4042023723378541e-05,
+      "loss": 0.1793,
+      "step": 7601
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8903951969647127,
+      "learning_rate": 1.4040517236462656e-05,
+      "loss": 0.1929,
+      "step": 7602
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3242813549316574,
+      "learning_rate": 1.4039010639947516e-05,
+      "loss": 0.1972,
+      "step": 7603
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.201248666535143,
+      "learning_rate": 1.4037503933873984e-05,
+      "loss": 0.2101,
+      "step": 7604
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1902203553720785,
+      "learning_rate": 1.403599711828293e-05,
+      "loss": 0.2285,
+      "step": 7605
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9544624554727649,
+      "learning_rate": 1.4034490193215224e-05,
+      "loss": 0.1957,
+      "step": 7606
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0195966094164068,
+      "learning_rate": 1.4032983158711744e-05,
+      "loss": 0.2003,
+      "step": 7607
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3276408942198594,
+      "learning_rate": 1.403147601481337e-05,
+      "loss": 0.2001,
+      "step": 7608
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7275677793246561,
+      "learning_rate": 1.4029968761560979e-05,
+      "loss": 0.1805,
+      "step": 7609
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8479951597951434,
+      "learning_rate": 1.402846139899546e-05,
+      "loss": 0.2048,
+      "step": 7610
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.838782389886137,
+      "learning_rate": 1.4026953927157698e-05,
+      "loss": 0.2025,
+      "step": 7611
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8453036827999479,
+      "learning_rate": 1.4025446346088582e-05,
+      "loss": 0.1987,
+      "step": 7612
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0457773409872628,
+      "learning_rate": 1.402393865582901e-05,
+      "loss": 0.1991,
+      "step": 7613
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8654771992999472,
+      "learning_rate": 1.4022430856419872e-05,
+      "loss": 0.1999,
+      "step": 7614
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1394541387656685,
+      "learning_rate": 1.4020922947902067e-05,
+      "loss": 0.2206,
+      "step": 7615
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8091480970897881,
+      "learning_rate": 1.4019414930316501e-05,
+      "loss": 0.1888,
+      "step": 7616
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0685495245375725,
+      "learning_rate": 1.4017906803704083e-05,
+      "loss": 0.1874,
+      "step": 7617
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.872612439298588,
+      "learning_rate": 1.401639856810571e-05,
+      "loss": 0.1935,
+      "step": 7618
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1548406631959833,
+      "learning_rate": 1.4014890223562303e-05,
+      "loss": 0.1862,
+      "step": 7619
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9949755985513199,
+      "learning_rate": 1.401338177011477e-05,
+      "loss": 0.1795,
+      "step": 7620
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9879919215858408,
+      "learning_rate": 1.401187320780403e-05,
+      "loss": 0.1913,
+      "step": 7621
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8082593760487321,
+      "learning_rate": 1.4010364536671004e-05,
+      "loss": 0.1858,
+      "step": 7622
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4274036461103985,
+      "learning_rate": 1.4008855756756612e-05,
+      "loss": 0.2293,
+      "step": 7623
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9867689521147746,
+      "learning_rate": 1.4007346868101779e-05,
+      "loss": 0.1839,
+      "step": 7624
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9102121091949064,
+      "learning_rate": 1.4005837870747439e-05,
+      "loss": 0.2006,
+      "step": 7625
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.854600912758212,
+      "learning_rate": 1.400432876473452e-05,
+      "loss": 0.1996,
+      "step": 7626
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2096446917432433,
+      "learning_rate": 1.4002819550103958e-05,
+      "loss": 0.1817,
+      "step": 7627
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0627923072240306,
+      "learning_rate": 1.4001310226896689e-05,
+      "loss": 0.2394,
+      "step": 7628
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2724119703874543,
+      "learning_rate": 1.3999800795153652e-05,
+      "loss": 0.186,
+      "step": 7629
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9999657164938784,
+      "learning_rate": 1.3998291254915797e-05,
+      "loss": 0.197,
+      "step": 7630
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2678985627552222,
+      "learning_rate": 1.3996781606224063e-05,
+      "loss": 0.1817,
+      "step": 7631
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9871813158684251,
+      "learning_rate": 1.3995271849119403e-05,
+      "loss": 0.2123,
+      "step": 7632
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7917965541568505,
+      "learning_rate": 1.3993761983642765e-05,
+      "loss": 0.1854,
+      "step": 7633
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.948713022523547,
+      "learning_rate": 1.3992252009835112e-05,
+      "loss": 0.1835,
+      "step": 7634
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.013801688872813,
+      "learning_rate": 1.3990741927737395e-05,
+      "loss": 0.1914,
+      "step": 7635
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0338027271382328,
+      "learning_rate": 1.3989231737390578e-05,
+      "loss": 0.2086,
+      "step": 7636
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0038262656980823,
+      "learning_rate": 1.3987721438835626e-05,
+      "loss": 0.2164,
+      "step": 7637
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.198776705998106,
+      "learning_rate": 1.3986211032113504e-05,
+      "loss": 0.2259,
+      "step": 7638
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1559578172326987,
+      "learning_rate": 1.3984700517265184e-05,
+      "loss": 0.2053,
+      "step": 7639
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9478082394204406,
+      "learning_rate": 1.3983189894331636e-05,
+      "loss": 0.2075,
+      "step": 7640
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0319933891708784,
+      "learning_rate": 1.3981679163353839e-05,
+      "loss": 0.1778,
+      "step": 7641
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2467961556378275,
+      "learning_rate": 1.3980168324372768e-05,
+      "loss": 0.1881,
+      "step": 7642
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.097771450032399,
+      "learning_rate": 1.3978657377429405e-05,
+      "loss": 0.2074,
+      "step": 7643
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7948181272258855,
+      "learning_rate": 1.3977146322564739e-05,
+      "loss": 0.1797,
+      "step": 7644
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6906764323928702,
+      "learning_rate": 1.3975635159819757e-05,
+      "loss": 0.1994,
+      "step": 7645
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8555549281608942,
+      "learning_rate": 1.3974123889235437e-05,
+      "loss": 0.1963,
+      "step": 7646
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8152633474642856,
+      "learning_rate": 1.3972612510852789e-05,
+      "loss": 0.2028,
+      "step": 7647
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.834483546768694,
+      "learning_rate": 1.3971101024712803e-05,
+      "loss": 0.2009,
+      "step": 7648
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8059419858590291,
+      "learning_rate": 1.3969589430856476e-05,
+      "loss": 0.1705,
+      "step": 7649
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8663444660848391,
+      "learning_rate": 1.3968077729324808e-05,
+      "loss": 0.2038,
+      "step": 7650
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9524959304972902,
+      "learning_rate": 1.3966565920158811e-05,
+      "loss": 0.1794,
+      "step": 7651
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9974758966169943,
+      "learning_rate": 1.3965054003399493e-05,
+      "loss": 0.1747,
+      "step": 7652
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0143608736447438,
+      "learning_rate": 1.3963541979087855e-05,
+      "loss": 0.1919,
+      "step": 7653
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8770771941762959,
+      "learning_rate": 1.396202984726492e-05,
+      "loss": 0.2121,
+      "step": 7654
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.058384880625591,
+      "learning_rate": 1.3960517607971697e-05,
+      "loss": 0.2058,
+      "step": 7655
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2266740375032155,
+      "learning_rate": 1.3959005261249217e-05,
+      "loss": 0.1884,
+      "step": 7656
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.803399173151223,
+      "learning_rate": 1.3957492807138491e-05,
+      "loss": 0.1858,
+      "step": 7657
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5027341436903714,
+      "learning_rate": 1.3955980245680551e-05,
+      "loss": 0.2012,
+      "step": 7658
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7537433177825293,
+      "learning_rate": 1.3954467576916422e-05,
+      "loss": 0.197,
+      "step": 7659
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0281275927478755,
+      "learning_rate": 1.3952954800887137e-05,
+      "loss": 0.1985,
+      "step": 7660
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.4007636048539807,
+      "learning_rate": 1.3951441917633732e-05,
+      "loss": 0.2119,
+      "step": 7661
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5874162488104218,
+      "learning_rate": 1.394992892719724e-05,
+      "loss": 0.1999,
+      "step": 7662
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.90732612942027,
+      "learning_rate": 1.3948415829618705e-05,
+      "loss": 0.1906,
+      "step": 7663
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8673104556121487,
+      "learning_rate": 1.3946902624939164e-05,
+      "loss": 0.2082,
+      "step": 7664
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.037346088119156,
+      "learning_rate": 1.3945389313199669e-05,
+      "loss": 0.2019,
+      "step": 7665
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.123917949735145,
+      "learning_rate": 1.3943875894441265e-05,
+      "loss": 0.2053,
+      "step": 7666
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9346108283378147,
+      "learning_rate": 1.3942362368705007e-05,
+      "loss": 0.1752,
+      "step": 7667
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8821888566223758,
+      "learning_rate": 1.3940848736031943e-05,
+      "loss": 0.1883,
+      "step": 7668
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0537674571428322,
+      "learning_rate": 1.3939334996463136e-05,
+      "loss": 0.2164,
+      "step": 7669
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0721094239860878,
+      "learning_rate": 1.3937821150039645e-05,
+      "loss": 0.2313,
+      "step": 7670
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8856670502546947,
+      "learning_rate": 1.3936307196802532e-05,
+      "loss": 0.1929,
+      "step": 7671
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.111960198508083,
+      "learning_rate": 1.3934793136792863e-05,
+      "loss": 0.2132,
+      "step": 7672
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1384490541523558,
+      "learning_rate": 1.3933278970051712e-05,
+      "loss": 0.2056,
+      "step": 7673
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2061262362828225,
+      "learning_rate": 1.3931764696620144e-05,
+      "loss": 0.2098,
+      "step": 7674
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6609050792349747,
+      "learning_rate": 1.3930250316539237e-05,
+      "loss": 0.1937,
+      "step": 7675
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0601825483540026,
+      "learning_rate": 1.3928735829850069e-05,
+      "loss": 0.2108,
+      "step": 7676
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9550895412085639,
+      "learning_rate": 1.3927221236593717e-05,
+      "loss": 0.1926,
+      "step": 7677
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8557196616439343,
+      "learning_rate": 1.392570653681127e-05,
+      "loss": 0.196,
+      "step": 7678
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9880124750398763,
+      "learning_rate": 1.392419173054381e-05,
+      "loss": 0.2059,
+      "step": 7679
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8231742504746878,
+      "learning_rate": 1.3922676817832432e-05,
+      "loss": 0.1963,
+      "step": 7680
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1466035230522185,
+      "learning_rate": 1.3921161798718217e-05,
+      "loss": 0.1918,
+      "step": 7681
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8208451969313353,
+      "learning_rate": 1.3919646673242272e-05,
+      "loss": 0.1905,
+      "step": 7682
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3997750841679093,
+      "learning_rate": 1.3918131441445687e-05,
+      "loss": 0.2143,
+      "step": 7683
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9636337485045307,
+      "learning_rate": 1.3916616103369567e-05,
+      "loss": 0.1944,
+      "step": 7684
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6144844863183354,
+      "learning_rate": 1.391510065905501e-05,
+      "loss": 0.1966,
+      "step": 7685
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9800808641664451,
+      "learning_rate": 1.3913585108543131e-05,
+      "loss": 0.1754,
+      "step": 7686
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8861425397209246,
+      "learning_rate": 1.3912069451875032e-05,
+      "loss": 0.192,
+      "step": 7687
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9087226051425511,
+      "learning_rate": 1.391055368909183e-05,
+      "loss": 0.1769,
+      "step": 7688
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.054864288688696,
+      "learning_rate": 1.3909037820234637e-05,
+      "loss": 0.1894,
+      "step": 7689
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1708274247339223,
+      "learning_rate": 1.3907521845344571e-05,
+      "loss": 0.1985,
+      "step": 7690
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0713949500783901,
+      "learning_rate": 1.3906005764462758e-05,
+      "loss": 0.2113,
+      "step": 7691
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5889139746273195,
+      "learning_rate": 1.3904489577630316e-05,
+      "loss": 0.1841,
+      "step": 7692
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.091357490656932,
+      "learning_rate": 1.3902973284888375e-05,
+      "loss": 0.1912,
+      "step": 7693
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8368483420261934,
+      "learning_rate": 1.3901456886278063e-05,
+      "loss": 0.204,
+      "step": 7694
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3583693284481406,
+      "learning_rate": 1.389994038184051e-05,
+      "loss": 0.203,
+      "step": 7695
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9312531164878359,
+      "learning_rate": 1.3898423771616854e-05,
+      "loss": 0.2161,
+      "step": 7696
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0903451646742162,
+      "learning_rate": 1.3896907055648235e-05,
+      "loss": 0.1956,
+      "step": 7697
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3439929901930354,
+      "learning_rate": 1.3895390233975793e-05,
+      "loss": 0.1906,
+      "step": 7698
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9825101816701041,
+      "learning_rate": 1.3893873306640669e-05,
+      "loss": 0.1943,
+      "step": 7699
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7285069594345552,
+      "learning_rate": 1.3892356273684012e-05,
+      "loss": 0.1971,
+      "step": 7700
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8837068614510146,
+      "learning_rate": 1.3890839135146972e-05,
+      "loss": 0.2041,
+      "step": 7701
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0181464717308626,
+      "learning_rate": 1.38893218910707e-05,
+      "loss": 0.1595,
+      "step": 7702
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1712417123030014,
+      "learning_rate": 1.388780454149635e-05,
+      "loss": 0.1995,
+      "step": 7703
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4386055659474735,
+      "learning_rate": 1.3886287086465085e-05,
+      "loss": 0.199,
+      "step": 7704
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9347759139666304,
+      "learning_rate": 1.3884769526018063e-05,
+      "loss": 0.1871,
+      "step": 7705
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.035172557245781,
+      "learning_rate": 1.3883251860196446e-05,
+      "loss": 0.2382,
+      "step": 7706
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0406406386957872,
+      "learning_rate": 1.3881734089041405e-05,
+      "loss": 0.1744,
+      "step": 7707
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0977754191926146,
+      "learning_rate": 1.388021621259411e-05,
+      "loss": 0.197,
+      "step": 7708
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0018304593302076,
+      "learning_rate": 1.3878698230895726e-05,
+      "loss": 0.1818,
+      "step": 7709
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9539595526956814,
+      "learning_rate": 1.3877180143987436e-05,
+      "loss": 0.1904,
+      "step": 7710
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3344670181819707,
+      "learning_rate": 1.3875661951910417e-05,
+      "loss": 0.1971,
+      "step": 7711
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.960361113128031,
+      "learning_rate": 1.3874143654705845e-05,
+      "loss": 0.1771,
+      "step": 7712
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2090564270325268,
+      "learning_rate": 1.387262525241491e-05,
+      "loss": 0.2206,
+      "step": 7713
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.986088483516514,
+      "learning_rate": 1.3871106745078798e-05,
+      "loss": 0.1889,
+      "step": 7714
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.434936865372642,
+      "learning_rate": 1.3869588132738696e-05,
+      "loss": 0.2354,
+      "step": 7715
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0088629743524504,
+      "learning_rate": 1.3868069415435795e-05,
+      "loss": 0.184,
+      "step": 7716
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6641815842457266,
+      "learning_rate": 1.3866550593211292e-05,
+      "loss": 0.2017,
+      "step": 7717
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3170396920934737,
+      "learning_rate": 1.386503166610639e-05,
+      "loss": 0.1842,
+      "step": 7718
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7855746745131865,
+      "learning_rate": 1.3863512634162283e-05,
+      "loss": 0.1922,
+      "step": 7719
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9041492336676009,
+      "learning_rate": 1.3861993497420176e-05,
+      "loss": 0.1913,
+      "step": 7720
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1314961895930964,
+      "learning_rate": 1.3860474255921279e-05,
+      "loss": 0.2246,
+      "step": 7721
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5367511841726282,
+      "learning_rate": 1.38589549097068e-05,
+      "loss": 0.1881,
+      "step": 7722
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9589789968589525,
+      "learning_rate": 1.3857435458817952e-05,
+      "loss": 0.2037,
+      "step": 7723
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1288332166730548,
+      "learning_rate": 1.3855915903295949e-05,
+      "loss": 0.1902,
+      "step": 7724
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.960376489981517,
+      "learning_rate": 1.3854396243182007e-05,
+      "loss": 0.2171,
+      "step": 7725
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9273880760394527,
+      "learning_rate": 1.3852876478517352e-05,
+      "loss": 0.2086,
+      "step": 7726
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2440017744030931,
+      "learning_rate": 1.3851356609343202e-05,
+      "loss": 0.1798,
+      "step": 7727
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8535773225164793,
+      "learning_rate": 1.3849836635700791e-05,
+      "loss": 0.1772,
+      "step": 7728
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.963785047508875,
+      "learning_rate": 1.384831655763134e-05,
+      "loss": 0.2068,
+      "step": 7729
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.214837663043247,
+      "learning_rate": 1.3846796375176083e-05,
+      "loss": 0.1957,
+      "step": 7730
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1073446266276163,
+      "learning_rate": 1.3845276088376262e-05,
+      "loss": 0.2112,
+      "step": 7731
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0268993364703471,
+      "learning_rate": 1.3843755697273109e-05,
+      "loss": 0.1911,
+      "step": 7732
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9552792651480384,
+      "learning_rate": 1.3842235201907865e-05,
+      "loss": 0.2289,
+      "step": 7733
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0202539534015398,
+      "learning_rate": 1.3840714602321774e-05,
+      "loss": 0.1571,
+      "step": 7734
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8635891135148347,
+      "learning_rate": 1.3839193898556083e-05,
+      "loss": 0.1876,
+      "step": 7735
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1217723335679453,
+      "learning_rate": 1.383767309065204e-05,
+      "loss": 0.1988,
+      "step": 7736
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0698837728586303,
+      "learning_rate": 1.3836152178650899e-05,
+      "loss": 0.1921,
+      "step": 7737
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0058637146346,
+      "learning_rate": 1.383463116259391e-05,
+      "loss": 0.2095,
+      "step": 7738
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9999314188011952,
+      "learning_rate": 1.3833110042522337e-05,
+      "loss": 0.2046,
+      "step": 7739
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4672213955523357,
+      "learning_rate": 1.3831588818477437e-05,
+      "loss": 0.2053,
+      "step": 7740
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0337800417999594,
+      "learning_rate": 1.3830067490500474e-05,
+      "loss": 0.2175,
+      "step": 7741
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2319011038386658,
+      "learning_rate": 1.3828546058632714e-05,
+      "loss": 0.1902,
+      "step": 7742
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3044079880196222,
+      "learning_rate": 1.3827024522915425e-05,
+      "loss": 0.2291,
+      "step": 7743
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8538781044258716,
+      "learning_rate": 1.382550288338988e-05,
+      "loss": 0.2123,
+      "step": 7744
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.879168221739137,
+      "learning_rate": 1.3823981140097354e-05,
+      "loss": 0.179,
+      "step": 7745
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9413694338433246,
+      "learning_rate": 1.3822459293079122e-05,
+      "loss": 0.1967,
+      "step": 7746
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1056894724859516,
+      "learning_rate": 1.3820937342376467e-05,
+      "loss": 0.191,
+      "step": 7747
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9126483606695378,
+      "learning_rate": 1.3819415288030672e-05,
+      "loss": 0.1762,
+      "step": 7748
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1355339214771147,
+      "learning_rate": 1.381789313008302e-05,
+      "loss": 0.2235,
+      "step": 7749
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9826179082927652,
+      "learning_rate": 1.3816370868574804e-05,
+      "loss": 0.2131,
+      "step": 7750
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8860565691463593,
+      "learning_rate": 1.3814848503547308e-05,
+      "loss": 0.1782,
+      "step": 7751
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5081861453338383,
+      "learning_rate": 1.3813326035041832e-05,
+      "loss": 0.2113,
+      "step": 7752
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3892948185606584,
+      "learning_rate": 1.3811803463099675e-05,
+      "loss": 0.2063,
+      "step": 7753
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9911639651533942,
+      "learning_rate": 1.3810280787762131e-05,
+      "loss": 0.1817,
+      "step": 7754
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.022925229717426,
+      "learning_rate": 1.3808758009070506e-05,
+      "loss": 0.1943,
+      "step": 7755
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3548483263654096,
+      "learning_rate": 1.3807235127066104e-05,
+      "loss": 0.2179,
+      "step": 7756
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1166557524441694,
+      "learning_rate": 1.3805712141790237e-05,
+      "loss": 0.2168,
+      "step": 7757
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9892567026131145,
+      "learning_rate": 1.380418905328421e-05,
+      "loss": 0.177,
+      "step": 7758
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0970740628385713,
+      "learning_rate": 1.3802665861589342e-05,
+      "loss": 0.229,
+      "step": 7759
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8689804869980224,
+      "learning_rate": 1.3801142566746945e-05,
+      "loss": 0.167,
+      "step": 7760
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.081339687802085,
+      "learning_rate": 1.3799619168798346e-05,
+      "loss": 0.1883,
+      "step": 7761
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7086114323382755,
+      "learning_rate": 1.3798095667784859e-05,
+      "loss": 0.2153,
+      "step": 7762
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2132432860413234,
+      "learning_rate": 1.3796572063747813e-05,
+      "loss": 0.2145,
+      "step": 7763
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9325342418973277,
+      "learning_rate": 1.3795048356728538e-05,
+      "loss": 0.173,
+      "step": 7764
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1566731200513392,
+      "learning_rate": 1.3793524546768358e-05,
+      "loss": 0.1983,
+      "step": 7765
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9601961276441815,
+      "learning_rate": 1.3792000633908612e-05,
+      "loss": 0.2059,
+      "step": 7766
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8099371655872485,
+      "learning_rate": 1.3790476618190634e-05,
+      "loss": 0.2094,
+      "step": 7767
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7473037502181532,
+      "learning_rate": 1.3788952499655765e-05,
+      "loss": 0.1907,
+      "step": 7768
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9096819251192658,
+      "learning_rate": 1.3787428278345344e-05,
+      "loss": 0.1857,
+      "step": 7769
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9874258926455962,
+      "learning_rate": 1.378590395430072e-05,
+      "loss": 0.1808,
+      "step": 7770
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9443362322677041,
+      "learning_rate": 1.3784379527563233e-05,
+      "loss": 0.1781,
+      "step": 7771
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1366189588352082,
+      "learning_rate": 1.3782854998174243e-05,
+      "loss": 0.2119,
+      "step": 7772
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9441430740776766,
+      "learning_rate": 1.3781330366175093e-05,
+      "loss": 0.1915,
+      "step": 7773
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1722347958655013,
+      "learning_rate": 1.3779805631607144e-05,
+      "loss": 0.2206,
+      "step": 7774
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.011282633832701,
+      "learning_rate": 1.3778280794511753e-05,
+      "loss": 0.1823,
+      "step": 7775
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1645863169044224,
+      "learning_rate": 1.3776755854930285e-05,
+      "loss": 0.2104,
+      "step": 7776
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1038162763376442,
+      "learning_rate": 1.3775230812904101e-05,
+      "loss": 0.2008,
+      "step": 7777
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5586894325978662,
+      "learning_rate": 1.3773705668474564e-05,
+      "loss": 0.2054,
+      "step": 7778
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0103587347743808,
+      "learning_rate": 1.377218042168305e-05,
+      "loss": 0.2099,
+      "step": 7779
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0698592267223566,
+      "learning_rate": 1.3770655072570929e-05,
+      "loss": 0.2056,
+      "step": 7780
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1252856387119448,
+      "learning_rate": 1.3769129621179578e-05,
+      "loss": 0.1959,
+      "step": 7781
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3938755135494856,
+      "learning_rate": 1.3767604067550369e-05,
+      "loss": 0.2072,
+      "step": 7782
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.043094931379382,
+      "learning_rate": 1.376607841172469e-05,
+      "loss": 0.1888,
+      "step": 7783
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9900803380343406,
+      "learning_rate": 1.376455265374392e-05,
+      "loss": 0.199,
+      "step": 7784
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3610036980785805,
+      "learning_rate": 1.376302679364945e-05,
+      "loss": 0.17,
+      "step": 7785
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8993769498920858,
+      "learning_rate": 1.3761500831482661e-05,
+      "loss": 0.1989,
+      "step": 7786
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9511935062881577,
+      "learning_rate": 1.3759974767284954e-05,
+      "loss": 0.1929,
+      "step": 7787
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.04482184811609,
+      "learning_rate": 1.3758448601097715e-05,
+      "loss": 0.1905,
+      "step": 7788
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0151552795044474,
+      "learning_rate": 1.3756922332962349e-05,
+      "loss": 0.1841,
+      "step": 7789
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1638260461131706,
+      "learning_rate": 1.375539596292025e-05,
+      "loss": 0.203,
+      "step": 7790
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1345663298986353,
+      "learning_rate": 1.3753869491012822e-05,
+      "loss": 0.1927,
+      "step": 7791
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.274166847278216,
+      "learning_rate": 1.3752342917281474e-05,
+      "loss": 0.1876,
+      "step": 7792
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2116466657736265,
+      "learning_rate": 1.3750816241767612e-05,
+      "loss": 0.2162,
+      "step": 7793
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.292280681415955,
+      "learning_rate": 1.374928946451265e-05,
+      "loss": 0.2137,
+      "step": 7794
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9538512883912856,
+      "learning_rate": 1.3747762585557995e-05,
+      "loss": 0.1932,
+      "step": 7795
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.902903017738043,
+      "learning_rate": 1.374623560494507e-05,
+      "loss": 0.2053,
+      "step": 7796
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9089488041497878,
+      "learning_rate": 1.3744708522715295e-05,
+      "loss": 0.1868,
+      "step": 7797
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.036402555926078,
+      "learning_rate": 1.3743181338910088e-05,
+      "loss": 0.175,
+      "step": 7798
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1412140081353452,
+      "learning_rate": 1.3741654053570877e-05,
+      "loss": 0.2069,
+      "step": 7799
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9812081753904945,
+      "learning_rate": 1.3740126666739086e-05,
+      "loss": 0.1921,
+      "step": 7800
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8599736416295825,
+      "learning_rate": 1.3738599178456149e-05,
+      "loss": 0.1993,
+      "step": 7801
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3933742441258217,
+      "learning_rate": 1.37370715887635e-05,
+      "loss": 0.2095,
+      "step": 7802
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9186112337500361,
+      "learning_rate": 1.3735543897702572e-05,
+      "loss": 0.2101,
+      "step": 7803
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7780230586723094,
+      "learning_rate": 1.3734016105314803e-05,
+      "loss": 0.1716,
+      "step": 7804
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.912468228908501,
+      "learning_rate": 1.3732488211641638e-05,
+      "loss": 0.1777,
+      "step": 7805
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0843294009632642,
+      "learning_rate": 1.3730960216724518e-05,
+      "loss": 0.1716,
+      "step": 7806
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3501717023837643,
+      "learning_rate": 1.3729432120604895e-05,
+      "loss": 0.1975,
+      "step": 7807
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.033915342936664,
+      "learning_rate": 1.3727903923324211e-05,
+      "loss": 0.1954,
+      "step": 7808
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9776691062281382,
+      "learning_rate": 1.3726375624923925e-05,
+      "loss": 0.1895,
+      "step": 7809
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8463396302747215,
+      "learning_rate": 1.3724847225445488e-05,
+      "loss": 0.1929,
+      "step": 7810
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1511623956776489,
+      "learning_rate": 1.3723318724930362e-05,
+      "loss": 0.168,
+      "step": 7811
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6733247641066649,
+      "learning_rate": 1.3721790123420002e-05,
+      "loss": 0.2306,
+      "step": 7812
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0738043696689645,
+      "learning_rate": 1.3720261420955874e-05,
+      "loss": 0.1865,
+      "step": 7813
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9350696789972645,
+      "learning_rate": 1.3718732617579449e-05,
+      "loss": 0.1919,
+      "step": 7814
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0062826557369187,
+      "learning_rate": 1.371720371333219e-05,
+      "loss": 0.178,
+      "step": 7815
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.095925858088011,
+      "learning_rate": 1.3715674708255571e-05,
+      "loss": 0.1966,
+      "step": 7816
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.072683440548299,
+      "learning_rate": 1.3714145602391063e-05,
+      "loss": 0.1699,
+      "step": 7817
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0745173488233095,
+      "learning_rate": 1.3712616395780148e-05,
+      "loss": 0.1882,
+      "step": 7818
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9455046613837105,
+      "learning_rate": 1.3711087088464303e-05,
+      "loss": 0.1703,
+      "step": 7819
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.875943906638408,
+      "learning_rate": 1.3709557680485013e-05,
+      "loss": 0.21,
+      "step": 7820
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.555443947600797,
+      "learning_rate": 1.3708028171883757e-05,
+      "loss": 0.2067,
+      "step": 7821
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.019118936893599,
+      "learning_rate": 1.3706498562702032e-05,
+      "loss": 0.1905,
+      "step": 7822
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9698314019007279,
+      "learning_rate": 1.3704968852981322e-05,
+      "loss": 0.2184,
+      "step": 7823
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9142833794886264,
+      "learning_rate": 1.3703439042763122e-05,
+      "loss": 0.2014,
+      "step": 7824
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9029009788321303,
+      "learning_rate": 1.3701909132088931e-05,
+      "loss": 0.2014,
+      "step": 7825
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.252169628887097,
+      "learning_rate": 1.3700379121000245e-05,
+      "loss": 0.2111,
+      "step": 7826
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3070075999421975,
+      "learning_rate": 1.3698849009538564e-05,
+      "loss": 0.2055,
+      "step": 7827
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3437122932148957,
+      "learning_rate": 1.3697318797745399e-05,
+      "loss": 0.1853,
+      "step": 7828
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0379265187071747,
+      "learning_rate": 1.3695788485662248e-05,
+      "loss": 0.1936,
+      "step": 7829
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1959800320200566,
+      "learning_rate": 1.3694258073330626e-05,
+      "loss": 0.1847,
+      "step": 7830
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0100684236799955,
+      "learning_rate": 1.3692727560792048e-05,
+      "loss": 0.2191,
+      "step": 7831
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0045741496998244,
+      "learning_rate": 1.3691196948088026e-05,
+      "loss": 0.2086,
+      "step": 7832
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.880542853536741,
+      "learning_rate": 1.3689666235260078e-05,
+      "loss": 0.2093,
+      "step": 7833
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8663881683353095,
+      "learning_rate": 1.3688135422349724e-05,
+      "loss": 0.1842,
+      "step": 7834
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1894788719346479,
+      "learning_rate": 1.3686604509398489e-05,
+      "loss": 0.215,
+      "step": 7835
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.239200767986605,
+      "learning_rate": 1.3685073496447898e-05,
+      "loss": 0.2123,
+      "step": 7836
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0676797817770391,
+      "learning_rate": 1.3683542383539482e-05,
+      "loss": 0.1925,
+      "step": 7837
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.179582556696482,
+      "learning_rate": 1.3682011170714771e-05,
+      "loss": 0.1993,
+      "step": 7838
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9022362728092863,
+      "learning_rate": 1.3680479858015297e-05,
+      "loss": 0.184,
+      "step": 7839
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8072628959257323,
+      "learning_rate": 1.3678948445482598e-05,
+      "loss": 0.1991,
+      "step": 7840
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1446276432557878,
+      "learning_rate": 1.3677416933158216e-05,
+      "loss": 0.1997,
+      "step": 7841
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3285163458160731,
+      "learning_rate": 1.3675885321083693e-05,
+      "loss": 0.1934,
+      "step": 7842
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0662290021048162,
+      "learning_rate": 1.3674353609300571e-05,
+      "loss": 0.1701,
+      "step": 7843
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8600687517356913,
+      "learning_rate": 1.3672821797850402e-05,
+      "loss": 0.1782,
+      "step": 7844
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9653570695824466,
+      "learning_rate": 1.3671289886774733e-05,
+      "loss": 0.2028,
+      "step": 7845
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9470181500627005,
+      "learning_rate": 1.3669757876115117e-05,
+      "loss": 0.2137,
+      "step": 7846
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3353820715794826,
+      "learning_rate": 1.3668225765913114e-05,
+      "loss": 0.1769,
+      "step": 7847
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0425801707777265,
+      "learning_rate": 1.3666693556210278e-05,
+      "loss": 0.1965,
+      "step": 7848
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9905064591861154,
+      "learning_rate": 1.3665161247048173e-05,
+      "loss": 0.1833,
+      "step": 7849
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2775710642059581,
+      "learning_rate": 1.3663628838468364e-05,
+      "loss": 0.1912,
+      "step": 7850
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2917881575654395,
+      "learning_rate": 1.3662096330512413e-05,
+      "loss": 0.2011,
+      "step": 7851
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0359447949370741,
+      "learning_rate": 1.3660563723221894e-05,
+      "loss": 0.2311,
+      "step": 7852
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8779461362875117,
+      "learning_rate": 1.3659031016638376e-05,
+      "loss": 0.2028,
+      "step": 7853
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0540674291123995,
+      "learning_rate": 1.3657498210803435e-05,
+      "loss": 0.1967,
+      "step": 7854
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4087280910830706,
+      "learning_rate": 1.3655965305758652e-05,
+      "loss": 0.1976,
+      "step": 7855
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8105676637666767,
+      "learning_rate": 1.36544323015456e-05,
+      "loss": 0.214,
+      "step": 7856
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9090277786069536,
+      "learning_rate": 1.3652899198205864e-05,
+      "loss": 0.1919,
+      "step": 7857
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0320170900219516,
+      "learning_rate": 1.3651365995781034e-05,
+      "loss": 0.1886,
+      "step": 7858
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8236703494092736,
+      "learning_rate": 1.3649832694312695e-05,
+      "loss": 0.2061,
+      "step": 7859
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.380715986676055,
+      "learning_rate": 1.3648299293842438e-05,
+      "loss": 0.1852,
+      "step": 7860
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7081578575084013,
+      "learning_rate": 1.3646765794411854e-05,
+      "loss": 0.172,
+      "step": 7861
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8488985864335844,
+      "learning_rate": 1.3645232196062544e-05,
+      "loss": 0.2182,
+      "step": 7862
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.006743983617717,
+      "learning_rate": 1.3643698498836104e-05,
+      "loss": 0.1942,
+      "step": 7863
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.14256700029364,
+      "learning_rate": 1.3642164702774137e-05,
+      "loss": 0.196,
+      "step": 7864
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1811893526465074,
+      "learning_rate": 1.3640630807918246e-05,
+      "loss": 0.1893,
+      "step": 7865
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8455848742795187,
+      "learning_rate": 1.3639096814310037e-05,
+      "loss": 0.2269,
+      "step": 7866
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7749192958494059,
+      "learning_rate": 1.3637562721991127e-05,
+      "loss": 0.1869,
+      "step": 7867
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8349972385642351,
+      "learning_rate": 1.3636028531003118e-05,
+      "loss": 0.2016,
+      "step": 7868
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9953791132956865,
+      "learning_rate": 1.3634494241387632e-05,
+      "loss": 0.2042,
+      "step": 7869
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9222005288634191,
+      "learning_rate": 1.363295985318628e-05,
+      "loss": 0.2072,
+      "step": 7870
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.075824838479233,
+      "learning_rate": 1.3631425366440691e-05,
+      "loss": 0.1992,
+      "step": 7871
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5194579873674021,
+      "learning_rate": 1.3629890781192486e-05,
+      "loss": 0.1939,
+      "step": 7872
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7643820525970437,
+      "learning_rate": 1.3628356097483288e-05,
+      "loss": 0.1971,
+      "step": 7873
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.197943921189177,
+      "learning_rate": 1.362682131535472e-05,
+      "loss": 0.1876,
+      "step": 7874
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2330583994530917,
+      "learning_rate": 1.3625286434848424e-05,
+      "loss": 0.1771,
+      "step": 7875
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1913529068812612,
+      "learning_rate": 1.3623751456006027e-05,
+      "loss": 0.201,
+      "step": 7876
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1474609620458969,
+      "learning_rate": 1.3622216378869169e-05,
+      "loss": 0.1877,
+      "step": 7877
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5636328342053893,
+      "learning_rate": 1.3620681203479484e-05,
+      "loss": 0.1911,
+      "step": 7878
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1944564472343593,
+      "learning_rate": 1.3619145929878617e-05,
+      "loss": 0.1999,
+      "step": 7879
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9179919973426882,
+      "learning_rate": 1.3617610558108214e-05,
+      "loss": 0.2123,
+      "step": 7880
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.060644212888549,
+      "learning_rate": 1.3616075088209921e-05,
+      "loss": 0.2137,
+      "step": 7881
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.857124640067879,
+      "learning_rate": 1.3614539520225388e-05,
+      "loss": 0.1664,
+      "step": 7882
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6058142618696125,
+      "learning_rate": 1.361300385419626e-05,
+      "loss": 0.1857,
+      "step": 7883
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.335899099030283,
+      "learning_rate": 1.3611468090164203e-05,
+      "loss": 0.2105,
+      "step": 7884
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0110629983418822,
+      "learning_rate": 1.3609932228170873e-05,
+      "loss": 0.189,
+      "step": 7885
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9841281444487436,
+      "learning_rate": 1.3608396268257922e-05,
+      "loss": 0.1885,
+      "step": 7886
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9728002662716808,
+      "learning_rate": 1.360686021046702e-05,
+      "loss": 0.2076,
+      "step": 7887
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1365569391346437,
+      "learning_rate": 1.360532405483983e-05,
+      "loss": 0.2065,
+      "step": 7888
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5428500470719355,
+      "learning_rate": 1.3603787801418025e-05,
+      "loss": 0.1843,
+      "step": 7889
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9149812568191804,
+      "learning_rate": 1.3602251450243273e-05,
+      "loss": 0.2111,
+      "step": 7890
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4272311709229732,
+      "learning_rate": 1.3600715001357241e-05,
+      "loss": 0.2076,
+      "step": 7891
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1625936777606585,
+      "learning_rate": 1.3599178454801615e-05,
+      "loss": 0.2171,
+      "step": 7892
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8890855284761452,
+      "learning_rate": 1.3597641810618071e-05,
+      "loss": 0.2132,
+      "step": 7893
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9647647431889569,
+      "learning_rate": 1.3596105068848289e-05,
+      "loss": 0.2097,
+      "step": 7894
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9228742641602002,
+      "learning_rate": 1.3594568229533953e-05,
+      "loss": 0.1896,
+      "step": 7895
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9523702934499297,
+      "learning_rate": 1.359303129271675e-05,
+      "loss": 0.2188,
+      "step": 7896
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2907224300111102,
+      "learning_rate": 1.3591494258438372e-05,
+      "loss": 0.2196,
+      "step": 7897
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4941226422465974,
+      "learning_rate": 1.3589957126740508e-05,
+      "loss": 0.1844,
+      "step": 7898
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.39262869659887,
+      "learning_rate": 1.3588419897664855e-05,
+      "loss": 0.2123,
+      "step": 7899
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3557954440930702,
+      "learning_rate": 1.358688257125311e-05,
+      "loss": 0.2364,
+      "step": 7900
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8274355379770744,
+      "learning_rate": 1.3585345147546971e-05,
+      "loss": 0.2012,
+      "step": 7901
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.445332020814645,
+      "learning_rate": 1.3583807626588143e-05,
+      "loss": 0.2053,
+      "step": 7902
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0809255064379315,
+      "learning_rate": 1.3582270008418332e-05,
+      "loss": 0.1686,
+      "step": 7903
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1250535830016777,
+      "learning_rate": 1.3580732293079244e-05,
+      "loss": 0.1923,
+      "step": 7904
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9187987560101761,
+      "learning_rate": 1.357919448061259e-05,
+      "loss": 0.1737,
+      "step": 7905
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9939882858974505,
+      "learning_rate": 1.3577656571060082e-05,
+      "loss": 0.1889,
+      "step": 7906
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1124667983403917,
+      "learning_rate": 1.357611856446344e-05,
+      "loss": 0.204,
+      "step": 7907
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2556400683929114,
+      "learning_rate": 1.3574580460864381e-05,
+      "loss": 0.1767,
+      "step": 7908
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.139433671837242,
+      "learning_rate": 1.3573042260304623e-05,
+      "loss": 0.2201,
+      "step": 7909
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7612526668488224,
+      "learning_rate": 1.3571503962825892e-05,
+      "loss": 0.1781,
+      "step": 7910
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9627113385215519,
+      "learning_rate": 1.3569965568469915e-05,
+      "loss": 0.1901,
+      "step": 7911
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8696829378181286,
+      "learning_rate": 1.3568427077278422e-05,
+      "loss": 0.1642,
+      "step": 7912
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8150121776016657,
+      "learning_rate": 1.356688848929314e-05,
+      "loss": 0.1953,
+      "step": 7913
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.192895214033769,
+      "learning_rate": 1.3565349804555805e-05,
+      "loss": 0.1792,
+      "step": 7914
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9312534145645553,
+      "learning_rate": 1.3563811023108157e-05,
+      "loss": 0.1973,
+      "step": 7915
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.102289423897505,
+      "learning_rate": 1.3562272144991934e-05,
+      "loss": 0.187,
+      "step": 7916
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5551690732592423,
+      "learning_rate": 1.3560733170248878e-05,
+      "loss": 0.2014,
+      "step": 7917
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9967216384377598,
+      "learning_rate": 1.3559194098920732e-05,
+      "loss": 0.1793,
+      "step": 7918
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.98329460157344,
+      "learning_rate": 1.3557654931049247e-05,
+      "loss": 0.1944,
+      "step": 7919
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.558620907919238,
+      "learning_rate": 1.355611566667617e-05,
+      "loss": 0.1906,
+      "step": 7920
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1066331357961985,
+      "learning_rate": 1.3554576305843257e-05,
+      "loss": 0.2229,
+      "step": 7921
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9693267653975826,
+      "learning_rate": 1.3553036848592261e-05,
+      "loss": 0.2108,
+      "step": 7922
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8412025440545284,
+      "learning_rate": 1.3551497294964935e-05,
+      "loss": 0.2106,
+      "step": 7923
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1110383384259537,
+      "learning_rate": 1.3549957645003046e-05,
+      "loss": 0.1907,
+      "step": 7924
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8034700067923729,
+      "learning_rate": 1.3548417898748361e-05,
+      "loss": 0.2206,
+      "step": 7925
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7850263892895291,
+      "learning_rate": 1.3546878056242632e-05,
+      "loss": 0.1809,
+      "step": 7926
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1270612963348563,
+      "learning_rate": 1.354533811752764e-05,
+      "loss": 0.1852,
+      "step": 7927
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9435404761329577,
+      "learning_rate": 1.3543798082645152e-05,
+      "loss": 0.2031,
+      "step": 7928
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8921086902767899,
+      "learning_rate": 1.3542257951636939e-05,
+      "loss": 0.2034,
+      "step": 7929
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6722635092982678,
+      "learning_rate": 1.354071772454478e-05,
+      "loss": 0.2064,
+      "step": 7930
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.201211190418289,
+      "learning_rate": 1.353917740141045e-05,
+      "loss": 0.2042,
+      "step": 7931
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9930918100552604,
+      "learning_rate": 1.3537636982275734e-05,
+      "loss": 0.1872,
+      "step": 7932
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9903922908206649,
+      "learning_rate": 1.3536096467182418e-05,
+      "loss": 0.1829,
+      "step": 7933
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8269314138269611,
+      "learning_rate": 1.3534555856172285e-05,
+      "loss": 0.2094,
+      "step": 7934
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.193805357692574,
+      "learning_rate": 1.3533015149287123e-05,
+      "loss": 0.1892,
+      "step": 7935
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.504864313457114,
+      "learning_rate": 1.3531474346568724e-05,
+      "loss": 0.1833,
+      "step": 7936
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8930310541385884,
+      "learning_rate": 1.3529933448058885e-05,
+      "loss": 0.1991,
+      "step": 7937
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.112814505702536,
+      "learning_rate": 1.3528392453799403e-05,
+      "loss": 0.2099,
+      "step": 7938
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1446346643591627,
+      "learning_rate": 1.352685136383208e-05,
+      "loss": 0.2001,
+      "step": 7939
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8912392415537557,
+      "learning_rate": 1.3525310178198707e-05,
+      "loss": 0.1905,
+      "step": 7940
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1065270220767909,
+      "learning_rate": 1.3523768896941101e-05,
+      "loss": 0.2287,
+      "step": 7941
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0090741292296308,
+      "learning_rate": 1.3522227520101064e-05,
+      "loss": 0.1835,
+      "step": 7942
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9676105550984178,
+      "learning_rate": 1.3520686047720409e-05,
+      "loss": 0.1913,
+      "step": 7943
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.073028571367729,
+      "learning_rate": 1.3519144479840942e-05,
+      "loss": 0.2014,
+      "step": 7944
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.796391365846085,
+      "learning_rate": 1.3517602816504482e-05,
+      "loss": 0.193,
+      "step": 7945
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0138190132905434,
+      "learning_rate": 1.351606105775285e-05,
+      "loss": 0.1995,
+      "step": 7946
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0018287764775855,
+      "learning_rate": 1.3514519203627863e-05,
+      "loss": 0.1938,
+      "step": 7947
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8105897346963113,
+      "learning_rate": 1.3512977254171343e-05,
+      "loss": 0.1963,
+      "step": 7948
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8502494921740894,
+      "learning_rate": 1.3511435209425115e-05,
+      "loss": 0.2106,
+      "step": 7949
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1600577932155047,
+      "learning_rate": 1.3509893069431011e-05,
+      "loss": 0.1868,
+      "step": 7950
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1029932924061727,
+      "learning_rate": 1.3508350834230857e-05,
+      "loss": 0.2033,
+      "step": 7951
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8956362559777115,
+      "learning_rate": 1.3506808503866491e-05,
+      "loss": 0.1889,
+      "step": 7952
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.144965652654032,
+      "learning_rate": 1.3505266078379741e-05,
+      "loss": 0.1805,
+      "step": 7953
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9835022229395287,
+      "learning_rate": 1.3503723557812455e-05,
+      "loss": 0.1973,
+      "step": 7954
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7857287589948215,
+      "learning_rate": 1.3502180942206472e-05,
+      "loss": 0.1822,
+      "step": 7955
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.045687291906436,
+      "learning_rate": 1.350063823160363e-05,
+      "loss": 0.1969,
+      "step": 7956
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8156672686588005,
+      "learning_rate": 1.3499095426045779e-05,
+      "loss": 0.1872,
+      "step": 7957
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2209587892466776,
+      "learning_rate": 1.3497552525574763e-05,
+      "loss": 0.2176,
+      "step": 7958
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8184576292391839,
+      "learning_rate": 1.3496009530232444e-05,
+      "loss": 0.2073,
+      "step": 7959
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7409523366673464,
+      "learning_rate": 1.3494466440060667e-05,
+      "loss": 0.1898,
+      "step": 7960
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9175046443611684,
+      "learning_rate": 1.349292325510129e-05,
+      "loss": 0.1953,
+      "step": 7961
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.991920168357191,
+      "learning_rate": 1.3491379975396171e-05,
+      "loss": 0.1846,
+      "step": 7962
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9289700337194272,
+      "learning_rate": 1.3489836600987173e-05,
+      "loss": 0.1884,
+      "step": 7963
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9731661635607622,
+      "learning_rate": 1.3488293131916161e-05,
+      "loss": 0.1642,
+      "step": 7964
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8785920997118186,
+      "learning_rate": 1.3486749568225002e-05,
+      "loss": 0.1948,
+      "step": 7965
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9600923761808118,
+      "learning_rate": 1.3485205909955562e-05,
+      "loss": 0.1848,
+      "step": 7966
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9832943610219766,
+      "learning_rate": 1.3483662157149713e-05,
+      "loss": 0.181,
+      "step": 7967
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.164552711160505,
+      "learning_rate": 1.3482118309849335e-05,
+      "loss": 0.2095,
+      "step": 7968
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8209572252079305,
+      "learning_rate": 1.34805743680963e-05,
+      "loss": 0.1783,
+      "step": 7969
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.279067529820421,
+      "learning_rate": 1.3479030331932488e-05,
+      "loss": 0.18,
+      "step": 7970
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.70461197632034,
+      "learning_rate": 1.347748620139978e-05,
+      "loss": 0.1622,
+      "step": 7971
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9765087054600745,
+      "learning_rate": 1.3475941976540066e-05,
+      "loss": 0.2011,
+      "step": 7972
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6868005859222541,
+      "learning_rate": 1.3474397657395231e-05,
+      "loss": 0.1957,
+      "step": 7973
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8892658074672084,
+      "learning_rate": 1.347285324400716e-05,
+      "loss": 0.2046,
+      "step": 7974
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2671014445922417,
+      "learning_rate": 1.347130873641775e-05,
+      "loss": 0.2059,
+      "step": 7975
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8699850071913555,
+      "learning_rate": 1.346976413466889e-05,
+      "loss": 0.1762,
+      "step": 7976
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8392292585051886,
+      "learning_rate": 1.3468219438802487e-05,
+      "loss": 0.1654,
+      "step": 7977
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5843249219827986,
+      "learning_rate": 1.3466674648860436e-05,
+      "loss": 0.2023,
+      "step": 7978
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8306870529471548,
+      "learning_rate": 1.3465129764884636e-05,
+      "loss": 0.1781,
+      "step": 7979
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9424208434182422,
+      "learning_rate": 1.3463584786916997e-05,
+      "loss": 0.2124,
+      "step": 7980
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0553131794893045,
+      "learning_rate": 1.3462039714999426e-05,
+      "loss": 0.2425,
+      "step": 7981
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8565393972030418,
+      "learning_rate": 1.3460494549173833e-05,
+      "loss": 0.1893,
+      "step": 7982
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3153057889606654,
+      "learning_rate": 1.3458949289482126e-05,
+      "loss": 0.174,
+      "step": 7983
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.972633249567405,
+      "learning_rate": 1.3457403935966227e-05,
+      "loss": 0.1992,
+      "step": 7984
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1382920500187748,
+      "learning_rate": 1.345585848866805e-05,
+      "loss": 0.2359,
+      "step": 7985
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8575969850834319,
+      "learning_rate": 1.3454312947629515e-05,
+      "loss": 0.1995,
+      "step": 7986
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0979026964467675,
+      "learning_rate": 1.345276731289255e-05,
+      "loss": 0.2002,
+      "step": 7987
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.079195049925033,
+      "learning_rate": 1.3451221584499073e-05,
+      "loss": 0.2269,
+      "step": 7988
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9035088056320559,
+      "learning_rate": 1.3449675762491017e-05,
+      "loss": 0.1953,
+      "step": 7989
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8710857518370761,
+      "learning_rate": 1.3448129846910312e-05,
+      "loss": 0.2033,
+      "step": 7990
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7995548269495811,
+      "learning_rate": 1.344658383779889e-05,
+      "loss": 0.2021,
+      "step": 7991
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.317150873310952,
+      "learning_rate": 1.3445037735198684e-05,
+      "loss": 0.2065,
+      "step": 7992
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8921681775584891,
+      "learning_rate": 1.3443491539151636e-05,
+      "loss": 0.2054,
+      "step": 7993
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8545579901327621,
+      "learning_rate": 1.3441945249699687e-05,
+      "loss": 0.2003,
+      "step": 7994
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0063728134691161,
+      "learning_rate": 1.3440398866884781e-05,
+      "loss": 0.2002,
+      "step": 7995
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9023502076409247,
+      "learning_rate": 1.343885239074886e-05,
+      "loss": 0.2044,
+      "step": 7996
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7519701313854469,
+      "learning_rate": 1.343730582133387e-05,
+      "loss": 0.1963,
+      "step": 7997
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9367432902659597,
+      "learning_rate": 1.3435759158681767e-05,
+      "loss": 0.1965,
+      "step": 7998
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9849913851261063,
+      "learning_rate": 1.3434212402834503e-05,
+      "loss": 0.1797,
+      "step": 7999
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9043783031126352,
+      "learning_rate": 1.3432665553834036e-05,
+      "loss": 0.1825,
+      "step": 8000
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9773402570390824,
+      "learning_rate": 1.3431118611722317e-05,
+      "loss": 0.1956,
+      "step": 8001
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9198481783329289,
+      "learning_rate": 1.3429571576541315e-05,
+      "loss": 0.1996,
+      "step": 8002
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9380731530654524,
+      "learning_rate": 1.3428024448332992e-05,
+      "loss": 0.1938,
+      "step": 8003
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1878699897617486,
+      "learning_rate": 1.342647722713931e-05,
+      "loss": 0.1915,
+      "step": 8004
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7892945827995911,
+      "learning_rate": 1.342492991300224e-05,
+      "loss": 0.1886,
+      "step": 8005
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8627428951792862,
+      "learning_rate": 1.3423382505963752e-05,
+      "loss": 0.1665,
+      "step": 8006
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8358965250113496,
+      "learning_rate": 1.3421835006065821e-05,
+      "loss": 0.1845,
+      "step": 8007
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8940788101936157,
+      "learning_rate": 1.3420287413350424e-05,
+      "loss": 0.1943,
+      "step": 8008
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8457027511164052,
+      "learning_rate": 1.3418739727859536e-05,
+      "loss": 0.1884,
+      "step": 8009
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9616924394345931,
+      "learning_rate": 1.3417191949635137e-05,
+      "loss": 0.2005,
+      "step": 8010
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9765691241789635,
+      "learning_rate": 1.3415644078719216e-05,
+      "loss": 0.2,
+      "step": 8011
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0198592589757345,
+      "learning_rate": 1.3414096115153758e-05,
+      "loss": 0.2008,
+      "step": 8012
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.078405481686966,
+      "learning_rate": 1.341254805898075e-05,
+      "loss": 0.1754,
+      "step": 8013
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1520658043098602,
+      "learning_rate": 1.341099991024218e-05,
+      "loss": 0.195,
+      "step": 8014
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8431762267412917,
+      "learning_rate": 1.3409451668980047e-05,
+      "loss": 0.1917,
+      "step": 8015
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9709340793317897,
+      "learning_rate": 1.3407903335236342e-05,
+      "loss": 0.1988,
+      "step": 8016
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9562143034143634,
+      "learning_rate": 1.3406354909053072e-05,
+      "loss": 0.1865,
+      "step": 8017
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8903264345603317,
+      "learning_rate": 1.340480639047223e-05,
+      "loss": 0.1795,
+      "step": 8018
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9493680900295547,
+      "learning_rate": 1.340325777953582e-05,
+      "loss": 0.196,
+      "step": 8019
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.892180765456923,
+      "learning_rate": 1.3401709076285854e-05,
+      "loss": 0.1962,
+      "step": 8020
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0172462805046316,
+      "learning_rate": 1.3400160280764334e-05,
+      "loss": 0.1945,
+      "step": 8021
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8633358976212213,
+      "learning_rate": 1.3398611393013276e-05,
+      "loss": 0.184,
+      "step": 8022
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8679448290497178,
+      "learning_rate": 1.3397062413074692e-05,
+      "loss": 0.2077,
+      "step": 8023
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8340346318477808,
+      "learning_rate": 1.3395513340990599e-05,
+      "loss": 0.1961,
+      "step": 8024
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2509060316174685,
+      "learning_rate": 1.3393964176803014e-05,
+      "loss": 0.2064,
+      "step": 8025
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7697603617815317,
+      "learning_rate": 1.3392414920553958e-05,
+      "loss": 0.1696,
+      "step": 8026
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8871225258106962,
+      "learning_rate": 1.3390865572285456e-05,
+      "loss": 0.191,
+      "step": 8027
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9158079068168967,
+      "learning_rate": 1.3389316132039534e-05,
+      "loss": 0.1744,
+      "step": 8028
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7186316771574106,
+      "learning_rate": 1.3387766599858223e-05,
+      "loss": 0.2006,
+      "step": 8029
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4655735167646904,
+      "learning_rate": 1.338621697578355e-05,
+      "loss": 0.1928,
+      "step": 8030
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.429526068500552,
+      "learning_rate": 1.338466725985755e-05,
+      "loss": 0.1723,
+      "step": 8031
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.057773856701616,
+      "learning_rate": 1.3383117452122259e-05,
+      "loss": 0.1977,
+      "step": 8032
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.035697937100189,
+      "learning_rate": 1.3381567552619716e-05,
+      "loss": 0.1796,
+      "step": 8033
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8543769830582006,
+      "learning_rate": 1.3380017561391964e-05,
+      "loss": 0.1993,
+      "step": 8034
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1232548014489392,
+      "learning_rate": 1.3378467478481043e-05,
+      "loss": 0.2145,
+      "step": 8035
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7160729312689819,
+      "learning_rate": 1.3376917303929e-05,
+      "loss": 0.1762,
+      "step": 8036
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0801627888125969,
+      "learning_rate": 1.3375367037777887e-05,
+      "loss": 0.223,
+      "step": 8037
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.247750747667892,
+      "learning_rate": 1.3373816680069749e-05,
+      "loss": 0.1975,
+      "step": 8038
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8611190036702995,
+      "learning_rate": 1.3372266230846647e-05,
+      "loss": 0.2073,
+      "step": 8039
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0154499916864541,
+      "learning_rate": 1.3370715690150631e-05,
+      "loss": 0.1905,
+      "step": 8040
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9680673347591338,
+      "learning_rate": 1.336916505802376e-05,
+      "loss": 0.1972,
+      "step": 8041
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9171675293224171,
+      "learning_rate": 1.3367614334508097e-05,
+      "loss": 0.2005,
+      "step": 8042
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7850773437724924,
+      "learning_rate": 1.3366063519645707e-05,
+      "loss": 0.2221,
+      "step": 8043
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9294976754762169,
+      "learning_rate": 1.3364512613478654e-05,
+      "loss": 0.175,
+      "step": 8044
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8474386472668382,
+      "learning_rate": 1.3362961616049006e-05,
+      "loss": 0.1965,
+      "step": 8045
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8462067143572174,
+      "learning_rate": 1.336141052739883e-05,
+      "loss": 0.2185,
+      "step": 8046
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.895323741516153,
+      "learning_rate": 1.335985934757021e-05,
+      "loss": 0.2025,
+      "step": 8047
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1842591109627791,
+      "learning_rate": 1.3358308076605213e-05,
+      "loss": 0.1718,
+      "step": 8048
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.358975880195966,
+      "learning_rate": 1.3356756714545917e-05,
+      "loss": 0.1839,
+      "step": 8049
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2801457031286867,
+      "learning_rate": 1.3355205261434408e-05,
+      "loss": 0.1898,
+      "step": 8050
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9469890610484961,
+      "learning_rate": 1.3353653717312767e-05,
+      "loss": 0.1808,
+      "step": 8051
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.865045301335842,
+      "learning_rate": 1.335210208222308e-05,
+      "loss": 0.2102,
+      "step": 8052
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0032497809315304,
+      "learning_rate": 1.3350550356207435e-05,
+      "loss": 0.2145,
+      "step": 8053
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1154342561170936,
+      "learning_rate": 1.3348998539307919e-05,
+      "loss": 0.1813,
+      "step": 8054
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0187690617804463,
+      "learning_rate": 1.334744663156663e-05,
+      "loss": 0.1844,
+      "step": 8055
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0827207692673682,
+      "learning_rate": 1.3345894633025662e-05,
+      "loss": 0.2129,
+      "step": 8056
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2555042431760213,
+      "learning_rate": 1.3344342543727115e-05,
+      "loss": 0.2022,
+      "step": 8057
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8291008640026881,
+      "learning_rate": 1.3342790363713088e-05,
+      "loss": 0.182,
+      "step": 8058
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4148639718049554,
+      "learning_rate": 1.3341238093025679e-05,
+      "loss": 0.1941,
+      "step": 8059
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9735277681041818,
+      "learning_rate": 1.3339685731707002e-05,
+      "loss": 0.1892,
+      "step": 8060
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.923651269778692,
+      "learning_rate": 1.3338133279799159e-05,
+      "loss": 0.1852,
+      "step": 8061
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.071944681667006,
+      "learning_rate": 1.3336580737344265e-05,
+      "loss": 0.2174,
+      "step": 8062
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6682334959961802,
+      "learning_rate": 1.3335028104384424e-05,
+      "loss": 0.2173,
+      "step": 8063
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2152012251443567,
+      "learning_rate": 1.3333475380961762e-05,
+      "loss": 0.1892,
+      "step": 8064
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0145161664772557,
+      "learning_rate": 1.3331922567118394e-05,
+      "loss": 0.2089,
+      "step": 8065
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1507511222245936,
+      "learning_rate": 1.3330369662896437e-05,
+      "loss": 0.214,
+      "step": 8066
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9486939578167846,
+      "learning_rate": 1.3328816668338012e-05,
+      "loss": 0.2021,
+      "step": 8067
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.114137032112211,
+      "learning_rate": 1.3327263583485248e-05,
+      "loss": 0.1919,
+      "step": 8068
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1410597004242617,
+      "learning_rate": 1.3325710408380272e-05,
+      "loss": 0.2153,
+      "step": 8069
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8798023649166551,
+      "learning_rate": 1.3324157143065213e-05,
+      "loss": 0.195,
+      "step": 8070
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9157107395312563,
+      "learning_rate": 1.3322603787582205e-05,
+      "loss": 0.1989,
+      "step": 8071
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8099024728710942,
+      "learning_rate": 1.3321050341973378e-05,
+      "loss": 0.2094,
+      "step": 8072
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0920822863428858,
+      "learning_rate": 1.3319496806280877e-05,
+      "loss": 0.1846,
+      "step": 8073
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.260886118708162,
+      "learning_rate": 1.3317943180546836e-05,
+      "loss": 0.1959,
+      "step": 8074
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0141664566644863,
+      "learning_rate": 1.3316389464813397e-05,
+      "loss": 0.1988,
+      "step": 8075
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8211926777508631,
+      "learning_rate": 1.3314835659122707e-05,
+      "loss": 0.1757,
+      "step": 8076
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.949669333859531,
+      "learning_rate": 1.3313281763516915e-05,
+      "loss": 0.1939,
+      "step": 8077
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9081621642327156,
+      "learning_rate": 1.3311727778038165e-05,
+      "loss": 0.1835,
+      "step": 8078
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9381499951139102,
+      "learning_rate": 1.3310173702728614e-05,
+      "loss": 0.196,
+      "step": 8079
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4136313347465652,
+      "learning_rate": 1.3308619537630416e-05,
+      "loss": 0.2122,
+      "step": 8080
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7596837076372943,
+      "learning_rate": 1.3307065282785723e-05,
+      "loss": 0.2217,
+      "step": 8081
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1243254028756764,
+      "learning_rate": 1.33055109382367e-05,
+      "loss": 0.1903,
+      "step": 8082
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9040116470936199,
+      "learning_rate": 1.3303956504025506e-05,
+      "loss": 0.2046,
+      "step": 8083
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1134412438267782,
+      "learning_rate": 1.3302401980194303e-05,
+      "loss": 0.1973,
+      "step": 8084
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.311444185148507,
+      "learning_rate": 1.3300847366785261e-05,
+      "loss": 0.2263,
+      "step": 8085
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6291026055605924,
+      "learning_rate": 1.3299292663840546e-05,
+      "loss": 0.205,
+      "step": 8086
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1327737332574543,
+      "learning_rate": 1.3297737871402333e-05,
+      "loss": 0.2018,
+      "step": 8087
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9299835949328888,
+      "learning_rate": 1.3296182989512794e-05,
+      "loss": 0.1824,
+      "step": 8088
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8818644971901961,
+      "learning_rate": 1.3294628018214105e-05,
+      "loss": 0.2143,
+      "step": 8089
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9567065848980094,
+      "learning_rate": 1.3293072957548443e-05,
+      "loss": 0.1917,
+      "step": 8090
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9484045798329989,
+      "learning_rate": 1.3291517807557994e-05,
+      "loss": 0.1908,
+      "step": 8091
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0639832334067085,
+      "learning_rate": 1.3289962568284937e-05,
+      "loss": 0.1944,
+      "step": 8092
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2129082756641967,
+      "learning_rate": 1.3288407239771462e-05,
+      "loss": 0.1931,
+      "step": 8093
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5496534950187957,
+      "learning_rate": 1.328685182205975e-05,
+      "loss": 0.1984,
+      "step": 8094
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7701213256886666,
+      "learning_rate": 1.3285296315192e-05,
+      "loss": 0.1883,
+      "step": 8095
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2667584968529702,
+      "learning_rate": 1.32837407192104e-05,
+      "loss": 0.196,
+      "step": 8096
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4217626914391113,
+      "learning_rate": 1.3282185034157151e-05,
+      "loss": 0.2033,
+      "step": 8097
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9766157940575106,
+      "learning_rate": 1.3280629260074442e-05,
+      "loss": 0.1865,
+      "step": 8098
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7803982157049401,
+      "learning_rate": 1.3279073397004485e-05,
+      "loss": 0.1695,
+      "step": 8099
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0041553192413004,
+      "learning_rate": 1.3277517444989476e-05,
+      "loss": 0.1876,
+      "step": 8100
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8422004848718346,
+      "learning_rate": 1.327596140407162e-05,
+      "loss": 0.195,
+      "step": 8101
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.126805813050145,
+      "learning_rate": 1.3274405274293122e-05,
+      "loss": 0.1922,
+      "step": 8102
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3489940250126804,
+      "learning_rate": 1.3272849055696203e-05,
+      "loss": 0.1994,
+      "step": 8103
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2051419046998122,
+      "learning_rate": 1.3271292748323064e-05,
+      "loss": 0.2189,
+      "step": 8104
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9861422782076765,
+      "learning_rate": 1.3269736352215925e-05,
+      "loss": 0.2058,
+      "step": 8105
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7770971461313525,
+      "learning_rate": 1.3268179867417004e-05,
+      "loss": 0.1905,
+      "step": 8106
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9277715775572186,
+      "learning_rate": 1.3266623293968518e-05,
+      "loss": 0.1919,
+      "step": 8107
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9695669138804043,
+      "learning_rate": 1.326506663191269e-05,
+      "loss": 0.1894,
+      "step": 8108
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9098778962285857,
+      "learning_rate": 1.3263509881291748e-05,
+      "loss": 0.1768,
+      "step": 8109
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8360349993777317,
+      "learning_rate": 1.3261953042147915e-05,
+      "loss": 0.1895,
+      "step": 8110
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.147216914825333,
+      "learning_rate": 1.326039611452342e-05,
+      "loss": 0.1896,
+      "step": 8111
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8526896466281547,
+      "learning_rate": 1.3258839098460496e-05,
+      "loss": 0.1937,
+      "step": 8112
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1435652297643273,
+      "learning_rate": 1.325728199400138e-05,
+      "loss": 0.2093,
+      "step": 8113
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8185991738538979,
+      "learning_rate": 1.3255724801188305e-05,
+      "loss": 0.1922,
+      "step": 8114
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0561514708661517,
+      "learning_rate": 1.325416752006351e-05,
+      "loss": 0.191,
+      "step": 8115
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7351805421960459,
+      "learning_rate": 1.3252610150669236e-05,
+      "loss": 0.1879,
+      "step": 8116
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.522389693987051,
+      "learning_rate": 1.3251052693047732e-05,
+      "loss": 0.1902,
+      "step": 8117
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9716643739300673,
+      "learning_rate": 1.324949514724124e-05,
+      "loss": 0.1864,
+      "step": 8118
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.372780519038876,
+      "learning_rate": 1.3247937513292007e-05,
+      "loss": 0.2246,
+      "step": 8119
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1258389056492442,
+      "learning_rate": 1.3246379791242284e-05,
+      "loss": 0.2059,
+      "step": 8120
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6779761807315811,
+      "learning_rate": 1.3244821981134326e-05,
+      "loss": 0.1742,
+      "step": 8121
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9024791913390288,
+      "learning_rate": 1.324326408301039e-05,
+      "loss": 0.214,
+      "step": 8122
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9387075044127428,
+      "learning_rate": 1.3241706096912731e-05,
+      "loss": 0.1813,
+      "step": 8123
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0895957947682025,
+      "learning_rate": 1.324014802288361e-05,
+      "loss": 0.2017,
+      "step": 8124
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8091951199061025,
+      "learning_rate": 1.3238589860965295e-05,
+      "loss": 0.1847,
+      "step": 8125
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9760857115992949,
+      "learning_rate": 1.3237031611200044e-05,
+      "loss": 0.2073,
+      "step": 8126
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8664661099865184,
+      "learning_rate": 1.3235473273630128e-05,
+      "loss": 0.1859,
+      "step": 8127
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.08176486103619,
+      "learning_rate": 1.3233914848297817e-05,
+      "loss": 0.1751,
+      "step": 8128
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0521867252969814,
+      "learning_rate": 1.3232356335245381e-05,
+      "loss": 0.1917,
+      "step": 8129
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8501926151086432,
+      "learning_rate": 1.3230797734515102e-05,
+      "loss": 0.1832,
+      "step": 8130
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8890514909572997,
+      "learning_rate": 1.3229239046149249e-05,
+      "loss": 0.1979,
+      "step": 8131
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0085527273462358,
+      "learning_rate": 1.3227680270190106e-05,
+      "loss": 0.1977,
+      "step": 8132
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7789829508973025,
+      "learning_rate": 1.322612140667995e-05,
+      "loss": 0.1936,
+      "step": 8133
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0270694342051736,
+      "learning_rate": 1.3224562455661069e-05,
+      "loss": 0.2294,
+      "step": 8134
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8857809528499025,
+      "learning_rate": 1.3223003417175755e-05,
+      "loss": 0.2033,
+      "step": 8135
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8872779913474352,
+      "learning_rate": 1.3221444291266288e-05,
+      "loss": 0.209,
+      "step": 8136
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9116143631173013,
+      "learning_rate": 1.3219885077974959e-05,
+      "loss": 0.1896,
+      "step": 8137
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8408863730472417,
+      "learning_rate": 1.321832577734407e-05,
+      "loss": 0.2315,
+      "step": 8138
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9435006909990574,
+      "learning_rate": 1.3216766389415909e-05,
+      "loss": 0.2047,
+      "step": 8139
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9122559504169311,
+      "learning_rate": 1.321520691423278e-05,
+      "loss": 0.1946,
+      "step": 8140
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9570472472937975,
+      "learning_rate": 1.3213647351836985e-05,
+      "loss": 0.1801,
+      "step": 8141
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3682519206358967,
+      "learning_rate": 1.3212087702270817e-05,
+      "loss": 0.2392,
+      "step": 8142
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7131643812133875,
+      "learning_rate": 1.3210527965576594e-05,
+      "loss": 0.165,
+      "step": 8143
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0247955772632986,
+      "learning_rate": 1.3208968141796616e-05,
+      "loss": 0.169,
+      "step": 8144
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7389961092282391,
+      "learning_rate": 1.3207408230973198e-05,
+      "loss": 0.1991,
+      "step": 8145
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.987690758612029,
+      "learning_rate": 1.3205848233148649e-05,
+      "loss": 0.2091,
+      "step": 8146
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.162320781384892,
+      "learning_rate": 1.3204288148365285e-05,
+      "loss": 0.1972,
+      "step": 8147
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.259879163113445,
+      "learning_rate": 1.3202727976665426e-05,
+      "loss": 0.1776,
+      "step": 8148
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6173310688838565,
+      "learning_rate": 1.320116771809139e-05,
+      "loss": 0.2075,
+      "step": 8149
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.15319608640954,
+      "learning_rate": 1.3199607372685497e-05,
+      "loss": 0.2367,
+      "step": 8150
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1085735482195873,
+      "learning_rate": 1.3198046940490072e-05,
+      "loss": 0.1925,
+      "step": 8151
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9526619299240466,
+      "learning_rate": 1.3196486421547447e-05,
+      "loss": 0.2255,
+      "step": 8152
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.186250935829803,
+      "learning_rate": 1.3194925815899946e-05,
+      "loss": 0.1938,
+      "step": 8153
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9462820876927331,
+      "learning_rate": 1.3193365123589904e-05,
+      "loss": 0.2193,
+      "step": 8154
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0680716000128938,
+      "learning_rate": 1.3191804344659647e-05,
+      "loss": 0.2051,
+      "step": 8155
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.890509111019689,
+      "learning_rate": 1.319024347915152e-05,
+      "loss": 0.197,
+      "step": 8156
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9695353673509965,
+      "learning_rate": 1.3188682527107856e-05,
+      "loss": 0.2027,
+      "step": 8157
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8790321917308074,
+      "learning_rate": 1.3187121488571001e-05,
+      "loss": 0.1975,
+      "step": 8158
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0904526851948093,
+      "learning_rate": 1.3185560363583294e-05,
+      "loss": 0.2031,
+      "step": 8159
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0304998013343536,
+      "learning_rate": 1.3183999152187084e-05,
+      "loss": 0.1907,
+      "step": 8160
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3517359749264424,
+      "learning_rate": 1.3182437854424716e-05,
+      "loss": 0.2229,
+      "step": 8161
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9239570982372851,
+      "learning_rate": 1.3180876470338545e-05,
+      "loss": 0.2058,
+      "step": 8162
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8717308844191373,
+      "learning_rate": 1.3179314999970915e-05,
+      "loss": 0.1981,
+      "step": 8163
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8246230871063063,
+      "learning_rate": 1.3177753443364188e-05,
+      "loss": 0.1909,
+      "step": 8164
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4695548290453426,
+      "learning_rate": 1.317619180056072e-05,
+      "loss": 0.1985,
+      "step": 8165
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8400845928983217,
+      "learning_rate": 1.317463007160287e-05,
+      "loss": 0.1846,
+      "step": 8166
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7093872820281457,
+      "learning_rate": 1.3173068256533e-05,
+      "loss": 0.1812,
+      "step": 8167
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9064441522202039,
+      "learning_rate": 1.3171506355393473e-05,
+      "loss": 0.2122,
+      "step": 8168
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.367856901127145,
+      "learning_rate": 1.3169944368226655e-05,
+      "loss": 0.2047,
+      "step": 8169
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8530899850369583,
+      "learning_rate": 1.3168382295074923e-05,
+      "loss": 0.1974,
+      "step": 8170
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9398278061149311,
+      "learning_rate": 1.316682013598064e-05,
+      "loss": 0.1778,
+      "step": 8171
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2818932363910829,
+      "learning_rate": 1.3165257890986178e-05,
+      "loss": 0.1952,
+      "step": 8172
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7649808941738352,
+      "learning_rate": 1.3163695560133922e-05,
+      "loss": 0.198,
+      "step": 8173
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2894358803233346,
+      "learning_rate": 1.3162133143466242e-05,
+      "loss": 0.1856,
+      "step": 8174
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.665193046043179,
+      "learning_rate": 1.3160570641025526e-05,
+      "loss": 0.1955,
+      "step": 8175
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.157724088983787,
+      "learning_rate": 1.3159008052854147e-05,
+      "loss": 0.1865,
+      "step": 8176
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0373369719316299,
+      "learning_rate": 1.3157445378994498e-05,
+      "loss": 0.2099,
+      "step": 8177
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8229843102712258,
+      "learning_rate": 1.3155882619488967e-05,
+      "loss": 0.1947,
+      "step": 8178
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0086904947534578,
+      "learning_rate": 1.315431977437994e-05,
+      "loss": 0.2151,
+      "step": 8179
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1766623038489123,
+      "learning_rate": 1.3152756843709814e-05,
+      "loss": 0.1941,
+      "step": 8180
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.190846556534522,
+      "learning_rate": 1.3151193827520975e-05,
+      "loss": 0.1877,
+      "step": 8181
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2251108654586467,
+      "learning_rate": 1.3149630725855828e-05,
+      "loss": 0.2026,
+      "step": 8182
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0803278763437667,
+      "learning_rate": 1.314806753875677e-05,
+      "loss": 0.1935,
+      "step": 8183
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9777490138647087,
+      "learning_rate": 1.3146504266266202e-05,
+      "loss": 0.2,
+      "step": 8184
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8264047600720581,
+      "learning_rate": 1.3144940908426532e-05,
+      "loss": 0.1841,
+      "step": 8185
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5967147063009872,
+      "learning_rate": 1.3143377465280155e-05,
+      "loss": 0.1837,
+      "step": 8186
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9558789187028566,
+      "learning_rate": 1.3141813936869494e-05,
+      "loss": 0.2139,
+      "step": 8187
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8320678707779134,
+      "learning_rate": 1.314025032323695e-05,
+      "loss": 0.201,
+      "step": 8188
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9471377988177363,
+      "learning_rate": 1.3138686624424937e-05,
+      "loss": 0.2062,
+      "step": 8189
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9599458958587899,
+      "learning_rate": 1.313712284047587e-05,
+      "loss": 0.1865,
+      "step": 8190
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3681755844392067,
+      "learning_rate": 1.3135558971432172e-05,
+      "loss": 0.1965,
+      "step": 8191
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8330320087338403,
+      "learning_rate": 1.3133995017336259e-05,
+      "loss": 0.2007,
+      "step": 8192
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9023380179478936,
+      "learning_rate": 1.3132430978230555e-05,
+      "loss": 0.182,
+      "step": 8193
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4851398984555777,
+      "learning_rate": 1.3130866854157482e-05,
+      "loss": 0.1946,
+      "step": 8194
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8570937044939294,
+      "learning_rate": 1.312930264515947e-05,
+      "loss": 0.1703,
+      "step": 8195
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3065861691308562,
+      "learning_rate": 1.3127738351278946e-05,
+      "loss": 0.1996,
+      "step": 8196
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3622707775354201,
+      "learning_rate": 1.3126173972558345e-05,
+      "loss": 0.1964,
+      "step": 8197
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1894897579210884,
+      "learning_rate": 1.3124609509040095e-05,
+      "loss": 0.1916,
+      "step": 8198
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0950019770268824,
+      "learning_rate": 1.3123044960766638e-05,
+      "loss": 0.1734,
+      "step": 8199
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4604954107534205,
+      "learning_rate": 1.3121480327780409e-05,
+      "loss": 0.1936,
+      "step": 8200
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0583170625721745,
+      "learning_rate": 1.311991561012385e-05,
+      "loss": 0.1843,
+      "step": 8201
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0871249935829725,
+      "learning_rate": 1.3118350807839404e-05,
+      "loss": 0.2015,
+      "step": 8202
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1809231282154058,
+      "learning_rate": 1.3116785920969517e-05,
+      "loss": 0.1884,
+      "step": 8203
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1723187003173714,
+      "learning_rate": 1.3115220949556635e-05,
+      "loss": 0.1776,
+      "step": 8204
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4721516041880833,
+      "learning_rate": 1.3113655893643208e-05,
+      "loss": 0.1993,
+      "step": 8205
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3186840370466455,
+      "learning_rate": 1.3112090753271693e-05,
+      "loss": 0.1927,
+      "step": 8206
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0011682661603702,
+      "learning_rate": 1.3110525528484535e-05,
+      "loss": 0.1978,
+      "step": 8207
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.07523059426114,
+      "learning_rate": 1.3108960219324201e-05,
+      "loss": 0.1978,
+      "step": 8208
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2301338016941046,
+      "learning_rate": 1.3107394825833142e-05,
+      "loss": 0.192,
+      "step": 8209
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.430169746032634,
+      "learning_rate": 1.3105829348053824e-05,
+      "loss": 0.2017,
+      "step": 8210
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9970915919775609,
+      "learning_rate": 1.310426378602871e-05,
+      "loss": 0.202,
+      "step": 8211
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7391855967398469,
+      "learning_rate": 1.3102698139800266e-05,
+      "loss": 0.2163,
+      "step": 8212
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.366594790036297,
+      "learning_rate": 1.3101132409410957e-05,
+      "loss": 0.2044,
+      "step": 8213
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.329529746225666,
+      "learning_rate": 1.309956659490326e-05,
+      "loss": 0.2073,
+      "step": 8214
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.864839755199251,
+      "learning_rate": 1.3098000696319642e-05,
+      "loss": 0.1674,
+      "step": 8215
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9127792797131793,
+      "learning_rate": 1.3096434713702579e-05,
+      "loss": 0.1944,
+      "step": 8216
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8344816953431473,
+      "learning_rate": 1.309486864709455e-05,
+      "loss": 0.1928,
+      "step": 8217
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7475117512260006,
+      "learning_rate": 1.3093302496538036e-05,
+      "loss": 0.1854,
+      "step": 8218
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8922062545828179,
+      "learning_rate": 1.3091736262075516e-05,
+      "loss": 0.182,
+      "step": 8219
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0794784191756126,
+      "learning_rate": 1.3090169943749475e-05,
+      "loss": 0.2091,
+      "step": 8220
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1386734232776148,
+      "learning_rate": 1.3088603541602401e-05,
+      "loss": 0.1989,
+      "step": 8221
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.97328561492684,
+      "learning_rate": 1.3087037055676782e-05,
+      "loss": 0.2353,
+      "step": 8222
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7729143032057222,
+      "learning_rate": 1.3085470486015106e-05,
+      "loss": 0.2031,
+      "step": 8223
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9819343230584475,
+      "learning_rate": 1.308390383265987e-05,
+      "loss": 0.2244,
+      "step": 8224
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8593104249308028,
+      "learning_rate": 1.3082337095653569e-05,
+      "loss": 0.1637,
+      "step": 8225
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.200272710074114,
+      "learning_rate": 1.30807702750387e-05,
+      "loss": 0.2117,
+      "step": 8226
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3002822205995266,
+      "learning_rate": 1.307920337085776e-05,
+      "loss": 0.2086,
+      "step": 8227
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.870542691222419,
+      "learning_rate": 1.3077636383153258e-05,
+      "loss": 0.1817,
+      "step": 8228
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9153276421843458,
+      "learning_rate": 1.3076069311967696e-05,
+      "loss": 0.2096,
+      "step": 8229
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2293731778608028,
+      "learning_rate": 1.3074502157343575e-05,
+      "loss": 0.2198,
+      "step": 8230
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8688580667543977,
+      "learning_rate": 1.3072934919323414e-05,
+      "loss": 0.1632,
+      "step": 8231
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.409556841159473,
+      "learning_rate": 1.307136759794972e-05,
+      "loss": 0.1991,
+      "step": 8232
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8755205631804428,
+      "learning_rate": 1.3069800193265005e-05,
+      "loss": 0.1977,
+      "step": 8233
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9340942948944939,
+      "learning_rate": 1.3068232705311784e-05,
+      "loss": 0.188,
+      "step": 8234
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9190292523868503,
+      "learning_rate": 1.3066665134132584e-05,
+      "loss": 0.2043,
+      "step": 8235
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1836600836195899,
+      "learning_rate": 1.3065097479769915e-05,
+      "loss": 0.2094,
+      "step": 8236
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8521202367831131,
+      "learning_rate": 1.3063529742266304e-05,
+      "loss": 0.2102,
+      "step": 8237
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0102823962086578,
+      "learning_rate": 1.3061961921664276e-05,
+      "loss": 0.1673,
+      "step": 8238
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8387168527811484,
+      "learning_rate": 1.3060394018006357e-05,
+      "loss": 0.1924,
+      "step": 8239
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0532826607045784,
+      "learning_rate": 1.305882603133508e-05,
+      "loss": 0.1784,
+      "step": 8240
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8118358299802393,
+      "learning_rate": 1.305725796169297e-05,
+      "loss": 0.1818,
+      "step": 8241
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9002539298249029,
+      "learning_rate": 1.3055689809122569e-05,
+      "loss": 0.2066,
+      "step": 8242
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9585852126504476,
+      "learning_rate": 1.3054121573666408e-05,
+      "loss": 0.168,
+      "step": 8243
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8460014460500977,
+      "learning_rate": 1.3052553255367024e-05,
+      "loss": 0.1781,
+      "step": 8244
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0891528902857406,
+      "learning_rate": 1.3050984854266963e-05,
+      "loss": 0.197,
+      "step": 8245
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7424589594293574,
+      "learning_rate": 1.3049416370408768e-05,
+      "loss": 0.1609,
+      "step": 8246
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.135546053361693,
+      "learning_rate": 1.3047847803834976e-05,
+      "loss": 0.2029,
+      "step": 8247
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0569142092023314,
+      "learning_rate": 1.3046279154588146e-05,
+      "loss": 0.2046,
+      "step": 8248
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7796741457759377,
+      "learning_rate": 1.3044710422710818e-05,
+      "loss": 0.2229,
+      "step": 8249
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0588020023041,
+      "learning_rate": 1.3043141608245551e-05,
+      "loss": 0.2038,
+      "step": 8250
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1450205997702398,
+      "learning_rate": 1.3041572711234893e-05,
+      "loss": 0.1869,
+      "step": 8251
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2504897105794466,
+      "learning_rate": 1.3040003731721402e-05,
+      "loss": 0.2002,
+      "step": 8252
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7303802878941769,
+      "learning_rate": 1.3038434669747644e-05,
+      "loss": 0.1793,
+      "step": 8253
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1163159311554531,
+      "learning_rate": 1.3036865525356168e-05,
+      "loss": 0.1896,
+      "step": 8254
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3143525222797499,
+      "learning_rate": 1.3035296298589549e-05,
+      "loss": 0.2188,
+      "step": 8255
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1341176124382015,
+      "learning_rate": 1.3033726989490341e-05,
+      "loss": 0.2011,
+      "step": 8256
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1201148079058703,
+      "learning_rate": 1.303215759810112e-05,
+      "loss": 0.1983,
+      "step": 8257
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.014726310011175,
+      "learning_rate": 1.3030588124464453e-05,
+      "loss": 0.1923,
+      "step": 8258
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5739238510028213,
+      "learning_rate": 1.302901856862291e-05,
+      "loss": 0.1954,
+      "step": 8259
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6029945998288895,
+      "learning_rate": 1.302744893061907e-05,
+      "loss": 0.2006,
+      "step": 8260
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.152759348936254,
+      "learning_rate": 1.3025879210495505e-05,
+      "loss": 0.2025,
+      "step": 8261
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9133265209563681,
+      "learning_rate": 1.3024309408294795e-05,
+      "loss": 0.1915,
+      "step": 8262
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8821852254147305,
+      "learning_rate": 1.3022739524059521e-05,
+      "loss": 0.1704,
+      "step": 8263
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7904481085547684,
+      "learning_rate": 1.3021169557832269e-05,
+      "loss": 0.187,
+      "step": 8264
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9535276347236445,
+      "learning_rate": 1.301959950965562e-05,
+      "loss": 0.1792,
+      "step": 8265
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4203925883538724,
+      "learning_rate": 1.3018029379572163e-05,
+      "loss": 0.2237,
+      "step": 8266
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.147777070235677,
+      "learning_rate": 1.3016459167624494e-05,
+      "loss": 0.2015,
+      "step": 8267
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8485290112978522,
+      "learning_rate": 1.3014888873855194e-05,
+      "loss": 0.2094,
+      "step": 8268
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8724810785819876,
+      "learning_rate": 1.3013318498306864e-05,
+      "loss": 0.1882,
+      "step": 8269
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0869723494866173,
+      "learning_rate": 1.3011748041022101e-05,
+      "loss": 0.2007,
+      "step": 8270
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8982814311359465,
+      "learning_rate": 1.3010177502043502e-05,
+      "loss": 0.2056,
+      "step": 8271
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9975730113802105,
+      "learning_rate": 1.3008606881413668e-05,
+      "loss": 0.2134,
+      "step": 8272
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1142646889120968,
+      "learning_rate": 1.3007036179175203e-05,
+      "loss": 0.2105,
+      "step": 8273
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8904410113100958,
+      "learning_rate": 1.300546539537071e-05,
+      "loss": 0.2045,
+      "step": 8274
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9203103644433339,
+      "learning_rate": 1.3003894530042803e-05,
+      "loss": 0.2076,
+      "step": 8275
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9444630885033088,
+      "learning_rate": 1.3002323583234082e-05,
+      "loss": 0.1697,
+      "step": 8276
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8244494027922004,
+      "learning_rate": 1.3000752554987166e-05,
+      "loss": 0.1908,
+      "step": 8277
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8985091790541165,
+      "learning_rate": 1.2999181445344666e-05,
+      "loss": 0.1874,
+      "step": 8278
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.04902462783325,
+      "learning_rate": 1.2997610254349203e-05,
+      "loss": 0.1777,
+      "step": 8279
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8914789760341593,
+      "learning_rate": 1.299603898204339e-05,
+      "loss": 0.2008,
+      "step": 8280
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8652214926690659,
+      "learning_rate": 1.2994467628469853e-05,
+      "loss": 0.1843,
+      "step": 8281
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7753537627462095,
+      "learning_rate": 1.299289619367121e-05,
+      "loss": 0.1718,
+      "step": 8282
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9956931480728481,
+      "learning_rate": 1.299132467769009e-05,
+      "loss": 0.1941,
+      "step": 8283
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.201701752879004,
+      "learning_rate": 1.2989753080569119e-05,
+      "loss": 0.2011,
+      "step": 8284
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9191160847591369,
+      "learning_rate": 1.2988181402350926e-05,
+      "loss": 0.1923,
+      "step": 8285
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9220969801360414,
+      "learning_rate": 1.2986609643078145e-05,
+      "loss": 0.2123,
+      "step": 8286
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9169015302630571,
+      "learning_rate": 1.2985037802793405e-05,
+      "loss": 0.2273,
+      "step": 8287
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8521753910449956,
+      "learning_rate": 1.298346588153935e-05,
+      "loss": 0.2115,
+      "step": 8288
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2120447235605898,
+      "learning_rate": 1.2981893879358616e-05,
+      "loss": 0.2047,
+      "step": 8289
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.101266182978855,
+      "learning_rate": 1.2980321796293838e-05,
+      "loss": 0.2001,
+      "step": 8290
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.174558001806414,
+      "learning_rate": 1.2978749632387665e-05,
+      "loss": 0.1848,
+      "step": 8291
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8953418782568198,
+      "learning_rate": 1.297717738768274e-05,
+      "loss": 0.2,
+      "step": 8292
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6475492947463053,
+      "learning_rate": 1.297560506222171e-05,
+      "loss": 0.1929,
+      "step": 8293
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.157541240282237,
+      "learning_rate": 1.2974032656047223e-05,
+      "loss": 0.2041,
+      "step": 8294
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4156656473816827,
+      "learning_rate": 1.2972460169201933e-05,
+      "loss": 0.1935,
+      "step": 8295
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9467195639374285,
+      "learning_rate": 1.2970887601728495e-05,
+      "loss": 0.2219,
+      "step": 8296
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6312858345818504,
+      "learning_rate": 1.2969314953669563e-05,
+      "loss": 0.1676,
+      "step": 8297
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4278720908189264,
+      "learning_rate": 1.2967742225067792e-05,
+      "loss": 0.1909,
+      "step": 8298
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9080688833167727,
+      "learning_rate": 1.2966169415965847e-05,
+      "loss": 0.1995,
+      "step": 8299
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9863305922628434,
+      "learning_rate": 1.296459652640639e-05,
+      "loss": 0.1974,
+      "step": 8300
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9664763132904046,
+      "learning_rate": 1.2963023556432083e-05,
+      "loss": 0.1962,
+      "step": 8301
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9776353817339578,
+      "learning_rate": 1.2961450506085597e-05,
+      "loss": 0.1841,
+      "step": 8302
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8443028312632579,
+      "learning_rate": 1.2959877375409598e-05,
+      "loss": 0.1786,
+      "step": 8303
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9857253074754555,
+      "learning_rate": 1.2958304164446758e-05,
+      "loss": 0.2038,
+      "step": 8304
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8825347887344253,
+      "learning_rate": 1.2956730873239746e-05,
+      "loss": 0.203,
+      "step": 8305
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9936270761193413,
+      "learning_rate": 1.2955157501831248e-05,
+      "loss": 0.206,
+      "step": 8306
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9087815807999129,
+      "learning_rate": 1.2953584050263935e-05,
+      "loss": 0.2033,
+      "step": 8307
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9204698155344063,
+      "learning_rate": 1.2952010518580487e-05,
+      "loss": 0.1886,
+      "step": 8308
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6209462461027426,
+      "learning_rate": 1.2950436906823584e-05,
+      "loss": 0.2012,
+      "step": 8309
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9357386669354606,
+      "learning_rate": 1.2948863215035918e-05,
+      "loss": 0.195,
+      "step": 8310
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8333558655932062,
+      "learning_rate": 1.2947289443260172e-05,
+      "loss": 0.1844,
+      "step": 8311
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3236927019756342,
+      "learning_rate": 1.2945715591539028e-05,
+      "loss": 0.208,
+      "step": 8312
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0149795519327343,
+      "learning_rate": 1.2944141659915184e-05,
+      "loss": 0.1809,
+      "step": 8313
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7928022468669075,
+      "learning_rate": 1.2942567648431333e-05,
+      "loss": 0.1779,
+      "step": 8314
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2245378017402813,
+      "learning_rate": 1.2940993557130166e-05,
+      "loss": 0.1938,
+      "step": 8315
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8517131351401706,
+      "learning_rate": 1.2939419386054384e-05,
+      "loss": 0.2013,
+      "step": 8316
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9578387527059876,
+      "learning_rate": 1.2937845135246682e-05,
+      "loss": 0.1977,
+      "step": 8317
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0551568804695122,
+      "learning_rate": 1.2936270804749769e-05,
+      "loss": 0.2133,
+      "step": 8318
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.849943397128867,
+      "learning_rate": 1.2934696394606344e-05,
+      "loss": 0.1761,
+      "step": 8319
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8992953513527367,
+      "learning_rate": 1.2933121904859111e-05,
+      "loss": 0.2131,
+      "step": 8320
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9666343163181608,
+      "learning_rate": 1.2931547335550782e-05,
+      "loss": 0.205,
+      "step": 8321
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8788238128220377,
+      "learning_rate": 1.2929972686724066e-05,
+      "loss": 0.1943,
+      "step": 8322
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9164644091486011,
+      "learning_rate": 1.2928397958421674e-05,
+      "loss": 0.1861,
+      "step": 8323
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0092684889010668,
+      "learning_rate": 1.2926823150686325e-05,
+      "loss": 0.1942,
+      "step": 8324
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7327146496169548,
+      "learning_rate": 1.2925248263560733e-05,
+      "loss": 0.1946,
+      "step": 8325
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9039676989792145,
+      "learning_rate": 1.2923673297087613e-05,
+      "loss": 0.1961,
+      "step": 8326
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.06902187233108,
+      "learning_rate": 1.2922098251309694e-05,
+      "loss": 0.204,
+      "step": 8327
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9821026300743326,
+      "learning_rate": 1.2920523126269692e-05,
+      "loss": 0.1915,
+      "step": 8328
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.905264461468117,
+      "learning_rate": 1.2918947922010336e-05,
+      "loss": 0.1825,
+      "step": 8329
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8558085429359792,
+      "learning_rate": 1.291737263857435e-05,
+      "loss": 0.1857,
+      "step": 8330
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8567074452471907,
+      "learning_rate": 1.2915797276004469e-05,
+      "loss": 0.1843,
+      "step": 8331
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8409836164037255,
+      "learning_rate": 1.2914221834343423e-05,
+      "loss": 0.2138,
+      "step": 8332
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0864523715789525,
+      "learning_rate": 1.2912646313633945e-05,
+      "loss": 0.1695,
+      "step": 8333
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9739854947766973,
+      "learning_rate": 1.2911070713918772e-05,
+      "loss": 0.2029,
+      "step": 8334
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0051465038200151,
+      "learning_rate": 1.2909495035240638e-05,
+      "loss": 0.2148,
+      "step": 8335
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2388383034363506,
+      "learning_rate": 1.2907919277642287e-05,
+      "loss": 0.1904,
+      "step": 8336
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7647215138884825,
+      "learning_rate": 1.2906343441166465e-05,
+      "loss": 0.1729,
+      "step": 8337
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7929162694707192,
+      "learning_rate": 1.290476752585591e-05,
+      "loss": 0.173,
+      "step": 8338
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8801109553531988,
+      "learning_rate": 1.2903191531753373e-05,
+      "loss": 0.1831,
+      "step": 8339
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8117842995770987,
+      "learning_rate": 1.2901615458901602e-05,
+      "loss": 0.1862,
+      "step": 8340
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8564129234299726,
+      "learning_rate": 1.2900039307343345e-05,
+      "loss": 0.1944,
+      "step": 8341
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8326250384944172,
+      "learning_rate": 1.2898463077121361e-05,
+      "loss": 0.1922,
+      "step": 8342
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8476141151237748,
+      "learning_rate": 1.2896886768278406e-05,
+      "loss": 0.2183,
+      "step": 8343
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9497331360658231,
+      "learning_rate": 1.2895310380857224e-05,
+      "loss": 0.1823,
+      "step": 8344
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.896734197061469,
+      "learning_rate": 1.2893733914900595e-05,
+      "loss": 0.1855,
+      "step": 8345
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9376838004640948,
+      "learning_rate": 1.2892157370451263e-05,
+      "loss": 0.1936,
+      "step": 8346
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2314498483898435,
+      "learning_rate": 1.2890580747552002e-05,
+      "loss": 0.2073,
+      "step": 8347
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9690556766269256,
+      "learning_rate": 1.2889004046245574e-05,
+      "loss": 0.1872,
+      "step": 8348
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8238006996316732,
+      "learning_rate": 1.2887427266574748e-05,
+      "loss": 0.2207,
+      "step": 8349
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8068928897313348,
+      "learning_rate": 1.2885850408582295e-05,
+      "loss": 0.1946,
+      "step": 8350
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.859272390151615,
+      "learning_rate": 1.2884273472310986e-05,
+      "loss": 0.2203,
+      "step": 8351
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.946612446604715,
+      "learning_rate": 1.2882696457803597e-05,
+      "loss": 0.159,
+      "step": 8352
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9089696914287012,
+      "learning_rate": 1.2881119365102901e-05,
+      "loss": 0.2278,
+      "step": 8353
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2959784235147993,
+      "learning_rate": 1.2879542194251681e-05,
+      "loss": 0.1795,
+      "step": 8354
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8042643958346923,
+      "learning_rate": 1.2877964945292717e-05,
+      "loss": 0.1843,
+      "step": 8355
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3629437841374252,
+      "learning_rate": 1.2876387618268793e-05,
+      "loss": 0.2049,
+      "step": 8356
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.958231619240513,
+      "learning_rate": 1.2874810213222689e-05,
+      "loss": 0.1776,
+      "step": 8357
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5744423044538112,
+      "learning_rate": 1.2873232730197197e-05,
+      "loss": 0.2109,
+      "step": 8358
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.936861568438983,
+      "learning_rate": 1.2871655169235104e-05,
+      "loss": 0.2103,
+      "step": 8359
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8437459959051726,
+      "learning_rate": 1.2870077530379205e-05,
+      "loss": 0.1947,
+      "step": 8360
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.081735761931492,
+      "learning_rate": 1.286849981367229e-05,
+      "loss": 0.1811,
+      "step": 8361
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1563276410499428,
+      "learning_rate": 1.2866922019157155e-05,
+      "loss": 0.1966,
+      "step": 8362
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.80787778076596,
+      "learning_rate": 1.28653441468766e-05,
+      "loss": 0.1989,
+      "step": 8363
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0611653379046442,
+      "learning_rate": 1.2863766196873419e-05,
+      "loss": 0.213,
+      "step": 8364
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.90607110241615,
+      "learning_rate": 1.2862188169190419e-05,
+      "loss": 0.2084,
+      "step": 8365
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8946529178848721,
+      "learning_rate": 1.2860610063870405e-05,
+      "loss": 0.1944,
+      "step": 8366
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8532649529386813,
+      "learning_rate": 1.2859031880956181e-05,
+      "loss": 0.1794,
+      "step": 8367
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.392909454230544,
+      "learning_rate": 1.2857453620490557e-05,
+      "loss": 0.1786,
+      "step": 8368
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8397705370290642,
+      "learning_rate": 1.2855875282516342e-05,
+      "loss": 0.2058,
+      "step": 8369
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0693770104817486,
+      "learning_rate": 1.2854296867076346e-05,
+      "loss": 0.1955,
+      "step": 8370
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7623931235591676,
+      "learning_rate": 1.2852718374213389e-05,
+      "loss": 0.17,
+      "step": 8371
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9831192248267664,
+      "learning_rate": 1.2851139803970285e-05,
+      "loss": 0.2333,
+      "step": 8372
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8845103966984852,
+      "learning_rate": 1.2849561156389851e-05,
+      "loss": 0.1856,
+      "step": 8373
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0431153361684566,
+      "learning_rate": 1.2847982431514911e-05,
+      "loss": 0.1819,
+      "step": 8374
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9885410602615909,
+      "learning_rate": 1.2846403629388285e-05,
+      "loss": 0.1831,
+      "step": 8375
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8649856620753934,
+      "learning_rate": 1.28448247500528e-05,
+      "loss": 0.1959,
+      "step": 8376
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3401866519005952,
+      "learning_rate": 1.2843245793551284e-05,
+      "loss": 0.2121,
+      "step": 8377
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9549802734325415,
+      "learning_rate": 1.2841666759926566e-05,
+      "loss": 0.1829,
+      "step": 8378
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8970244098474727,
+      "learning_rate": 1.2840087649221476e-05,
+      "loss": 0.1897,
+      "step": 8379
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9181250871587772,
+      "learning_rate": 1.283850846147885e-05,
+      "loss": 0.1814,
+      "step": 8380
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9677772504052429,
+      "learning_rate": 1.2836929196741518e-05,
+      "loss": 0.2052,
+      "step": 8381
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3704789262684547,
+      "learning_rate": 1.2835349855052324e-05,
+      "loss": 0.1896,
+      "step": 8382
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0717251508841295,
+      "learning_rate": 1.2833770436454103e-05,
+      "loss": 0.2023,
+      "step": 8383
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9885072449972804,
+      "learning_rate": 1.2832190940989699e-05,
+      "loss": 0.1995,
+      "step": 8384
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.896312751966259,
+      "learning_rate": 1.2830611368701957e-05,
+      "loss": 0.2039,
+      "step": 8385
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7110044731285126,
+      "learning_rate": 1.2829031719633722e-05,
+      "loss": 0.1865,
+      "step": 8386
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9079049703949122,
+      "learning_rate": 1.2827451993827841e-05,
+      "loss": 0.1777,
+      "step": 8387
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2645822150236385,
+      "learning_rate": 1.2825872191327164e-05,
+      "loss": 0.1932,
+      "step": 8388
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7242785606944482,
+      "learning_rate": 1.2824292312174547e-05,
+      "loss": 0.1961,
+      "step": 8389
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8742480736548195,
+      "learning_rate": 1.282271235641284e-05,
+      "loss": 0.1985,
+      "step": 8390
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0083321320072796,
+      "learning_rate": 1.28211323240849e-05,
+      "loss": 0.1837,
+      "step": 8391
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7590299682094305,
+      "learning_rate": 1.2819552215233585e-05,
+      "loss": 0.1861,
+      "step": 8392
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7632714309789075,
+      "learning_rate": 1.2817972029901759e-05,
+      "loss": 0.1687,
+      "step": 8393
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8250873902419275,
+      "learning_rate": 1.2816391768132284e-05,
+      "loss": 0.2067,
+      "step": 8394
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9130439612947502,
+      "learning_rate": 1.2814811429968022e-05,
+      "loss": 0.1961,
+      "step": 8395
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1247865595713258,
+      "learning_rate": 1.2813231015451842e-05,
+      "loss": 0.1934,
+      "step": 8396
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.418166635996874,
+      "learning_rate": 1.2811650524626608e-05,
+      "loss": 0.2263,
+      "step": 8397
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4470070889450257,
+      "learning_rate": 1.2810069957535198e-05,
+      "loss": 0.1728,
+      "step": 8398
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.002144849625383,
+      "learning_rate": 1.2808489314220483e-05,
+      "loss": 0.1923,
+      "step": 8399
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3487231469057854,
+      "learning_rate": 1.2806908594725335e-05,
+      "loss": 0.1714,
+      "step": 8400
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0348615877972056,
+      "learning_rate": 1.280532779909263e-05,
+      "loss": 0.19,
+      "step": 8401
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.818483891086596,
+      "learning_rate": 1.2803746927365252e-05,
+      "loss": 0.1832,
+      "step": 8402
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9141276415044727,
+      "learning_rate": 1.2802165979586084e-05,
+      "loss": 0.2016,
+      "step": 8403
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0645918152401999,
+      "learning_rate": 1.2800584955798e-05,
+      "loss": 0.1963,
+      "step": 8404
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9973262286041973,
+      "learning_rate": 1.2799003856043893e-05,
+      "loss": 0.207,
+      "step": 8405
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9579121869369324,
+      "learning_rate": 1.2797422680366649e-05,
+      "loss": 0.1928,
+      "step": 8406
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0122691887181539,
+      "learning_rate": 1.2795841428809155e-05,
+      "loss": 0.1898,
+      "step": 8407
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8437227917655225,
+      "learning_rate": 1.2794260101414307e-05,
+      "loss": 0.2134,
+      "step": 8408
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.298694300803485,
+      "learning_rate": 1.2792678698224995e-05,
+      "loss": 0.1703,
+      "step": 8409
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8327681044213492,
+      "learning_rate": 1.2791097219284115e-05,
+      "loss": 0.2284,
+      "step": 8410
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6873455193552223,
+      "learning_rate": 1.2789515664634564e-05,
+      "loss": 0.2054,
+      "step": 8411
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0429920359977503,
+      "learning_rate": 1.2787934034319245e-05,
+      "loss": 0.1796,
+      "step": 8412
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.831414425233925,
+      "learning_rate": 1.2786352328381057e-05,
+      "loss": 0.1732,
+      "step": 8413
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0479045262632114,
+      "learning_rate": 1.2784770546862905e-05,
+      "loss": 0.2097,
+      "step": 8414
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9688634782356663,
+      "learning_rate": 1.2783188689807697e-05,
+      "loss": 0.1877,
+      "step": 8415
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9186954634465992,
+      "learning_rate": 1.2781606757258335e-05,
+      "loss": 0.2137,
+      "step": 8416
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8804643222421765,
+      "learning_rate": 1.2780024749257736e-05,
+      "loss": 0.187,
+      "step": 8417
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0031110219542805,
+      "learning_rate": 1.2778442665848805e-05,
+      "loss": 0.2042,
+      "step": 8418
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7537420190969669,
+      "learning_rate": 1.277686050707446e-05,
+      "loss": 0.1772,
+      "step": 8419
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0582699108537157,
+      "learning_rate": 1.277527827297762e-05,
+      "loss": 0.1974,
+      "step": 8420
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1413213859696738,
+      "learning_rate": 1.2773695963601199e-05,
+      "loss": 0.1822,
+      "step": 8421
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8847067204436639,
+      "learning_rate": 1.2772113578988117e-05,
+      "loss": 0.1918,
+      "step": 8422
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3751260454860468,
+      "learning_rate": 1.2770531119181295e-05,
+      "loss": 0.1992,
+      "step": 8423
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.181669385070735,
+      "learning_rate": 1.2768948584223666e-05,
+      "loss": 0.1808,
+      "step": 8424
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9003848410651017,
+      "learning_rate": 1.2767365974158146e-05,
+      "loss": 0.1724,
+      "step": 8425
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1547567937475915,
+      "learning_rate": 1.2765783289027671e-05,
+      "loss": 0.2013,
+      "step": 8426
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9531012582250008,
+      "learning_rate": 1.2764200528875164e-05,
+      "loss": 0.2085,
+      "step": 8427
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4375307799373926,
+      "learning_rate": 1.2762617693743562e-05,
+      "loss": 0.1851,
+      "step": 8428
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9412421733587898,
+      "learning_rate": 1.2761034783675803e-05,
+      "loss": 0.1906,
+      "step": 8429
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9939292690651369,
+      "learning_rate": 1.2759451798714816e-05,
+      "loss": 0.1956,
+      "step": 8430
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7829017756475471,
+      "learning_rate": 1.2757868738903545e-05,
+      "loss": 0.1985,
+      "step": 8431
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1982992843299392,
+      "learning_rate": 1.2756285604284928e-05,
+      "loss": 0.1817,
+      "step": 8432
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9110729910834504,
+      "learning_rate": 1.275470239490191e-05,
+      "loss": 0.2057,
+      "step": 8433
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2121332217693976,
+      "learning_rate": 1.2753119110797432e-05,
+      "loss": 0.2112,
+      "step": 8434
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2017977376703233,
+      "learning_rate": 1.2751535752014444e-05,
+      "loss": 0.1876,
+      "step": 8435
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.162603626637288,
+      "learning_rate": 1.274995231859589e-05,
+      "loss": 0.161,
+      "step": 8436
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8380059596350338,
+      "learning_rate": 1.2748368810584725e-05,
+      "loss": 0.1758,
+      "step": 8437
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.424109898645111,
+      "learning_rate": 1.2746785228023904e-05,
+      "loss": 0.1961,
+      "step": 8438
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3728472348261007,
+      "learning_rate": 1.2745201570956379e-05,
+      "loss": 0.1872,
+      "step": 8439
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2101805295026007,
+      "learning_rate": 1.27436178394251e-05,
+      "loss": 0.1694,
+      "step": 8440
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9012241586461739,
+      "learning_rate": 1.2742034033473037e-05,
+      "loss": 0.1645,
+      "step": 8441
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4716426150058821,
+      "learning_rate": 1.2740450153143144e-05,
+      "loss": 0.1965,
+      "step": 8442
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8971699492952784,
+      "learning_rate": 1.2738866198478388e-05,
+      "loss": 0.2057,
+      "step": 8443
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8749718882080885,
+      "learning_rate": 1.2737282169521732e-05,
+      "loss": 0.1899,
+      "step": 8444
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9340663748135312,
+      "learning_rate": 1.2735698066316138e-05,
+      "loss": 0.2032,
+      "step": 8445
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9371509604709034,
+      "learning_rate": 1.2734113888904584e-05,
+      "loss": 0.1909,
+      "step": 8446
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2213004935273502,
+      "learning_rate": 1.2732529637330036e-05,
+      "loss": 0.1849,
+      "step": 8447
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9992218155973259,
+      "learning_rate": 1.2730945311635465e-05,
+      "loss": 0.198,
+      "step": 8448
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8369839626639749,
+      "learning_rate": 1.272936091186385e-05,
+      "loss": 0.1964,
+      "step": 8449
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9779530137060174,
+      "learning_rate": 1.2727776438058166e-05,
+      "loss": 0.2013,
+      "step": 8450
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.746936294451183,
+      "learning_rate": 1.2726191890261393e-05,
+      "loss": 0.19,
+      "step": 8451
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2654470125593666,
+      "learning_rate": 1.272460726851651e-05,
+      "loss": 0.1716,
+      "step": 8452
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6691442469544424,
+      "learning_rate": 1.2723022572866497e-05,
+      "loss": 0.1844,
+      "step": 8453
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4040660915184713,
+      "learning_rate": 1.2721437803354348e-05,
+      "loss": 0.2178,
+      "step": 8454
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.168892417400603,
+      "learning_rate": 1.2719852960023043e-05,
+      "loss": 0.206,
+      "step": 8455
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2688061946410185,
+      "learning_rate": 1.2718268042915574e-05,
+      "loss": 0.1774,
+      "step": 8456
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2904127430747385,
+      "learning_rate": 1.271668305207493e-05,
+      "loss": 0.206,
+      "step": 8457
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1834309738463864,
+      "learning_rate": 1.2715097987544104e-05,
+      "loss": 0.2018,
+      "step": 8458
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3807424297364006,
+      "learning_rate": 1.2713512849366092e-05,
+      "loss": 0.1826,
+      "step": 8459
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1125232250625734,
+      "learning_rate": 1.2711927637583892e-05,
+      "loss": 0.1961,
+      "step": 8460
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8912732082138243,
+      "learning_rate": 1.2710342352240498e-05,
+      "loss": 0.2065,
+      "step": 8461
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2823631484358686,
+      "learning_rate": 1.270875699337892e-05,
+      "loss": 0.2041,
+      "step": 8462
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9186526027198525,
+      "learning_rate": 1.270717156104215e-05,
+      "loss": 0.2414,
+      "step": 8463
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9547704881649449,
+      "learning_rate": 1.2705586055273202e-05,
+      "loss": 0.2094,
+      "step": 8464
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9810576545703948,
+      "learning_rate": 1.2704000476115079e-05,
+      "loss": 0.218,
+      "step": 8465
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8829659202427127,
+      "learning_rate": 1.2702414823610791e-05,
+      "loss": 0.1928,
+      "step": 8466
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0581841757246446,
+      "learning_rate": 1.2700829097803347e-05,
+      "loss": 0.1974,
+      "step": 8467
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9613440736632844,
+      "learning_rate": 1.2699243298735762e-05,
+      "loss": 0.189,
+      "step": 8468
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1715634892409108,
+      "learning_rate": 1.2697657426451051e-05,
+      "loss": 0.1986,
+      "step": 8469
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9137834082634491,
+      "learning_rate": 1.2696071480992229e-05,
+      "loss": 0.2278,
+      "step": 8470
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8783778890963508,
+      "learning_rate": 1.2694485462402315e-05,
+      "loss": 0.1965,
+      "step": 8471
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1505391364195132,
+      "learning_rate": 1.269289937072433e-05,
+      "loss": 0.2037,
+      "step": 8472
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4161874301172996,
+      "learning_rate": 1.2691313206001298e-05,
+      "loss": 0.2112,
+      "step": 8473
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.647897930955205,
+      "learning_rate": 1.2689726968276246e-05,
+      "loss": 0.2061,
+      "step": 8474
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9562300082535361,
+      "learning_rate": 1.2688140657592195e-05,
+      "loss": 0.2063,
+      "step": 8475
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6988876400429254,
+      "learning_rate": 1.2686554273992177e-05,
+      "loss": 0.2024,
+      "step": 8476
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0352568876456878,
+      "learning_rate": 1.2684967817519222e-05,
+      "loss": 0.2233,
+      "step": 8477
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6628966240070668,
+      "learning_rate": 1.2683381288216368e-05,
+      "loss": 0.2163,
+      "step": 8478
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0610061654242153,
+      "learning_rate": 1.268179468612664e-05,
+      "loss": 0.2103,
+      "step": 8479
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0703551730464311,
+      "learning_rate": 1.268020801129308e-05,
+      "loss": 0.1937,
+      "step": 8480
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9026742355707935,
+      "learning_rate": 1.2678621263758728e-05,
+      "loss": 0.2067,
+      "step": 8481
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1872590164339674,
+      "learning_rate": 1.2677034443566623e-05,
+      "loss": 0.1944,
+      "step": 8482
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3809906224953725,
+      "learning_rate": 1.2675447550759807e-05,
+      "loss": 0.1892,
+      "step": 8483
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1300445482997878,
+      "learning_rate": 1.2673860585381329e-05,
+      "loss": 0.1978,
+      "step": 8484
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0810033265516914,
+      "learning_rate": 1.2672273547474225e-05,
+      "loss": 0.17,
+      "step": 8485
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.190920506332146,
+      "learning_rate": 1.2670686437081554e-05,
+      "loss": 0.1965,
+      "step": 8486
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.093795744759877,
+      "learning_rate": 1.2669099254246363e-05,
+      "loss": 0.2049,
+      "step": 8487
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4062279100264012,
+      "learning_rate": 1.2667511999011699e-05,
+      "loss": 0.1928,
+      "step": 8488
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3611497401758716,
+      "learning_rate": 1.2665924671420626e-05,
+      "loss": 0.1897,
+      "step": 8489
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8179653065978323,
+      "learning_rate": 1.2664337271516194e-05,
+      "loss": 0.1996,
+      "step": 8490
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.221152195388021,
+      "learning_rate": 1.2662749799341464e-05,
+      "loss": 0.1651,
+      "step": 8491
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.174205904888376,
+      "learning_rate": 1.2661162254939496e-05,
+      "loss": 0.2103,
+      "step": 8492
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3784424859747435,
+      "learning_rate": 1.2659574638353349e-05,
+      "loss": 0.1805,
+      "step": 8493
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1577475217917517,
+      "learning_rate": 1.2657986949626091e-05,
+      "loss": 0.1849,
+      "step": 8494
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8962746567132944,
+      "learning_rate": 1.2656399188800788e-05,
+      "loss": 0.1944,
+      "step": 8495
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9910380022034406,
+      "learning_rate": 1.2654811355920505e-05,
+      "loss": 0.1889,
+      "step": 8496
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0426069901807775,
+      "learning_rate": 1.2653223451028316e-05,
+      "loss": 0.1993,
+      "step": 8497
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1887980320486227,
+      "learning_rate": 1.2651635474167287e-05,
+      "loss": 0.177,
+      "step": 8498
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0126773212647495,
+      "learning_rate": 1.2650047425380501e-05,
+      "loss": 0.2027,
+      "step": 8499
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.165218024798249,
+      "learning_rate": 1.2648459304711026e-05,
+      "loss": 0.1823,
+      "step": 8500
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9797609981321793,
+      "learning_rate": 1.2646871112201943e-05,
+      "loss": 0.1933,
+      "step": 8501
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.344103428058884,
+      "learning_rate": 1.2645282847896335e-05,
+      "loss": 0.1963,
+      "step": 8502
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2433867429641254,
+      "learning_rate": 1.2643694511837278e-05,
+      "loss": 0.1712,
+      "step": 8503
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3329144189269893,
+      "learning_rate": 1.2642106104067857e-05,
+      "loss": 0.1836,
+      "step": 8504
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7401092387768005,
+      "learning_rate": 1.264051762463116e-05,
+      "loss": 0.2108,
+      "step": 8505
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7026921755147951,
+      "learning_rate": 1.2638929073570273e-05,
+      "loss": 0.1618,
+      "step": 8506
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.188762430576397,
+      "learning_rate": 1.2637340450928284e-05,
+      "loss": 0.2004,
+      "step": 8507
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.169245700509231,
+      "learning_rate": 1.263575175674829e-05,
+      "loss": 0.1656,
+      "step": 8508
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9169441692984407,
+      "learning_rate": 1.2634162991073376e-05,
+      "loss": 0.1608,
+      "step": 8509
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.239622696551675,
+      "learning_rate": 1.2632574153946646e-05,
+      "loss": 0.1836,
+      "step": 8510
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0826608464510832,
+      "learning_rate": 1.263098524541119e-05,
+      "loss": 0.1777,
+      "step": 8511
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5205226845830233,
+      "learning_rate": 1.2629396265510113e-05,
+      "loss": 0.1969,
+      "step": 8512
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9138966349728019,
+      "learning_rate": 1.2627807214286514e-05,
+      "loss": 0.1927,
+      "step": 8513
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9706735417872003,
+      "learning_rate": 1.2626218091783496e-05,
+      "loss": 0.1932,
+      "step": 8514
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0667045902349253,
+      "learning_rate": 1.262462889804416e-05,
+      "loss": 0.1853,
+      "step": 8515
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9375839118068797,
+      "learning_rate": 1.2623039633111623e-05,
+      "loss": 0.1862,
+      "step": 8516
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9210059162958971,
+      "learning_rate": 1.2621450297028984e-05,
+      "loss": 0.1892,
+      "step": 8517
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.321499182230024,
+      "learning_rate": 1.261986088983936e-05,
+      "loss": 0.1814,
+      "step": 8518
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2698154007000069,
+      "learning_rate": 1.2618271411585859e-05,
+      "loss": 0.1871,
+      "step": 8519
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8777129749240363,
+      "learning_rate": 1.26166818623116e-05,
+      "loss": 0.2097,
+      "step": 8520
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.712414234101878,
+      "learning_rate": 1.2615092242059697e-05,
+      "loss": 0.1877,
+      "step": 8521
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0140691144950047,
+      "learning_rate": 1.2613502550873269e-05,
+      "loss": 0.1742,
+      "step": 8522
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5850344252998092,
+      "learning_rate": 1.2611912788795437e-05,
+      "loss": 0.2046,
+      "step": 8523
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.329465016086951,
+      "learning_rate": 1.261032295586932e-05,
+      "loss": 0.1894,
+      "step": 8524
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8296339341431521,
+      "learning_rate": 1.260873305213805e-05,
+      "loss": 0.202,
+      "step": 8525
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9733137634418483,
+      "learning_rate": 1.2607143077644746e-05,
+      "loss": 0.194,
+      "step": 8526
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.160550697515113,
+      "learning_rate": 1.260555303243254e-05,
+      "loss": 0.2085,
+      "step": 8527
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8414652418885628,
+      "learning_rate": 1.2603962916544558e-05,
+      "loss": 0.1943,
+      "step": 8528
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0728029220031154,
+      "learning_rate": 1.2602372730023938e-05,
+      "loss": 0.1731,
+      "step": 8529
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8849396662531649,
+      "learning_rate": 1.2600782472913811e-05,
+      "loss": 0.1975,
+      "step": 8530
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7267393674010236,
+      "learning_rate": 1.259919214525731e-05,
+      "loss": 0.1781,
+      "step": 8531
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8599653618866973,
+      "learning_rate": 1.2597601747097578e-05,
+      "loss": 0.212,
+      "step": 8532
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0116049169051977,
+      "learning_rate": 1.259601127847775e-05,
+      "loss": 0.1909,
+      "step": 8533
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0711190542073004,
+      "learning_rate": 1.259442073944097e-05,
+      "loss": 0.1807,
+      "step": 8534
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0228121141488749,
+      "learning_rate": 1.259283013003038e-05,
+      "loss": 0.2351,
+      "step": 8535
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9452815167442381,
+      "learning_rate": 1.2591239450289127e-05,
+      "loss": 0.2026,
+      "step": 8536
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9610403379351731,
+      "learning_rate": 1.2589648700260359e-05,
+      "loss": 0.221,
+      "step": 8537
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6244814059214097,
+      "learning_rate": 1.2588057879987223e-05,
+      "loss": 0.1703,
+      "step": 8538
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9149145552404859,
+      "learning_rate": 1.2586466989512872e-05,
+      "loss": 0.2023,
+      "step": 8539
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9812119245902255,
+      "learning_rate": 1.2584876028880455e-05,
+      "loss": 0.1755,
+      "step": 8540
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0256757321539571,
+      "learning_rate": 1.258328499813313e-05,
+      "loss": 0.1977,
+      "step": 8541
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0537374073758372,
+      "learning_rate": 1.2581693897314056e-05,
+      "loss": 0.1901,
+      "step": 8542
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9823156564187766,
+      "learning_rate": 1.2580102726466388e-05,
+      "loss": 0.2047,
+      "step": 8543
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1843880734627192,
+      "learning_rate": 1.2578511485633288e-05,
+      "loss": 0.1823,
+      "step": 8544
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.93633345218317,
+      "learning_rate": 1.2576920174857917e-05,
+      "loss": 0.2008,
+      "step": 8545
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.040963176406591,
+      "learning_rate": 1.2575328794183439e-05,
+      "loss": 0.1999,
+      "step": 8546
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0780583875307697,
+      "learning_rate": 1.2573737343653026e-05,
+      "loss": 0.2144,
+      "step": 8547
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9516385201913629,
+      "learning_rate": 1.257214582330984e-05,
+      "loss": 0.1737,
+      "step": 8548
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9613205197820011,
+      "learning_rate": 1.2570554233197054e-05,
+      "loss": 0.1744,
+      "step": 8549
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.019145823439213,
+      "learning_rate": 1.2568962573357837e-05,
+      "loss": 0.2127,
+      "step": 8550
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.873002514272877,
+      "learning_rate": 1.256737084383537e-05,
+      "loss": 0.2042,
+      "step": 8551
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.050032016240503,
+      "learning_rate": 1.2565779044672821e-05,
+      "loss": 0.1894,
+      "step": 8552
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8068052044940806,
+      "learning_rate": 1.256418717591337e-05,
+      "loss": 0.1846,
+      "step": 8553
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3979721822608953,
+      "learning_rate": 1.25625952376002e-05,
+      "loss": 0.1842,
+      "step": 8554
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.547408364045404,
+      "learning_rate": 1.2561003229776485e-05,
+      "loss": 0.182,
+      "step": 8555
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9049443458327565,
+      "learning_rate": 1.2559411152485414e-05,
+      "loss": 0.1992,
+      "step": 8556
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8407326410671497,
+      "learning_rate": 1.2557819005770174e-05,
+      "loss": 0.1639,
+      "step": 8557
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9925588954179055,
+      "learning_rate": 1.2556226789673946e-05,
+      "loss": 0.2002,
+      "step": 8558
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8053222521118775,
+      "learning_rate": 1.2554634504239923e-05,
+      "loss": 0.2065,
+      "step": 8559
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8122917580352957,
+      "learning_rate": 1.2553042149511295e-05,
+      "loss": 0.1815,
+      "step": 8560
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9270132598078689,
+      "learning_rate": 1.2551449725531254e-05,
+      "loss": 0.1895,
+      "step": 8561
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9402093869050032,
+      "learning_rate": 1.2549857232342995e-05,
+      "loss": 0.198,
+      "step": 8562
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9951990611606287,
+      "learning_rate": 1.2548264669989712e-05,
+      "loss": 0.2181,
+      "step": 8563
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.195070825105065,
+      "learning_rate": 1.2546672038514608e-05,
+      "loss": 0.1918,
+      "step": 8564
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8772802211101891,
+      "learning_rate": 1.2545079337960883e-05,
+      "loss": 0.1969,
+      "step": 8565
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2160331913974665,
+      "learning_rate": 1.2543486568371736e-05,
+      "loss": 0.195,
+      "step": 8566
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.994132559894767,
+      "learning_rate": 1.2541893729790374e-05,
+      "loss": 0.1901,
+      "step": 8567
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7510679233061346,
+      "learning_rate": 1.2540300822259996e-05,
+      "loss": 0.1633,
+      "step": 8568
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0775744359597887,
+      "learning_rate": 1.253870784582382e-05,
+      "loss": 0.1956,
+      "step": 8569
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0562615754525093,
+      "learning_rate": 1.2537114800525047e-05,
+      "loss": 0.182,
+      "step": 8570
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1632556247766093,
+      "learning_rate": 1.2535521686406892e-05,
+      "loss": 0.1942,
+      "step": 8571
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1858005068205464,
+      "learning_rate": 1.253392850351257e-05,
+      "loss": 0.2178,
+      "step": 8572
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9958119742587856,
+      "learning_rate": 1.2532335251885295e-05,
+      "loss": 0.1953,
+      "step": 8573
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1614657700601225,
+      "learning_rate": 1.253074193156828e-05,
+      "loss": 0.1933,
+      "step": 8574
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7794605022464707,
+      "learning_rate": 1.252914854260475e-05,
+      "loss": 0.1924,
+      "step": 8575
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.52489821490543,
+      "learning_rate": 1.2527555085037919e-05,
+      "loss": 0.1658,
+      "step": 8576
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8420982181362046,
+      "learning_rate": 1.2525961558911018e-05,
+      "loss": 0.1941,
+      "step": 8577
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2260277438965748,
+      "learning_rate": 1.2524367964267264e-05,
+      "loss": 0.1826,
+      "step": 8578
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4095294544722203,
+      "learning_rate": 1.252277430114989e-05,
+      "loss": 0.2074,
+      "step": 8579
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.062831301945892,
+      "learning_rate": 1.2521180569602117e-05,
+      "loss": 0.1782,
+      "step": 8580
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1477816960609704,
+      "learning_rate": 1.2519586769667178e-05,
+      "loss": 0.1647,
+      "step": 8581
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9151787324161524,
+      "learning_rate": 1.2517992901388308e-05,
+      "loss": 0.2288,
+      "step": 8582
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8072999718676849,
+      "learning_rate": 1.2516398964808735e-05,
+      "loss": 0.2063,
+      "step": 8583
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9536553411143446,
+      "learning_rate": 1.2514804959971703e-05,
+      "loss": 0.1889,
+      "step": 8584
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.374498857012934,
+      "learning_rate": 1.251321088692044e-05,
+      "loss": 0.191,
+      "step": 8585
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7731476687954281,
+      "learning_rate": 1.2511616745698192e-05,
+      "loss": 0.1846,
+      "step": 8586
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0778842872188044,
+      "learning_rate": 1.2510022536348198e-05,
+      "loss": 0.1892,
+      "step": 8587
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9253449989932073,
+      "learning_rate": 1.2508428258913701e-05,
+      "loss": 0.187,
+      "step": 8588
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.793090613678402,
+      "learning_rate": 1.2506833913437946e-05,
+      "loss": 0.1968,
+      "step": 8589
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.019474888579202,
+      "learning_rate": 1.2505239499964179e-05,
+      "loss": 0.1851,
+      "step": 8590
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8630947482479875,
+      "learning_rate": 1.2503645018535649e-05,
+      "loss": 0.1793,
+      "step": 8591
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9521707788414885,
+      "learning_rate": 1.2502050469195609e-05,
+      "loss": 0.1933,
+      "step": 8592
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2404247773503827,
+      "learning_rate": 1.2500455851987306e-05,
+      "loss": 0.1703,
+      "step": 8593
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1472370304032953,
+      "learning_rate": 1.2498861166953995e-05,
+      "loss": 0.182,
+      "step": 8594
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9263589657793647,
+      "learning_rate": 1.2497266414138935e-05,
+      "loss": 0.2068,
+      "step": 8595
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0346555947683853,
+      "learning_rate": 1.2495671593585384e-05,
+      "loss": 0.182,
+      "step": 8596
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2960625843711235,
+      "learning_rate": 1.2494076705336599e-05,
+      "loss": 0.1918,
+      "step": 8597
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9689840253768952,
+      "learning_rate": 1.249248174943584e-05,
+      "loss": 0.2027,
+      "step": 8598
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8396890836603846,
+      "learning_rate": 1.2490886725926376e-05,
+      "loss": 0.1822,
+      "step": 8599
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7729532838421248,
+      "learning_rate": 1.248929163485147e-05,
+      "loss": 0.191,
+      "step": 8600
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.332863405617956,
+      "learning_rate": 1.2487696476254385e-05,
+      "loss": 0.1975,
+      "step": 8601
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8057277936675971,
+      "learning_rate": 1.2486101250178394e-05,
+      "loss": 0.1848,
+      "step": 8602
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8889873160686912,
+      "learning_rate": 1.2484505956666765e-05,
+      "loss": 0.1798,
+      "step": 8603
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7659955529390862,
+      "learning_rate": 1.2482910595762774e-05,
+      "loss": 0.1898,
+      "step": 8604
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3749674199455812,
+      "learning_rate": 1.2481315167509691e-05,
+      "loss": 0.2143,
+      "step": 8605
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.045463034779189,
+      "learning_rate": 1.2479719671950794e-05,
+      "loss": 0.1893,
+      "step": 8606
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.957490242738464,
+      "learning_rate": 1.247812410912936e-05,
+      "loss": 0.1875,
+      "step": 8607
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8685192921852094,
+      "learning_rate": 1.2476528479088672e-05,
+      "loss": 0.1966,
+      "step": 8608
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7153859251377935,
+      "learning_rate": 1.247493278187201e-05,
+      "loss": 0.166,
+      "step": 8609
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9957059549307132,
+      "learning_rate": 1.2473337017522653e-05,
+      "loss": 0.1893,
+      "step": 8610
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8473339655405635,
+      "learning_rate": 1.247174118608389e-05,
+      "loss": 0.2085,
+      "step": 8611
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9887067152293455,
+      "learning_rate": 1.247014528759901e-05,
+      "loss": 0.202,
+      "step": 8612
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7880322623898394,
+      "learning_rate": 1.2468549322111299e-05,
+      "loss": 0.1973,
+      "step": 8613
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2680939216214728,
+      "learning_rate": 1.2466953289664047e-05,
+      "loss": 0.2221,
+      "step": 8614
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7955595735621311,
+      "learning_rate": 1.246535719030055e-05,
+      "loss": 0.2079,
+      "step": 8615
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9892586690670219,
+      "learning_rate": 1.2463761024064093e-05,
+      "loss": 0.1772,
+      "step": 8616
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8184148596800643,
+      "learning_rate": 1.2462164790997986e-05,
+      "loss": 0.1804,
+      "step": 8617
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0393142650583775,
+      "learning_rate": 1.246056849114552e-05,
+      "loss": 0.1957,
+      "step": 8618
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7627059001636234,
+      "learning_rate": 1.2458972124549993e-05,
+      "loss": 0.1964,
+      "step": 8619
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9186241307525763,
+      "learning_rate": 1.2457375691254707e-05,
+      "loss": 0.1814,
+      "step": 8620
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9818593160010094,
+      "learning_rate": 1.2455779191302968e-05,
+      "loss": 0.2123,
+      "step": 8621
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9572114580336248,
+      "learning_rate": 1.2454182624738079e-05,
+      "loss": 0.2078,
+      "step": 8622
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9285814703568301,
+      "learning_rate": 1.2452585991603347e-05,
+      "loss": 0.2018,
+      "step": 8623
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8095708824494355,
+      "learning_rate": 1.2450989291942084e-05,
+      "loss": 0.1798,
+      "step": 8624
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8823515473747471,
+      "learning_rate": 1.2449392525797597e-05,
+      "loss": 0.1979,
+      "step": 8625
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8613819549391174,
+      "learning_rate": 1.24477956932132e-05,
+      "loss": 0.1853,
+      "step": 8626
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8301435562156122,
+      "learning_rate": 1.2446198794232206e-05,
+      "loss": 0.2087,
+      "step": 8627
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9817789102752862,
+      "learning_rate": 1.2444601828897932e-05,
+      "loss": 0.1995,
+      "step": 8628
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8351663415207485,
+      "learning_rate": 1.2443004797253692e-05,
+      "loss": 0.1902,
+      "step": 8629
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.132073956766278,
+      "learning_rate": 1.244140769934281e-05,
+      "loss": 0.2077,
+      "step": 8630
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7589643983073007,
+      "learning_rate": 1.243981053520861e-05,
+      "loss": 0.2104,
+      "step": 8631
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8034555252532624,
+      "learning_rate": 1.2438213304894408e-05,
+      "loss": 0.1927,
+      "step": 8632
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1157291738011732,
+      "learning_rate": 1.2436616008443534e-05,
+      "loss": 0.1892,
+      "step": 8633
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8209965863930293,
+      "learning_rate": 1.2435018645899312e-05,
+      "loss": 0.1695,
+      "step": 8634
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8722595948991724,
+      "learning_rate": 1.2433421217305072e-05,
+      "loss": 0.1867,
+      "step": 8635
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0124487539884681,
+      "learning_rate": 1.2431823722704147e-05,
+      "loss": 0.1798,
+      "step": 8636
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7754073159722372,
+      "learning_rate": 1.2430226162139863e-05,
+      "loss": 0.1993,
+      "step": 8637
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7968838934251788,
+      "learning_rate": 1.2428628535655557e-05,
+      "loss": 0.1781,
+      "step": 8638
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0162402030663464,
+      "learning_rate": 1.2427030843294562e-05,
+      "loss": 0.2094,
+      "step": 8639
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9488292199973177,
+      "learning_rate": 1.2425433085100224e-05,
+      "loss": 0.2202,
+      "step": 8640
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8235359537085621,
+      "learning_rate": 1.2423835261115875e-05,
+      "loss": 0.1932,
+      "step": 8641
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0966483166359258,
+      "learning_rate": 1.2422237371384857e-05,
+      "loss": 0.2221,
+      "step": 8642
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7067805471174222,
+      "learning_rate": 1.2420639415950512e-05,
+      "loss": 0.2121,
+      "step": 8643
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9066505172572226,
+      "learning_rate": 1.241904139485619e-05,
+      "loss": 0.1914,
+      "step": 8644
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8482878763826599,
+      "learning_rate": 1.2417443308145231e-05,
+      "loss": 0.2082,
+      "step": 8645
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.902057632643992,
+      "learning_rate": 1.2415845155860985e-05,
+      "loss": 0.2329,
+      "step": 8646
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8031248362414674,
+      "learning_rate": 1.2414246938046803e-05,
+      "loss": 0.1731,
+      "step": 8647
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.212999570557578,
+      "learning_rate": 1.2412648654746038e-05,
+      "loss": 0.1741,
+      "step": 8648
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.054499635574648,
+      "learning_rate": 1.2411050306002042e-05,
+      "loss": 0.1829,
+      "step": 8649
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9135542379212512,
+      "learning_rate": 1.240945189185817e-05,
+      "loss": 0.1867,
+      "step": 8650
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.405275321330997,
+      "learning_rate": 1.2407853412357775e-05,
+      "loss": 0.1889,
+      "step": 8651
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0048655495585312,
+      "learning_rate": 1.2406254867544225e-05,
+      "loss": 0.2119,
+      "step": 8652
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0534382352091345,
+      "learning_rate": 1.2404656257460875e-05,
+      "loss": 0.1997,
+      "step": 8653
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8011986524483066,
+      "learning_rate": 1.2403057582151088e-05,
+      "loss": 0.2032,
+      "step": 8654
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1453188395860927,
+      "learning_rate": 1.2401458841658227e-05,
+      "loss": 0.2022,
+      "step": 8655
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9429234081767248,
+      "learning_rate": 1.239986003602566e-05,
+      "loss": 0.1846,
+      "step": 8656
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2492399758575994,
+      "learning_rate": 1.2398261165296755e-05,
+      "loss": 0.186,
+      "step": 8657
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1247829996000132,
+      "learning_rate": 1.239666222951488e-05,
+      "loss": 0.1907,
+      "step": 8658
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9308794957701302,
+      "learning_rate": 1.2395063228723405e-05,
+      "loss": 0.1974,
+      "step": 8659
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8476358733321918,
+      "learning_rate": 1.2393464162965708e-05,
+      "loss": 0.2024,
+      "step": 8660
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8318120223195576,
+      "learning_rate": 1.239186503228516e-05,
+      "loss": 0.1754,
+      "step": 8661
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9189154775984179,
+      "learning_rate": 1.2390265836725136e-05,
+      "loss": 0.1938,
+      "step": 8662
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8411064285076155,
+      "learning_rate": 1.2388666576329016e-05,
+      "loss": 0.2287,
+      "step": 8663
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0606276167362687,
+      "learning_rate": 1.2387067251140178e-05,
+      "loss": 0.1818,
+      "step": 8664
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6362813671613476,
+      "learning_rate": 1.238546786120201e-05,
+      "loss": 0.2011,
+      "step": 8665
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.162047843891554,
+      "learning_rate": 1.238386840655789e-05,
+      "loss": 0.1743,
+      "step": 8666
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8695016899684824,
+      "learning_rate": 1.2382268887251207e-05,
+      "loss": 0.1849,
+      "step": 8667
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9264986670392068,
+      "learning_rate": 1.2380669303325346e-05,
+      "loss": 0.1989,
+      "step": 8668
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1435490303453728,
+      "learning_rate": 1.237906965482369e-05,
+      "loss": 0.2015,
+      "step": 8669
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7978001269514786,
+      "learning_rate": 1.2377469941789639e-05,
+      "loss": 0.1993,
+      "step": 8670
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1432618438727802,
+      "learning_rate": 1.2375870164266584e-05,
+      "loss": 0.1714,
+      "step": 8671
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9594113068655842,
+      "learning_rate": 1.2374270322297912e-05,
+      "loss": 0.1852,
+      "step": 8672
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7286130896672868,
+      "learning_rate": 1.2372670415927023e-05,
+      "loss": 0.1797,
+      "step": 8673
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.938152450920805,
+      "learning_rate": 1.2371070445197319e-05,
+      "loss": 0.2,
+      "step": 8674
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8560302080257822,
+      "learning_rate": 1.2369470410152195e-05,
+      "loss": 0.1802,
+      "step": 8675
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.923050463393855,
+      "learning_rate": 1.236787031083505e-05,
+      "loss": 0.2041,
+      "step": 8676
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9821030954763804,
+      "learning_rate": 1.2366270147289292e-05,
+      "loss": 0.1782,
+      "step": 8677
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9208630447896724,
+      "learning_rate": 1.2364669919558321e-05,
+      "loss": 0.1772,
+      "step": 8678
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8104379453709678,
+      "learning_rate": 1.2363069627685545e-05,
+      "loss": 0.1839,
+      "step": 8679
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8191806600272565,
+      "learning_rate": 1.2361469271714372e-05,
+      "loss": 0.1912,
+      "step": 8680
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9325590163572611,
+      "learning_rate": 1.235986885168821e-05,
+      "loss": 0.1951,
+      "step": 8681
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0513257766154047,
+      "learning_rate": 1.2358268367650472e-05,
+      "loss": 0.1946,
+      "step": 8682
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2444637911653882,
+      "learning_rate": 1.2356667819644575e-05,
+      "loss": 0.1932,
+      "step": 8683
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9103909492122905,
+      "learning_rate": 1.2355067207713927e-05,
+      "loss": 0.1942,
+      "step": 8684
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9457849858116818,
+      "learning_rate": 1.235346653190195e-05,
+      "loss": 0.2188,
+      "step": 8685
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9842814315695442,
+      "learning_rate": 1.2351865792252056e-05,
+      "loss": 0.1956,
+      "step": 8686
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8901122886363906,
+      "learning_rate": 1.2350264988807674e-05,
+      "loss": 0.1979,
+      "step": 8687
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7798350694565372,
+      "learning_rate": 1.2348664121612219e-05,
+      "loss": 0.2042,
+      "step": 8688
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9090186696452542,
+      "learning_rate": 1.234706319070912e-05,
+      "loss": 0.185,
+      "step": 8689
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7624892317314006,
+      "learning_rate": 1.2345462196141797e-05,
+      "loss": 0.1861,
+      "step": 8690
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3538646256449431,
+      "learning_rate": 1.2343861137953678e-05,
+      "loss": 0.1867,
+      "step": 8691
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9224124338770778,
+      "learning_rate": 1.2342260016188197e-05,
+      "loss": 0.1833,
+      "step": 8692
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8073268179435598,
+      "learning_rate": 1.234065883088878e-05,
+      "loss": 0.1896,
+      "step": 8693
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8844325994489721,
+      "learning_rate": 1.2339057582098859e-05,
+      "loss": 0.1878,
+      "step": 8694
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.91483931694918,
+      "learning_rate": 1.233745626986187e-05,
+      "loss": 0.2062,
+      "step": 8695
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1966454284281745,
+      "learning_rate": 1.2335854894221247e-05,
+      "loss": 0.1903,
+      "step": 8696
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0333757994680617,
+      "learning_rate": 1.2334253455220429e-05,
+      "loss": 0.1973,
+      "step": 8697
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8421908931074247,
+      "learning_rate": 1.2332651952902852e-05,
+      "loss": 0.1971,
+      "step": 8698
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9574744317950326,
+      "learning_rate": 1.2331050387311957e-05,
+      "loss": 0.2113,
+      "step": 8699
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9223770881917542,
+      "learning_rate": 1.2329448758491195e-05,
+      "loss": 0.1758,
+      "step": 8700
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9606924406565363,
+      "learning_rate": 1.2327847066484e-05,
+      "loss": 0.178,
+      "step": 8701
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2263409336211948,
+      "learning_rate": 1.2326245311333823e-05,
+      "loss": 0.1908,
+      "step": 8702
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.184016379817194,
+      "learning_rate": 1.232464349308411e-05,
+      "loss": 0.1893,
+      "step": 8703
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8384118727319358,
+      "learning_rate": 1.2323041611778309e-05,
+      "loss": 0.2002,
+      "step": 8704
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9282092440214317,
+      "learning_rate": 1.2321439667459876e-05,
+      "loss": 0.1757,
+      "step": 8705
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7959667284276463,
+      "learning_rate": 1.2319837660172258e-05,
+      "loss": 0.1889,
+      "step": 8706
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.750400860745341,
+      "learning_rate": 1.2318235589958916e-05,
+      "loss": 0.1915,
+      "step": 8707
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7735865678129724,
+      "learning_rate": 1.2316633456863299e-05,
+      "loss": 0.1712,
+      "step": 8708
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8724522023035849,
+      "learning_rate": 1.2315031260928872e-05,
+      "loss": 0.2235,
+      "step": 8709
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.915554410318466,
+      "learning_rate": 1.2313429002199088e-05,
+      "loss": 0.2089,
+      "step": 8710
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.78490783447735,
+      "learning_rate": 1.2311826680717416e-05,
+      "loss": 0.1776,
+      "step": 8711
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8774957388654196,
+      "learning_rate": 1.231022429652731e-05,
+      "loss": 0.1975,
+      "step": 8712
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.775234896911333,
+      "learning_rate": 1.2308621849672244e-05,
+      "loss": 0.1829,
+      "step": 8713
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9202591567715592,
+      "learning_rate": 1.2307019340195679e-05,
+      "loss": 0.2139,
+      "step": 8714
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9882359732294717,
+      "learning_rate": 1.2305416768141082e-05,
+      "loss": 0.2124,
+      "step": 8715
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.443374205415701,
+      "learning_rate": 1.2303814133551926e-05,
+      "loss": 0.1838,
+      "step": 8716
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8106286511053232,
+      "learning_rate": 1.230221143647168e-05,
+      "loss": 0.1945,
+      "step": 8717
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2442072877651187,
+      "learning_rate": 1.230060867694382e-05,
+      "loss": 0.1881,
+      "step": 8718
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9047866963976102,
+      "learning_rate": 1.229900585501182e-05,
+      "loss": 0.1872,
+      "step": 8719
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.072887532874953,
+      "learning_rate": 1.2297402970719157e-05,
+      "loss": 0.2219,
+      "step": 8720
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8320286552867004,
+      "learning_rate": 1.2295800024109306e-05,
+      "loss": 0.2063,
+      "step": 8721
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9128948927663452,
+      "learning_rate": 1.2294197015225751e-05,
+      "loss": 0.1819,
+      "step": 8722
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7655232053293657,
+      "learning_rate": 1.2292593944111972e-05,
+      "loss": 0.1741,
+      "step": 8723
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0419077231983973,
+      "learning_rate": 1.2290990810811456e-05,
+      "loss": 0.1952,
+      "step": 8724
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7758100286467379,
+      "learning_rate": 1.2289387615367684e-05,
+      "loss": 0.1884,
+      "step": 8725
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7020149836767687,
+      "learning_rate": 1.2287784357824138e-05,
+      "loss": 0.2069,
+      "step": 8726
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8239198097590726,
+      "learning_rate": 1.2286181038224316e-05,
+      "loss": 0.1886,
+      "step": 8727
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.928750257315147,
+      "learning_rate": 1.2284577656611706e-05,
+      "loss": 0.1989,
+      "step": 8728
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0055916699279317,
+      "learning_rate": 1.2282974213029797e-05,
+      "loss": 0.1863,
+      "step": 8729
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8478829767427825,
+      "learning_rate": 1.2281370707522083e-05,
+      "loss": 0.2072,
+      "step": 8730
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.765500143011328,
+      "learning_rate": 1.2279767140132059e-05,
+      "loss": 0.21,
+      "step": 8731
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8223953910269926,
+      "learning_rate": 1.2278163510903222e-05,
+      "loss": 0.2074,
+      "step": 8732
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0654584856574105,
+      "learning_rate": 1.2276559819879075e-05,
+      "loss": 0.2128,
+      "step": 8733
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9679552401237536,
+      "learning_rate": 1.227495606710311e-05,
+      "loss": 0.2025,
+      "step": 8734
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9519636449630081,
+      "learning_rate": 1.2273352252618834e-05,
+      "loss": 0.1731,
+      "step": 8735
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7244513931276524,
+      "learning_rate": 1.2271748376469753e-05,
+      "loss": 0.1691,
+      "step": 8736
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.253965355212945,
+      "learning_rate": 1.2270144438699365e-05,
+      "loss": 0.1948,
+      "step": 8737
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8691465810010405,
+      "learning_rate": 1.2268540439351183e-05,
+      "loss": 0.2224,
+      "step": 8738
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9405752587134205,
+      "learning_rate": 1.226693637846871e-05,
+      "loss": 0.1693,
+      "step": 8739
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9853075377180287,
+      "learning_rate": 1.2265332256095463e-05,
+      "loss": 0.1713,
+      "step": 8740
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.899562324586662,
+      "learning_rate": 1.2263728072274952e-05,
+      "loss": 0.1964,
+      "step": 8741
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8181308450571411,
+      "learning_rate": 1.2262123827050686e-05,
+      "loss": 0.1864,
+      "step": 8742
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8100402362729071,
+      "learning_rate": 1.2260519520466185e-05,
+      "loss": 0.2008,
+      "step": 8743
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7879823698755953,
+      "learning_rate": 1.2258915152564964e-05,
+      "loss": 0.1935,
+      "step": 8744
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8888512146131431,
+      "learning_rate": 1.2257310723390541e-05,
+      "loss": 0.1891,
+      "step": 8745
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0620541290396086,
+      "learning_rate": 1.2255706232986438e-05,
+      "loss": 0.184,
+      "step": 8746
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8531629248382205,
+      "learning_rate": 1.2254101681396177e-05,
+      "loss": 0.1811,
+      "step": 8747
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7726202393329789,
+      "learning_rate": 1.2252497068663281e-05,
+      "loss": 0.2278,
+      "step": 8748
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.094567252642259,
+      "learning_rate": 1.2250892394831272e-05,
+      "loss": 0.2069,
+      "step": 8749
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9823153214494921,
+      "learning_rate": 1.2249287659943682e-05,
+      "loss": 0.2044,
+      "step": 8750
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0498614567686715,
+      "learning_rate": 1.2247682864044037e-05,
+      "loss": 0.1827,
+      "step": 8751
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7545538618223903,
+      "learning_rate": 1.2246078007175866e-05,
+      "loss": 0.1968,
+      "step": 8752
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.061379383257261,
+      "learning_rate": 1.2244473089382702e-05,
+      "loss": 0.2101,
+      "step": 8753
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9009511671275275,
+      "learning_rate": 1.2242868110708079e-05,
+      "loss": 0.1945,
+      "step": 8754
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.790125825791761,
+      "learning_rate": 1.2241263071195535e-05,
+      "loss": 0.1818,
+      "step": 8755
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8315887432214394,
+      "learning_rate": 1.2239657970888598e-05,
+      "loss": 0.1918,
+      "step": 8756
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2449738158638226,
+      "learning_rate": 1.2238052809830816e-05,
+      "loss": 0.1872,
+      "step": 8757
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.867437863360933,
+      "learning_rate": 1.2236447588065723e-05,
+      "loss": 0.1945,
+      "step": 8758
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0749293306637857,
+      "learning_rate": 1.2234842305636865e-05,
+      "loss": 0.2122,
+      "step": 8759
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0562899061895583,
+      "learning_rate": 1.2233236962587782e-05,
+      "loss": 0.1708,
+      "step": 8760
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.034263512526097,
+      "learning_rate": 1.223163155896202e-05,
+      "loss": 0.195,
+      "step": 8761
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5855367010272412,
+      "learning_rate": 1.2230026094803127e-05,
+      "loss": 0.2104,
+      "step": 8762
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7252618781768035,
+      "learning_rate": 1.2228420570154649e-05,
+      "loss": 0.1752,
+      "step": 8763
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0131144932423213,
+      "learning_rate": 1.222681498506014e-05,
+      "loss": 0.2102,
+      "step": 8764
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8044676222681848,
+      "learning_rate": 1.2225209339563144e-05,
+      "loss": 0.167,
+      "step": 8765
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.867224693956808,
+      "learning_rate": 1.2223603633707224e-05,
+      "loss": 0.1935,
+      "step": 8766
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.339476780240116,
+      "learning_rate": 1.2221997867535926e-05,
+      "loss": 0.1887,
+      "step": 8767
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8640242341151145,
+      "learning_rate": 1.2220392041092813e-05,
+      "loss": 0.2016,
+      "step": 8768
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1395164976092698,
+      "learning_rate": 1.2218786154421439e-05,
+      "loss": 0.1965,
+      "step": 8769
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0955555345668755,
+      "learning_rate": 1.2217180207565365e-05,
+      "loss": 0.1814,
+      "step": 8770
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.080909186673661,
+      "learning_rate": 1.2215574200568155e-05,
+      "loss": 0.2092,
+      "step": 8771
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.868806952837568,
+      "learning_rate": 1.2213968133473366e-05,
+      "loss": 0.2031,
+      "step": 8772
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4526469492698382,
+      "learning_rate": 1.221236200632457e-05,
+      "loss": 0.1876,
+      "step": 8773
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.78784617514379,
+      "learning_rate": 1.2210755819165325e-05,
+      "loss": 0.1934,
+      "step": 8774
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1377878874747358,
+      "learning_rate": 1.2209149572039207e-05,
+      "loss": 0.2003,
+      "step": 8775
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1566898421232799,
+      "learning_rate": 1.220754326498978e-05,
+      "loss": 0.188,
+      "step": 8776
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4577368626043437,
+      "learning_rate": 1.2205936898060619e-05,
+      "loss": 0.2484,
+      "step": 8777
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.076111363988673,
+      "learning_rate": 1.2204330471295296e-05,
+      "loss": 0.1786,
+      "step": 8778
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2501558204678522,
+      "learning_rate": 1.2202723984737381e-05,
+      "loss": 0.1708,
+      "step": 8779
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9423711066705222,
+      "learning_rate": 1.2201117438430456e-05,
+      "loss": 0.1917,
+      "step": 8780
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.069649524013828,
+      "learning_rate": 1.2199510832418095e-05,
+      "loss": 0.1663,
+      "step": 8781
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.181212300392255,
+      "learning_rate": 1.2197904166743882e-05,
+      "loss": 0.1815,
+      "step": 8782
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7986369608289404,
+      "learning_rate": 1.2196297441451392e-05,
+      "loss": 0.179,
+      "step": 8783
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9746185125165827,
+      "learning_rate": 1.2194690656584209e-05,
+      "loss": 0.2062,
+      "step": 8784
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8134645721588003,
+      "learning_rate": 1.2193083812185919e-05,
+      "loss": 0.1861,
+      "step": 8785
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.052563619953177,
+      "learning_rate": 1.2191476908300107e-05,
+      "loss": 0.1774,
+      "step": 8786
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9659584698227893,
+      "learning_rate": 1.2189869944970356e-05,
+      "loss": 0.1747,
+      "step": 8787
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7230070985388103,
+      "learning_rate": 1.2188262922240263e-05,
+      "loss": 0.2043,
+      "step": 8788
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0231134064007024,
+      "learning_rate": 1.2186655840153413e-05,
+      "loss": 0.1926,
+      "step": 8789
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8286619944444407,
+      "learning_rate": 1.2185048698753403e-05,
+      "loss": 0.1684,
+      "step": 8790
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9708291745003088,
+      "learning_rate": 1.2183441498083821e-05,
+      "loss": 0.1807,
+      "step": 8791
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7781997192222662,
+      "learning_rate": 1.2181834238188264e-05,
+      "loss": 0.1757,
+      "step": 8792
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9225471816554169,
+      "learning_rate": 1.2180226919110332e-05,
+      "loss": 0.1986,
+      "step": 8793
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9767934398909636,
+      "learning_rate": 1.2178619540893621e-05,
+      "loss": 0.1835,
+      "step": 8794
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0241007879543205,
+      "learning_rate": 1.2177012103581733e-05,
+      "loss": 0.1993,
+      "step": 8795
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9077009583754211,
+      "learning_rate": 1.2175404607218267e-05,
+      "loss": 0.1816,
+      "step": 8796
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7274573376917196,
+      "learning_rate": 1.217379705184683e-05,
+      "loss": 0.1782,
+      "step": 8797
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.072798711913059,
+      "learning_rate": 1.2172189437511023e-05,
+      "loss": 0.2116,
+      "step": 8798
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2541852954716943,
+      "learning_rate": 1.2170581764254458e-05,
+      "loss": 0.1911,
+      "step": 8799
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2076914791500577,
+      "learning_rate": 1.2168974032120737e-05,
+      "loss": 0.2009,
+      "step": 8800
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9503860641385818,
+      "learning_rate": 1.2167366241153475e-05,
+      "loss": 0.1913,
+      "step": 8801
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0390296904907494,
+      "learning_rate": 1.2165758391396281e-05,
+      "loss": 0.1984,
+      "step": 8802
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3874804890400194,
+      "learning_rate": 1.2164150482892768e-05,
+      "loss": 0.2067,
+      "step": 8803
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6212319706782996,
+      "learning_rate": 1.2162542515686551e-05,
+      "loss": 0.1909,
+      "step": 8804
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9567611095220859,
+      "learning_rate": 1.2160934489821244e-05,
+      "loss": 0.1839,
+      "step": 8805
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2799465065164584,
+      "learning_rate": 1.2159326405340468e-05,
+      "loss": 0.1921,
+      "step": 8806
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.086003363283199,
+      "learning_rate": 1.2157718262287841e-05,
+      "loss": 0.1784,
+      "step": 8807
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7401550610848462,
+      "learning_rate": 1.2156110060706986e-05,
+      "loss": 0.1891,
+      "step": 8808
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8915999284651739,
+      "learning_rate": 1.215450180064152e-05,
+      "loss": 0.1977,
+      "step": 8809
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0195758291399881,
+      "learning_rate": 1.2152893482135075e-05,
+      "loss": 0.1806,
+      "step": 8810
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1749497925696335,
+      "learning_rate": 1.2151285105231273e-05,
+      "loss": 0.2189,
+      "step": 8811
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8225497459206069,
+      "learning_rate": 1.214967666997374e-05,
+      "loss": 0.1824,
+      "step": 8812
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.214799801660971,
+      "learning_rate": 1.2148068176406104e-05,
+      "loss": 0.178,
+      "step": 8813
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9197690708431241,
+      "learning_rate": 1.2146459624571998e-05,
+      "loss": 0.1859,
+      "step": 8814
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9091858741271089,
+      "learning_rate": 1.2144851014515055e-05,
+      "loss": 0.1755,
+      "step": 8815
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.086043007279395,
+      "learning_rate": 1.2143242346278908e-05,
+      "loss": 0.1923,
+      "step": 8816
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3372441168269635,
+      "learning_rate": 1.214163361990719e-05,
+      "loss": 0.1931,
+      "step": 8817
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.963781956668138,
+      "learning_rate": 1.2140024835443537e-05,
+      "loss": 0.2051,
+      "step": 8818
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2980605044900115,
+      "learning_rate": 1.213841599293159e-05,
+      "loss": 0.1746,
+      "step": 8819
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8386179927027012,
+      "learning_rate": 1.2136807092414992e-05,
+      "loss": 0.1855,
+      "step": 8820
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8201498327266299,
+      "learning_rate": 1.2135198133937381e-05,
+      "loss": 0.1677,
+      "step": 8821
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9651597614348006,
+      "learning_rate": 1.2133589117542395e-05,
+      "loss": 0.1907,
+      "step": 8822
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9814233612261183,
+      "learning_rate": 1.2131980043273685e-05,
+      "loss": 0.2056,
+      "step": 8823
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8653900828234407,
+      "learning_rate": 1.2130370911174898e-05,
+      "loss": 0.1729,
+      "step": 8824
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3058003760102206,
+      "learning_rate": 1.212876172128968e-05,
+      "loss": 0.1978,
+      "step": 8825
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0536411372349836,
+      "learning_rate": 1.2127152473661678e-05,
+      "loss": 0.1949,
+      "step": 8826
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9398760268893137,
+      "learning_rate": 1.2125543168334546e-05,
+      "loss": 0.1978,
+      "step": 8827
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9603591809895711,
+      "learning_rate": 1.2123933805351934e-05,
+      "loss": 0.2136,
+      "step": 8828
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8674456935843604,
+      "learning_rate": 1.21223243847575e-05,
+      "loss": 0.1757,
+      "step": 8829
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.959780207158066,
+      "learning_rate": 1.2120714906594897e-05,
+      "loss": 0.2059,
+      "step": 8830
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.250047070583227,
+      "learning_rate": 1.211910537090778e-05,
+      "loss": 0.1826,
+      "step": 8831
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8414534827990336,
+      "learning_rate": 1.2117495777739815e-05,
+      "loss": 0.174,
+      "step": 8832
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.105316291874254,
+      "learning_rate": 1.2115886127134653e-05,
+      "loss": 0.2109,
+      "step": 8833
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8495791290530903,
+      "learning_rate": 1.2114276419135964e-05,
+      "loss": 0.1905,
+      "step": 8834
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4008403268681422,
+      "learning_rate": 1.2112666653787404e-05,
+      "loss": 0.2,
+      "step": 8835
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0376823697906254,
+      "learning_rate": 1.2111056831132641e-05,
+      "loss": 0.1758,
+      "step": 8836
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9164332739485018,
+      "learning_rate": 1.2109446951215347e-05,
+      "loss": 0.196,
+      "step": 8837
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0077455559186472,
+      "learning_rate": 1.2107837014079182e-05,
+      "loss": 0.2039,
+      "step": 8838
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8600414769402724,
+      "learning_rate": 1.210622701976782e-05,
+      "loss": 0.1759,
+      "step": 8839
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.83784665366274,
+      "learning_rate": 1.2104616968324928e-05,
+      "loss": 0.1779,
+      "step": 8840
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1928432869020684,
+      "learning_rate": 1.2103006859794184e-05,
+      "loss": 0.193,
+      "step": 8841
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9301657792242488,
+      "learning_rate": 1.2101396694219262e-05,
+      "loss": 0.176,
+      "step": 8842
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1520056763985131,
+      "learning_rate": 1.2099786471643834e-05,
+      "loss": 0.1946,
+      "step": 8843
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1447231599498406,
+      "learning_rate": 1.2098176192111578e-05,
+      "loss": 0.2055,
+      "step": 8844
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8944520715875532,
+      "learning_rate": 1.2096565855666178e-05,
+      "loss": 0.2033,
+      "step": 8845
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1372848444010801,
+      "learning_rate": 1.209495546235131e-05,
+      "loss": 0.2061,
+      "step": 8846
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7647781497580564,
+      "learning_rate": 1.2093345012210656e-05,
+      "loss": 0.1932,
+      "step": 8847
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8912513425073467,
+      "learning_rate": 1.20917345052879e-05,
+      "loss": 0.1796,
+      "step": 8848
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5089662709207687,
+      "learning_rate": 1.2090123941626726e-05,
+      "loss": 0.1784,
+      "step": 8849
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0933341421872094,
+      "learning_rate": 1.2088513321270823e-05,
+      "loss": 0.1995,
+      "step": 8850
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9191939310469205,
+      "learning_rate": 1.2086902644263878e-05,
+      "loss": 0.1889,
+      "step": 8851
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8287536485776805,
+      "learning_rate": 1.2085291910649585e-05,
+      "loss": 0.1873,
+      "step": 8852
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9271405890498322,
+      "learning_rate": 1.2083681120471626e-05,
+      "loss": 0.1982,
+      "step": 8853
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9672085990082487,
+      "learning_rate": 1.20820702737737e-05,
+      "loss": 0.1924,
+      "step": 8854
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.050213589768552,
+      "learning_rate": 1.2080459370599502e-05,
+      "loss": 0.184,
+      "step": 8855
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9957672183719678,
+      "learning_rate": 1.2078848410992726e-05,
+      "loss": 0.1982,
+      "step": 8856
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7800792332036594,
+      "learning_rate": 1.2077237394997065e-05,
+      "loss": 0.1702,
+      "step": 8857
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0934036596223062,
+      "learning_rate": 1.2075626322656227e-05,
+      "loss": 0.1914,
+      "step": 8858
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8329459545117248,
+      "learning_rate": 1.2074015194013906e-05,
+      "loss": 0.1869,
+      "step": 8859
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8139405261848299,
+      "learning_rate": 1.2072404009113808e-05,
+      "loss": 0.2056,
+      "step": 8860
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9862417615264362,
+      "learning_rate": 1.2070792767999633e-05,
+      "loss": 0.1837,
+      "step": 8861
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0204044344285164,
+      "learning_rate": 1.2069181470715083e-05,
+      "loss": 0.2264,
+      "step": 8862
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1435750249832166,
+      "learning_rate": 1.2067570117303872e-05,
+      "loss": 0.198,
+      "step": 8863
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9935613757091443,
+      "learning_rate": 1.2065958707809705e-05,
+      "loss": 0.2109,
+      "step": 8864
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0114797607164516,
+      "learning_rate": 1.2064347242276293e-05,
+      "loss": 0.192,
+      "step": 8865
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.118130182359597,
+      "learning_rate": 1.2062735720747343e-05,
+      "loss": 0.1888,
+      "step": 8866
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3928760503882156,
+      "learning_rate": 1.2061124143266571e-05,
+      "loss": 0.1989,
+      "step": 8867
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.020045704762623,
+      "learning_rate": 1.2059512509877691e-05,
+      "loss": 0.1943,
+      "step": 8868
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7282918073861949,
+      "learning_rate": 1.205790082062442e-05,
+      "loss": 0.1686,
+      "step": 8869
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8817405987482647,
+      "learning_rate": 1.2056289075550467e-05,
+      "loss": 0.1743,
+      "step": 8870
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2194815943547086,
+      "learning_rate": 1.205467727469956e-05,
+      "loss": 0.1815,
+      "step": 8871
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1515293942216966,
+      "learning_rate": 1.2053065418115418e-05,
+      "loss": 0.1901,
+      "step": 8872
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8144674738067087,
+      "learning_rate": 1.2051453505841757e-05,
+      "loss": 0.194,
+      "step": 8873
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8586989430386209,
+      "learning_rate": 1.2049841537922307e-05,
+      "loss": 0.196,
+      "step": 8874
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7018719551715215,
+      "learning_rate": 1.2048229514400785e-05,
+      "loss": 0.1723,
+      "step": 8875
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0214569381837684,
+      "learning_rate": 1.2046617435320924e-05,
+      "loss": 0.1901,
+      "step": 8876
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.005899395849749,
+      "learning_rate": 1.2045005300726452e-05,
+      "loss": 0.1923,
+      "step": 8877
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5862559489013401,
+      "learning_rate": 1.2043393110661092e-05,
+      "loss": 0.1911,
+      "step": 8878
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1185356170292566,
+      "learning_rate": 1.2041780865168577e-05,
+      "loss": 0.2061,
+      "step": 8879
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0316089428982487,
+      "learning_rate": 1.2040168564292644e-05,
+      "loss": 0.19,
+      "step": 8880
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6316760066241547,
+      "learning_rate": 1.2038556208077026e-05,
+      "loss": 0.1756,
+      "step": 8881
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9262215986725295,
+      "learning_rate": 1.2036943796565453e-05,
+      "loss": 0.1677,
+      "step": 8882
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8161375418600838,
+      "learning_rate": 1.2035331329801663e-05,
+      "loss": 0.2222,
+      "step": 8883
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9862382320843694,
+      "learning_rate": 1.2033718807829395e-05,
+      "loss": 0.2013,
+      "step": 8884
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.924590732113376,
+      "learning_rate": 1.2032106230692394e-05,
+      "loss": 0.1705,
+      "step": 8885
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7549429941943393,
+      "learning_rate": 1.2030493598434392e-05,
+      "loss": 0.1843,
+      "step": 8886
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8770179406545869,
+      "learning_rate": 1.2028880911099141e-05,
+      "loss": 0.199,
+      "step": 8887
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0535531729638505,
+      "learning_rate": 1.2027268168730378e-05,
+      "loss": 0.1917,
+      "step": 8888
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9816682275340497,
+      "learning_rate": 1.202565537137185e-05,
+      "loss": 0.1873,
+      "step": 8889
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.111591525187513,
+      "learning_rate": 1.202404251906731e-05,
+      "loss": 0.2159,
+      "step": 8890
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9930935676272202,
+      "learning_rate": 1.20224296118605e-05,
+      "loss": 0.1959,
+      "step": 8891
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0744553729919295,
+      "learning_rate": 1.202081664979517e-05,
+      "loss": 0.1829,
+      "step": 8892
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0903638199210839,
+      "learning_rate": 1.2019203632915078e-05,
+      "loss": 0.193,
+      "step": 8893
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8433993464195626,
+      "learning_rate": 1.2017590561263973e-05,
+      "loss": 0.1989,
+      "step": 8894
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9218661790570869,
+      "learning_rate": 1.2015977434885608e-05,
+      "loss": 0.1988,
+      "step": 8895
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.086711989423081,
+      "learning_rate": 1.2014364253823742e-05,
+      "loss": 0.1997,
+      "step": 8896
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9552204008080996,
+      "learning_rate": 1.2012751018122132e-05,
+      "loss": 0.1918,
+      "step": 8897
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1356827714958893,
+      "learning_rate": 1.2011137727824536e-05,
+      "loss": 0.1838,
+      "step": 8898
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1033408198461252,
+      "learning_rate": 1.2009524382974717e-05,
+      "loss": 0.2092,
+      "step": 8899
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.026359269650774,
+      "learning_rate": 1.2007910983616435e-05,
+      "loss": 0.1791,
+      "step": 8900
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7595638197151642,
+      "learning_rate": 1.2006297529793456e-05,
+      "loss": 0.185,
+      "step": 8901
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8070567756184621,
+      "learning_rate": 1.200468402154954e-05,
+      "loss": 0.1862,
+      "step": 8902
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9054734943926448,
+      "learning_rate": 1.2003070458928458e-05,
+      "loss": 0.215,
+      "step": 8903
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.4442105832569023,
+      "learning_rate": 1.200145684197398e-05,
+      "loss": 0.1978,
+      "step": 8904
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2466883918482339,
+      "learning_rate": 1.1999843170729866e-05,
+      "loss": 0.1729,
+      "step": 8905
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.941852164338417,
+      "learning_rate": 1.1998229445239898e-05,
+      "loss": 0.1966,
+      "step": 8906
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1718913462872127,
+      "learning_rate": 1.1996615665547841e-05,
+      "loss": 0.1927,
+      "step": 8907
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.994145976940982,
+      "learning_rate": 1.1995001831697472e-05,
+      "loss": 0.2004,
+      "step": 8908
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8267161858948318,
+      "learning_rate": 1.1993387943732567e-05,
+      "loss": 0.2053,
+      "step": 8909
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7850285872276042,
+      "learning_rate": 1.1991774001696896e-05,
+      "loss": 0.1807,
+      "step": 8910
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9629868160271702,
+      "learning_rate": 1.1990160005634248e-05,
+      "loss": 0.1959,
+      "step": 8911
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0503190559300086,
+      "learning_rate": 1.1988545955588395e-05,
+      "loss": 0.1844,
+      "step": 8912
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9957325838836562,
+      "learning_rate": 1.1986931851603122e-05,
+      "loss": 0.2011,
+      "step": 8913
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0639985315351588,
+      "learning_rate": 1.1985317693722212e-05,
+      "loss": 0.1996,
+      "step": 8914
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9629188523199401,
+      "learning_rate": 1.1983703481989443e-05,
+      "loss": 0.2051,
+      "step": 8915
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9569474788455568,
+      "learning_rate": 1.1982089216448607e-05,
+      "loss": 0.2102,
+      "step": 8916
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0215416358534117,
+      "learning_rate": 1.198047489714349e-05,
+      "loss": 0.2206,
+      "step": 8917
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1053432675497292,
+      "learning_rate": 1.197886052411788e-05,
+      "loss": 0.2172,
+      "step": 8918
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2538967899336353,
+      "learning_rate": 1.1977246097415565e-05,
+      "loss": 0.1804,
+      "step": 8919
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7904544707992178,
+      "learning_rate": 1.1975631617080339e-05,
+      "loss": 0.2404,
+      "step": 8920
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.010512958427013,
+      "learning_rate": 1.1974017083155993e-05,
+      "loss": 0.2069,
+      "step": 8921
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9187475659348551,
+      "learning_rate": 1.1972402495686323e-05,
+      "loss": 0.195,
+      "step": 8922
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1264834567954618,
+      "learning_rate": 1.1970787854715123e-05,
+      "loss": 0.1797,
+      "step": 8923
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.902873139466147,
+      "learning_rate": 1.1969173160286191e-05,
+      "loss": 0.1672,
+      "step": 8924
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8147847023995554,
+      "learning_rate": 1.1967558412443328e-05,
+      "loss": 0.2024,
+      "step": 8925
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.426225608926197,
+      "learning_rate": 1.1965943611230331e-05,
+      "loss": 0.1954,
+      "step": 8926
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9450555610864556,
+      "learning_rate": 1.1964328756691e-05,
+      "loss": 0.1639,
+      "step": 8927
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.239627680378109,
+      "learning_rate": 1.196271384886914e-05,
+      "loss": 0.1813,
+      "step": 8928
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.110179894405812,
+      "learning_rate": 1.196109888780856e-05,
+      "loss": 0.1652,
+      "step": 8929
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8597255073498342,
+      "learning_rate": 1.1959483873553059e-05,
+      "loss": 0.1854,
+      "step": 8930
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5145937177785822,
+      "learning_rate": 1.1957868806146449e-05,
+      "loss": 0.2137,
+      "step": 8931
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1041447036829297,
+      "learning_rate": 1.1956253685632534e-05,
+      "loss": 0.1661,
+      "step": 8932
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9525616525645625,
+      "learning_rate": 1.1954638512055131e-05,
+      "loss": 0.2112,
+      "step": 8933
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3595966325094786,
+      "learning_rate": 1.1953023285458047e-05,
+      "loss": 0.1958,
+      "step": 8934
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9975591742736065,
+      "learning_rate": 1.1951408005885098e-05,
+      "loss": 0.183,
+      "step": 8935
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.773323065930237,
+      "learning_rate": 1.1949792673380094e-05,
+      "loss": 0.1689,
+      "step": 8936
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0978246199548096,
+      "learning_rate": 1.194817728798685e-05,
+      "loss": 0.2031,
+      "step": 8937
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9025386810836243,
+      "learning_rate": 1.1946561849749192e-05,
+      "loss": 0.1739,
+      "step": 8938
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0924668110309403,
+      "learning_rate": 1.1944946358710936e-05,
+      "loss": 0.1892,
+      "step": 8939
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8475628728718165,
+      "learning_rate": 1.1943330814915897e-05,
+      "loss": 0.1962,
+      "step": 8940
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.144332901565883,
+      "learning_rate": 1.1941715218407898e-05,
+      "loss": 0.1789,
+      "step": 8941
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9641545428843032,
+      "learning_rate": 1.1940099569230767e-05,
+      "loss": 0.2029,
+      "step": 8942
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0694630086624335,
+      "learning_rate": 1.1938483867428326e-05,
+      "loss": 0.1833,
+      "step": 8943
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6671110565068417,
+      "learning_rate": 1.19368681130444e-05,
+      "loss": 0.1973,
+      "step": 8944
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0086089261741662,
+      "learning_rate": 1.1935252306122812e-05,
+      "loss": 0.1949,
+      "step": 8945
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8401339480562876,
+      "learning_rate": 1.1933636446707401e-05,
+      "loss": 0.159,
+      "step": 8946
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1733262203357395,
+      "learning_rate": 1.1932020534841992e-05,
+      "loss": 0.1925,
+      "step": 8947
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2062640371557118,
+      "learning_rate": 1.1930404570570417e-05,
+      "loss": 0.2117,
+      "step": 8948
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.58050795448441,
+      "learning_rate": 1.1928788553936507e-05,
+      "loss": 0.1997,
+      "step": 8949
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2136141398608247,
+      "learning_rate": 1.19271724849841e-05,
+      "loss": 0.2107,
+      "step": 8950
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5287755275959543,
+      "learning_rate": 1.192555636375703e-05,
+      "loss": 0.1869,
+      "step": 8951
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7695097792624204,
+      "learning_rate": 1.1923940190299135e-05,
+      "loss": 0.1838,
+      "step": 8952
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.046415308205685,
+      "learning_rate": 1.1922323964654254e-05,
+      "loss": 0.1898,
+      "step": 8953
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9872118514792896,
+      "learning_rate": 1.1920707686866227e-05,
+      "loss": 0.1898,
+      "step": 8954
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9799211018153227,
+      "learning_rate": 1.1919091356978894e-05,
+      "loss": 0.1642,
+      "step": 8955
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.155722364648013,
+      "learning_rate": 1.19174749750361e-05,
+      "loss": 0.1772,
+      "step": 8956
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2290325356839902,
+      "learning_rate": 1.1915858541081693e-05,
+      "loss": 0.1663,
+      "step": 8957
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.30211196478441,
+      "learning_rate": 1.191424205515951e-05,
+      "loss": 0.2073,
+      "step": 8958
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0336619162587886,
+      "learning_rate": 1.1912625517313406e-05,
+      "loss": 0.2175,
+      "step": 8959
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.015136790947612,
+      "learning_rate": 1.1911008927587224e-05,
+      "loss": 0.1947,
+      "step": 8960
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0525402971892104,
+      "learning_rate": 1.190939228602482e-05,
+      "loss": 0.2008,
+      "step": 8961
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0309507319611686,
+      "learning_rate": 1.190777559267004e-05,
+      "loss": 0.1712,
+      "step": 8962
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.042308202194219,
+      "learning_rate": 1.190615884756674e-05,
+      "loss": 0.2174,
+      "step": 8963
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2486820476818925,
+      "learning_rate": 1.1904542050758774e-05,
+      "loss": 0.2019,
+      "step": 8964
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9681283258309303,
+      "learning_rate": 1.1902925202289997e-05,
+      "loss": 0.1748,
+      "step": 8965
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8824864392862095,
+      "learning_rate": 1.1901308302204267e-05,
+      "loss": 0.1827,
+      "step": 8966
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0792711159266326,
+      "learning_rate": 1.189969135054544e-05,
+      "loss": 0.1817,
+      "step": 8967
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9004669137625958,
+      "learning_rate": 1.1898074347357377e-05,
+      "loss": 0.1905,
+      "step": 8968
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8946191807568328,
+      "learning_rate": 1.1896457292683945e-05,
+      "loss": 0.2004,
+      "step": 8969
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9348331603869736,
+      "learning_rate": 1.1894840186569e-05,
+      "loss": 0.1854,
+      "step": 8970
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7410801074117348,
+      "learning_rate": 1.1893223029056406e-05,
+      "loss": 0.1813,
+      "step": 8971
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1059641875465824,
+      "learning_rate": 1.1891605820190031e-05,
+      "loss": 0.1687,
+      "step": 8972
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9228413836557761,
+      "learning_rate": 1.1889988560013741e-05,
+      "loss": 0.1917,
+      "step": 8973
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8623997445511834,
+      "learning_rate": 1.1888371248571409e-05,
+      "loss": 0.1745,
+      "step": 8974
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3696551052147947,
+      "learning_rate": 1.1886753885906895e-05,
+      "loss": 0.2035,
+      "step": 8975
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1247215692096408,
+      "learning_rate": 1.1885136472064076e-05,
+      "loss": 0.2072,
+      "step": 8976
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1875189683505865,
+      "learning_rate": 1.1883519007086824e-05,
+      "loss": 0.217,
+      "step": 8977
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2625623423472383,
+      "learning_rate": 1.1881901491019014e-05,
+      "loss": 0.1774,
+      "step": 8978
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9240639516641106,
+      "learning_rate": 1.1880283923904518e-05,
+      "loss": 0.192,
+      "step": 8979
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8776495632201888,
+      "learning_rate": 1.1878666305787214e-05,
+      "loss": 0.1982,
+      "step": 8980
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9658463095638202,
+      "learning_rate": 1.1877048636710981e-05,
+      "loss": 0.1716,
+      "step": 8981
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2894250472685753,
+      "learning_rate": 1.1875430916719698e-05,
+      "loss": 0.1957,
+      "step": 8982
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0529602099603186,
+      "learning_rate": 1.187381314585725e-05,
+      "loss": 0.18,
+      "step": 8983
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.159065460851313,
+      "learning_rate": 1.1872195324167508e-05,
+      "loss": 0.2057,
+      "step": 8984
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0456114107643657,
+      "learning_rate": 1.1870577451694363e-05,
+      "loss": 0.1668,
+      "step": 8985
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8648956682920848,
+      "learning_rate": 1.18689595284817e-05,
+      "loss": 0.1842,
+      "step": 8986
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0062545746069842,
+      "learning_rate": 1.1867341554573405e-05,
+      "loss": 0.1752,
+      "step": 8987
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9052168926450945,
+      "learning_rate": 1.1865723530013367e-05,
+      "loss": 0.178,
+      "step": 8988
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1849884410504337,
+      "learning_rate": 1.1864105454845467e-05,
+      "loss": 0.1898,
+      "step": 8989
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1768091963894038,
+      "learning_rate": 1.1862487329113606e-05,
+      "loss": 0.2171,
+      "step": 8990
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9149338679136048,
+      "learning_rate": 1.186086915286167e-05,
+      "loss": 0.191,
+      "step": 8991
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8723532019990747,
+      "learning_rate": 1.1859250926133554e-05,
+      "loss": 0.1828,
+      "step": 8992
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4707240199120826,
+      "learning_rate": 1.185763264897315e-05,
+      "loss": 0.1637,
+      "step": 8993
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7403666829101415,
+      "learning_rate": 1.1856014321424356e-05,
+      "loss": 0.1928,
+      "step": 8994
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9394086856654486,
+      "learning_rate": 1.185439594353107e-05,
+      "loss": 0.1723,
+      "step": 8995
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7670705541749999,
+      "learning_rate": 1.1852777515337186e-05,
+      "loss": 0.1841,
+      "step": 8996
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1035336169056782,
+      "learning_rate": 1.185115903688661e-05,
+      "loss": 0.1886,
+      "step": 8997
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2146454611707889,
+      "learning_rate": 1.1849540508223238e-05,
+      "loss": 0.1969,
+      "step": 8998
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0393209712450067,
+      "learning_rate": 1.1847921929390977e-05,
+      "loss": 0.1946,
+      "step": 8999
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.008870498761671,
+      "learning_rate": 1.184630330043373e-05,
+      "loss": 0.1847,
+      "step": 9000
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9349115695325397,
+      "learning_rate": 1.1844684621395401e-05,
+      "loss": 0.2054,
+      "step": 9001
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.924911367937286,
+      "learning_rate": 1.1843065892319895e-05,
+      "loss": 0.1891,
+      "step": 9002
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1563745043827904,
+      "learning_rate": 1.1841447113251126e-05,
+      "loss": 0.2084,
+      "step": 9003
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8965980604277103,
+      "learning_rate": 1.1839828284233e-05,
+      "loss": 0.1612,
+      "step": 9004
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9034527880843174,
+      "learning_rate": 1.1838209405309427e-05,
+      "loss": 0.1886,
+      "step": 9005
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0077625098069756,
+      "learning_rate": 1.183659047652432e-05,
+      "loss": 0.1865,
+      "step": 9006
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9196710553785808,
+      "learning_rate": 1.1834971497921591e-05,
+      "loss": 0.1953,
+      "step": 9007
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9536834374306841,
+      "learning_rate": 1.1833352469545158e-05,
+      "loss": 0.2062,
+      "step": 9008
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2212574071626323,
+      "learning_rate": 1.1831733391438937e-05,
+      "loss": 0.1907,
+      "step": 9009
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5537835504342088,
+      "learning_rate": 1.1830114263646844e-05,
+      "loss": 0.2087,
+      "step": 9010
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8431492083254475,
+      "learning_rate": 1.1828495086212794e-05,
+      "loss": 0.1801,
+      "step": 9011
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0463466456718005,
+      "learning_rate": 1.1826875859180718e-05,
+      "loss": 0.186,
+      "step": 9012
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1162356038874017,
+      "learning_rate": 1.1825256582594526e-05,
+      "loss": 0.207,
+      "step": 9013
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9375444806124058,
+      "learning_rate": 1.1823637256498149e-05,
+      "loss": 0.2037,
+      "step": 9014
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.180452055645475,
+      "learning_rate": 1.1822017880935507e-05,
+      "loss": 0.1722,
+      "step": 9015
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8274008918640854,
+      "learning_rate": 1.1820398455950526e-05,
+      "loss": 0.1889,
+      "step": 9016
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7722948608780377,
+      "learning_rate": 1.1818778981587135e-05,
+      "loss": 0.1851,
+      "step": 9017
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8623477671210378,
+      "learning_rate": 1.1817159457889264e-05,
+      "loss": 0.1758,
+      "step": 9018
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7589775449226215,
+      "learning_rate": 1.1815539884900837e-05,
+      "loss": 0.1954,
+      "step": 9019
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.054810457169351,
+      "learning_rate": 1.1813920262665788e-05,
+      "loss": 0.1878,
+      "step": 9020
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.293935543595433,
+      "learning_rate": 1.1812300591228052e-05,
+      "loss": 0.212,
+      "step": 9021
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8558181092026415,
+      "learning_rate": 1.1810680870631558e-05,
+      "loss": 0.176,
+      "step": 9022
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7573874055928003,
+      "learning_rate": 1.1809061100920245e-05,
+      "loss": 0.1755,
+      "step": 9023
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1342319197752257,
+      "learning_rate": 1.1807441282138045e-05,
+      "loss": 0.1882,
+      "step": 9024
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.052170888869555,
+      "learning_rate": 1.1805821414328897e-05,
+      "loss": 0.1855,
+      "step": 9025
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9040062958108732,
+      "learning_rate": 1.1804201497536746e-05,
+      "loss": 0.1781,
+      "step": 9026
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0001531848845238,
+      "learning_rate": 1.1802581531805525e-05,
+      "loss": 0.1835,
+      "step": 9027
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9702189858557619,
+      "learning_rate": 1.1800961517179177e-05,
+      "loss": 0.1925,
+      "step": 9028
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9391996907219045,
+      "learning_rate": 1.179934145370165e-05,
+      "loss": 0.1968,
+      "step": 9029
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.169550256959143,
+      "learning_rate": 1.1797721341416882e-05,
+      "loss": 0.2095,
+      "step": 9030
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4986948128088586,
+      "learning_rate": 1.179610118036882e-05,
+      "loss": 0.1983,
+      "step": 9031
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2257047082649992,
+      "learning_rate": 1.1794480970601413e-05,
+      "loss": 0.2061,
+      "step": 9032
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1231876572801334,
+      "learning_rate": 1.1792860712158608e-05,
+      "loss": 0.1956,
+      "step": 9033
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3543258643514862,
+      "learning_rate": 1.1791240405084355e-05,
+      "loss": 0.1994,
+      "step": 9034
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8771805946884635,
+      "learning_rate": 1.1789620049422603e-05,
+      "loss": 0.1673,
+      "step": 9035
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.028897040551911,
+      "learning_rate": 1.1787999645217309e-05,
+      "loss": 0.2148,
+      "step": 9036
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0308527874745395,
+      "learning_rate": 1.178637919251242e-05,
+      "loss": 0.2075,
+      "step": 9037
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2276060915073248,
+      "learning_rate": 1.1784758691351898e-05,
+      "loss": 0.1925,
+      "step": 9038
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2425534898000836,
+      "learning_rate": 1.1783138141779695e-05,
+      "loss": 0.1499,
+      "step": 9039
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9302029211692885,
+      "learning_rate": 1.178151754383977e-05,
+      "loss": 0.1992,
+      "step": 9040
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0435356139538756,
+      "learning_rate": 1.177989689757608e-05,
+      "loss": 0.1983,
+      "step": 9041
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9754457914366035,
+      "learning_rate": 1.1778276203032584e-05,
+      "loss": 0.1784,
+      "step": 9042
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0053469624790905,
+      "learning_rate": 1.1776655460253248e-05,
+      "loss": 0.1921,
+      "step": 9043
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8778961911792674,
+      "learning_rate": 1.1775034669282034e-05,
+      "loss": 0.1955,
+      "step": 9044
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1203395252317987,
+      "learning_rate": 1.1773413830162902e-05,
+      "loss": 0.1981,
+      "step": 9045
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0830248882097275,
+      "learning_rate": 1.177179294293982e-05,
+      "loss": 0.1805,
+      "step": 9046
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9531798772402397,
+      "learning_rate": 1.1770172007656756e-05,
+      "loss": 0.195,
+      "step": 9047
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3430591008501616,
+      "learning_rate": 1.1768551024357673e-05,
+      "loss": 0.2009,
+      "step": 9048
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9824402645449075,
+      "learning_rate": 1.1766929993086547e-05,
+      "loss": 0.2128,
+      "step": 9049
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6385073345023113,
+      "learning_rate": 1.1765308913887341e-05,
+      "loss": 0.1901,
+      "step": 9050
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.964258379970069,
+      "learning_rate": 1.1763687786804034e-05,
+      "loss": 0.1797,
+      "step": 9051
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1455400657359058,
+      "learning_rate": 1.1762066611880596e-05,
+      "loss": 0.1977,
+      "step": 9052
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8002007770933686,
+      "learning_rate": 1.1760445389161002e-05,
+      "loss": 0.1776,
+      "step": 9053
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.943694401307601,
+      "learning_rate": 1.1758824118689225e-05,
+      "loss": 0.2099,
+      "step": 9054
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0283652595881423,
+      "learning_rate": 1.1757202800509245e-05,
+      "loss": 0.1996,
+      "step": 9055
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9931005687344918,
+      "learning_rate": 1.1755581434665043e-05,
+      "loss": 0.18,
+      "step": 9056
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.238537129714373,
+      "learning_rate": 1.1753960021200593e-05,
+      "loss": 0.1984,
+      "step": 9057
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7553415966956479,
+      "learning_rate": 1.1752338560159878e-05,
+      "loss": 0.1678,
+      "step": 9058
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0408684215597828,
+      "learning_rate": 1.1750717051586882e-05,
+      "loss": 0.1895,
+      "step": 9059
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0384347110753673,
+      "learning_rate": 1.1749095495525584e-05,
+      "loss": 0.1906,
+      "step": 9060
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.044673644971104,
+      "learning_rate": 1.1747473892019977e-05,
+      "loss": 0.1983,
+      "step": 9061
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.109794063699578,
+      "learning_rate": 1.1745852241114038e-05,
+      "loss": 0.2028,
+      "step": 9062
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.084289158802793,
+      "learning_rate": 1.1744230542851758e-05,
+      "loss": 0.1999,
+      "step": 9063
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.777145804319658,
+      "learning_rate": 1.174260879727713e-05,
+      "loss": 0.1711,
+      "step": 9064
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0703078027644066,
+      "learning_rate": 1.1740987004434137e-05,
+      "loss": 0.1918,
+      "step": 9065
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8781926448135073,
+      "learning_rate": 1.1739365164366775e-05,
+      "loss": 0.198,
+      "step": 9066
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3489089861431531,
+      "learning_rate": 1.1737743277119031e-05,
+      "loss": 0.1865,
+      "step": 9067
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9097810356618725,
+      "learning_rate": 1.1736121342734903e-05,
+      "loss": 0.1857,
+      "step": 9068
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0893069302099345,
+      "learning_rate": 1.173449936125839e-05,
+      "loss": 0.186,
+      "step": 9069
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8277620769173616,
+      "learning_rate": 1.1732877332733479e-05,
+      "loss": 0.1812,
+      "step": 9070
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3249669025073985,
+      "learning_rate": 1.1731255257204174e-05,
+      "loss": 0.1903,
+      "step": 9071
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1042809564014204,
+      "learning_rate": 1.1729633134714475e-05,
+      "loss": 0.2004,
+      "step": 9072
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9206027756962021,
+      "learning_rate": 1.1728010965308377e-05,
+      "loss": 0.1911,
+      "step": 9073
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0423116504989947,
+      "learning_rate": 1.1726388749029884e-05,
+      "loss": 0.188,
+      "step": 9074
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7968196186056415,
+      "learning_rate": 1.1724766485922998e-05,
+      "loss": 0.1874,
+      "step": 9075
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9243988557355096,
+      "learning_rate": 1.1723144176031727e-05,
+      "loss": 0.1881,
+      "step": 9076
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.042151183072485,
+      "learning_rate": 1.1721521819400068e-05,
+      "loss": 0.1965,
+      "step": 9077
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9206453558272957,
+      "learning_rate": 1.1719899416072037e-05,
+      "loss": 0.1782,
+      "step": 9078
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9587520930099989,
+      "learning_rate": 1.1718276966091638e-05,
+      "loss": 0.1903,
+      "step": 9079
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2943426276561494,
+      "learning_rate": 1.1716654469502875e-05,
+      "loss": 0.218,
+      "step": 9080
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.124829320688035,
+      "learning_rate": 1.1715031926349763e-05,
+      "loss": 0.1913,
+      "step": 9081
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8456629905306549,
+      "learning_rate": 1.1713409336676313e-05,
+      "loss": 0.1845,
+      "step": 9082
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8710014737852795,
+      "learning_rate": 1.1711786700526541e-05,
+      "loss": 0.1871,
+      "step": 9083
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2990723070408405,
+      "learning_rate": 1.1710164017944456e-05,
+      "loss": 0.1872,
+      "step": 9084
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.236419696060247,
+      "learning_rate": 1.1708541288974074e-05,
+      "loss": 0.2054,
+      "step": 9085
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0682702113309683,
+      "learning_rate": 1.1706918513659416e-05,
+      "loss": 0.1932,
+      "step": 9086
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8068307461764347,
+      "learning_rate": 1.1705295692044496e-05,
+      "loss": 0.173,
+      "step": 9087
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3880576279247103,
+      "learning_rate": 1.1703672824173333e-05,
+      "loss": 0.1704,
+      "step": 9088
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.270121404719117,
+      "learning_rate": 1.170204991008995e-05,
+      "loss": 0.1879,
+      "step": 9089
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0435643904778924,
+      "learning_rate": 1.1700426949838364e-05,
+      "loss": 0.1885,
+      "step": 9090
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8568140770975008,
+      "learning_rate": 1.1698803943462604e-05,
+      "loss": 0.1848,
+      "step": 9091
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0820704614596093,
+      "learning_rate": 1.169718089100669e-05,
+      "loss": 0.212,
+      "step": 9092
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0800354792918512,
+      "learning_rate": 1.1695557792514648e-05,
+      "loss": 0.1971,
+      "step": 9093
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1750232689570579,
+      "learning_rate": 1.1693934648030508e-05,
+      "loss": 0.1914,
+      "step": 9094
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3779916093593438,
+      "learning_rate": 1.1692311457598291e-05,
+      "loss": 0.184,
+      "step": 9095
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8684209284055365,
+      "learning_rate": 1.1690688221262035e-05,
+      "loss": 0.1927,
+      "step": 9096
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.697502018051808,
+      "learning_rate": 1.1689064939065762e-05,
+      "loss": 0.1879,
+      "step": 9097
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0772614966661544,
+      "learning_rate": 1.1687441611053505e-05,
+      "loss": 0.1772,
+      "step": 9098
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1014332449661124,
+      "learning_rate": 1.1685818237269302e-05,
+      "loss": 0.1901,
+      "step": 9099
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.013613860724744,
+      "learning_rate": 1.1684194817757184e-05,
+      "loss": 0.2189,
+      "step": 9100
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0590001336397885,
+      "learning_rate": 1.1682571352561187e-05,
+      "loss": 0.1967,
+      "step": 9101
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1143069368763483,
+      "learning_rate": 1.1680947841725348e-05,
+      "loss": 0.1986,
+      "step": 9102
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9463708285613962,
+      "learning_rate": 1.1679324285293698e-05,
+      "loss": 0.1896,
+      "step": 9103
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0438085528108987,
+      "learning_rate": 1.1677700683310286e-05,
+      "loss": 0.1854,
+      "step": 9104
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.92252065973537,
+      "learning_rate": 1.1676077035819148e-05,
+      "loss": 0.2096,
+      "step": 9105
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1214717138794612,
+      "learning_rate": 1.1674453342864327e-05,
+      "loss": 0.1925,
+      "step": 9106
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0099356972626463,
+      "learning_rate": 1.1672829604489864e-05,
+      "loss": 0.1649,
+      "step": 9107
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1449027127665157,
+      "learning_rate": 1.16712058207398e-05,
+      "loss": 0.215,
+      "step": 9108
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.911336039272148,
+      "learning_rate": 1.1669581991658187e-05,
+      "loss": 0.1711,
+      "step": 9109
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4580166182319019,
+      "learning_rate": 1.1667958117289068e-05,
+      "loss": 0.2201,
+      "step": 9110
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0857664654881272,
+      "learning_rate": 1.1666334197676492e-05,
+      "loss": 0.1829,
+      "step": 9111
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8859738895387417,
+      "learning_rate": 1.1664710232864505e-05,
+      "loss": 0.1673,
+      "step": 9112
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7798264118555199,
+      "learning_rate": 1.1663086222897157e-05,
+      "loss": 0.2089,
+      "step": 9113
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9975538121325291,
+      "learning_rate": 1.1661462167818507e-05,
+      "loss": 0.1873,
+      "step": 9114
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0709575867805985,
+      "learning_rate": 1.16598380676726e-05,
+      "loss": 0.1843,
+      "step": 9115
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9233783373327914,
+      "learning_rate": 1.1658213922503488e-05,
+      "loss": 0.1582,
+      "step": 9116
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0645080907936217,
+      "learning_rate": 1.1656589732355233e-05,
+      "loss": 0.1973,
+      "step": 9117
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9392700054586304,
+      "learning_rate": 1.1654965497271886e-05,
+      "loss": 0.1929,
+      "step": 9118
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.114398197170294,
+      "learning_rate": 1.1653341217297507e-05,
+      "loss": 0.1915,
+      "step": 9119
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9674687619310163,
+      "learning_rate": 1.1651716892476154e-05,
+      "loss": 0.1944,
+      "step": 9120
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9367552026693869,
+      "learning_rate": 1.1650092522851885e-05,
+      "loss": 0.1898,
+      "step": 9121
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0708496935895828,
+      "learning_rate": 1.1648468108468767e-05,
+      "loss": 0.1894,
+      "step": 9122
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8815096620043407,
+      "learning_rate": 1.1646843649370858e-05,
+      "loss": 0.1887,
+      "step": 9123
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9908797866048731,
+      "learning_rate": 1.164521914560222e-05,
+      "loss": 0.1956,
+      "step": 9124
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0955654895177573,
+      "learning_rate": 1.164359459720692e-05,
+      "loss": 0.1917,
+      "step": 9125
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9897343143053348,
+      "learning_rate": 1.1641970004229025e-05,
+      "loss": 0.1836,
+      "step": 9126
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7886372445419565,
+      "learning_rate": 1.16403453667126e-05,
+      "loss": 0.2013,
+      "step": 9127
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3670924199438759,
+      "learning_rate": 1.1638720684701714e-05,
+      "loss": 0.1719,
+      "step": 9128
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6908065901193217,
+      "learning_rate": 1.1637095958240439e-05,
+      "loss": 0.2039,
+      "step": 9129
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8688739782273087,
+      "learning_rate": 1.163547118737284e-05,
+      "loss": 0.1984,
+      "step": 9130
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8693621221208244,
+      "learning_rate": 1.1633846372142997e-05,
+      "loss": 0.1758,
+      "step": 9131
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0033795035288822,
+      "learning_rate": 1.1632221512594977e-05,
+      "loss": 0.1829,
+      "step": 9132
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.203854492354105,
+      "learning_rate": 1.1630596608772855e-05,
+      "loss": 0.1991,
+      "step": 9133
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7739788349490826,
+      "learning_rate": 1.1628971660720707e-05,
+      "loss": 0.2161,
+      "step": 9134
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3333093970645196,
+      "learning_rate": 1.1627346668482614e-05,
+      "loss": 0.2002,
+      "step": 9135
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3961606707042902,
+      "learning_rate": 1.1625721632102649e-05,
+      "loss": 0.181,
+      "step": 9136
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7972776439352925,
+      "learning_rate": 1.1624096551624893e-05,
+      "loss": 0.1848,
+      "step": 9137
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8296410166518111,
+      "learning_rate": 1.1622471427093424e-05,
+      "loss": 0.1726,
+      "step": 9138
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9506650235484677,
+      "learning_rate": 1.162084625855233e-05,
+      "loss": 0.1721,
+      "step": 9139
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7818432041827315,
+      "learning_rate": 1.1619221046045688e-05,
+      "loss": 0.1942,
+      "step": 9140
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1575333574213695,
+      "learning_rate": 1.1617595789617585e-05,
+      "loss": 0.2027,
+      "step": 9141
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.434177466973586,
+      "learning_rate": 1.1615970489312102e-05,
+      "loss": 0.194,
+      "step": 9142
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.245727117900203,
+      "learning_rate": 1.1614345145173329e-05,
+      "loss": 0.2023,
+      "step": 9143
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4528276599772527,
+      "learning_rate": 1.1612719757245353e-05,
+      "loss": 0.1841,
+      "step": 9144
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.883066228974052,
+      "learning_rate": 1.1611094325572263e-05,
+      "loss": 0.1973,
+      "step": 9145
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.153096925156812,
+      "learning_rate": 1.1609468850198149e-05,
+      "loss": 0.1883,
+      "step": 9146
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2894960675682319,
+      "learning_rate": 1.1607843331167099e-05,
+      "loss": 0.2053,
+      "step": 9147
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8747914564760497,
+      "learning_rate": 1.160621776852321e-05,
+      "loss": 0.1882,
+      "step": 9148
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1131978812303405,
+      "learning_rate": 1.1604592162310575e-05,
+      "loss": 0.2121,
+      "step": 9149
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5693726179321674,
+      "learning_rate": 1.1602966512573286e-05,
+      "loss": 0.1881,
+      "step": 9150
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0216723824159144,
+      "learning_rate": 1.1601340819355437e-05,
+      "loss": 0.1912,
+      "step": 9151
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7815028235272667,
+      "learning_rate": 1.159971508270113e-05,
+      "loss": 0.1954,
+      "step": 9152
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9754311439115084,
+      "learning_rate": 1.159808930265446e-05,
+      "loss": 0.1918,
+      "step": 9153
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0278402602039418,
+      "learning_rate": 1.159646347925953e-05,
+      "loss": 0.1974,
+      "step": 9154
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9032850279132243,
+      "learning_rate": 1.1594837612560437e-05,
+      "loss": 0.1778,
+      "step": 9155
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8774455184882909,
+      "learning_rate": 1.159321170260128e-05,
+      "loss": 0.2006,
+      "step": 9156
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9127838115789007,
+      "learning_rate": 1.159158574942617e-05,
+      "loss": 0.1733,
+      "step": 9157
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8448337659172754,
+      "learning_rate": 1.1589959753079203e-05,
+      "loss": 0.1759,
+      "step": 9158
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2241639207791157,
+      "learning_rate": 1.1588333713604491e-05,
+      "loss": 0.2184,
+      "step": 9159
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7723198053308619,
+      "learning_rate": 1.1586707631046135e-05,
+      "loss": 0.1973,
+      "step": 9160
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8615544161233657,
+      "learning_rate": 1.1585081505448246e-05,
+      "loss": 0.2044,
+      "step": 9161
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8611882561608455,
+      "learning_rate": 1.1583455336854932e-05,
+      "loss": 0.1845,
+      "step": 9162
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0851258693274464,
+      "learning_rate": 1.1581829125310302e-05,
+      "loss": 0.179,
+      "step": 9163
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7737070155432831,
+      "learning_rate": 1.1580202870858468e-05,
+      "loss": 0.1768,
+      "step": 9164
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9287374921183846,
+      "learning_rate": 1.1578576573543541e-05,
+      "loss": 0.1976,
+      "step": 9165
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8834705337603755,
+      "learning_rate": 1.1576950233409638e-05,
+      "loss": 0.1852,
+      "step": 9166
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8774831452300471,
+      "learning_rate": 1.1575323850500868e-05,
+      "loss": 0.1702,
+      "step": 9167
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4809191685656273,
+      "learning_rate": 1.1573697424861353e-05,
+      "loss": 0.1834,
+      "step": 9168
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9047569129335316,
+      "learning_rate": 1.1572070956535202e-05,
+      "loss": 0.1878,
+      "step": 9169
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0114729627753714,
+      "learning_rate": 1.1570444445566538e-05,
+      "loss": 0.1841,
+      "step": 9170
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0533951490813414,
+      "learning_rate": 1.1568817891999482e-05,
+      "loss": 0.2076,
+      "step": 9171
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0908645916592143,
+      "learning_rate": 1.1567191295878152e-05,
+      "loss": 0.1858,
+      "step": 9172
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9703974422296601,
+      "learning_rate": 1.1565564657246667e-05,
+      "loss": 0.2004,
+      "step": 9173
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0118705478228984,
+      "learning_rate": 1.1563937976149153e-05,
+      "loss": 0.1953,
+      "step": 9174
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8260904591199736,
+      "learning_rate": 1.1562311252629736e-05,
+      "loss": 0.2017,
+      "step": 9175
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7339715507228588,
+      "learning_rate": 1.1560684486732537e-05,
+      "loss": 0.1916,
+      "step": 9176
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8509635990034636,
+      "learning_rate": 1.1559057678501682e-05,
+      "loss": 0.1829,
+      "step": 9177
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9624278814958298,
+      "learning_rate": 1.1557430827981297e-05,
+      "loss": 0.2154,
+      "step": 9178
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.479984283272646,
+      "learning_rate": 1.1555803935215516e-05,
+      "loss": 0.2187,
+      "step": 9179
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9547459427889696,
+      "learning_rate": 1.1554177000248466e-05,
+      "loss": 0.192,
+      "step": 9180
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0713234269482474,
+      "learning_rate": 1.1552550023124276e-05,
+      "loss": 0.1993,
+      "step": 9181
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.084265326853148,
+      "learning_rate": 1.155092300388708e-05,
+      "loss": 0.1896,
+      "step": 9182
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.823005443489001,
+      "learning_rate": 1.1549295942581007e-05,
+      "loss": 0.1774,
+      "step": 9183
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6002085217547777,
+      "learning_rate": 1.1547668839250199e-05,
+      "loss": 0.1919,
+      "step": 9184
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8626993942068936,
+      "learning_rate": 1.1546041693938784e-05,
+      "loss": 0.192,
+      "step": 9185
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0077475214317404,
+      "learning_rate": 1.1544414506690897e-05,
+      "loss": 0.1848,
+      "step": 9186
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0757353918928187,
+      "learning_rate": 1.1542787277550683e-05,
+      "loss": 0.2179,
+      "step": 9187
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.08680507436135,
+      "learning_rate": 1.1541160006562275e-05,
+      "loss": 0.1883,
+      "step": 9188
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9751665996225382,
+      "learning_rate": 1.1539532693769818e-05,
+      "loss": 0.1974,
+      "step": 9189
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.48160369171706,
+      "learning_rate": 1.1537905339217448e-05,
+      "loss": 0.1883,
+      "step": 9190
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.975263185759104,
+      "learning_rate": 1.1536277942949305e-05,
+      "loss": 0.2017,
+      "step": 9191
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8240373691871361,
+      "learning_rate": 1.1534650505009542e-05,
+      "loss": 0.1775,
+      "step": 9192
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3049402644744645,
+      "learning_rate": 1.1533023025442294e-05,
+      "loss": 0.2072,
+      "step": 9193
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9742705420566906,
+      "learning_rate": 1.1531395504291711e-05,
+      "loss": 0.1864,
+      "step": 9194
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9472394210632011,
+      "learning_rate": 1.1529767941601937e-05,
+      "loss": 0.1836,
+      "step": 9195
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9238803406730269,
+      "learning_rate": 1.1528140337417121e-05,
+      "loss": 0.1921,
+      "step": 9196
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8351639923254849,
+      "learning_rate": 1.1526512691781415e-05,
+      "loss": 0.1927,
+      "step": 9197
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2967717494314384,
+      "learning_rate": 1.1524885004738966e-05,
+      "loss": 0.1871,
+      "step": 9198
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.915047277005369,
+      "learning_rate": 1.1523257276333924e-05,
+      "loss": 0.2058,
+      "step": 9199
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8859798604811852,
+      "learning_rate": 1.1521629506610439e-05,
+      "loss": 0.2082,
+      "step": 9200
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9162438028535452,
+      "learning_rate": 1.1520001695612675e-05,
+      "loss": 0.2077,
+      "step": 9201
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8868233681525796,
+      "learning_rate": 1.1518373843384774e-05,
+      "loss": 0.1949,
+      "step": 9202
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.893103606569373,
+      "learning_rate": 1.1516745949970897e-05,
+      "loss": 0.1734,
+      "step": 9203
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0350623362067328,
+      "learning_rate": 1.15151180154152e-05,
+      "loss": 0.1957,
+      "step": 9204
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7472797149020382,
+      "learning_rate": 1.1513490039761843e-05,
+      "loss": 0.1719,
+      "step": 9205
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8943678903413873,
+      "learning_rate": 1.1511862023054983e-05,
+      "loss": 0.1849,
+      "step": 9206
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0615004595864916,
+      "learning_rate": 1.151023396533878e-05,
+      "loss": 0.1849,
+      "step": 9207
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0065551689099037,
+      "learning_rate": 1.1508605866657392e-05,
+      "loss": 0.2165,
+      "step": 9208
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8278316868482662,
+      "learning_rate": 1.1506977727054988e-05,
+      "loss": 0.1768,
+      "step": 9209
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0617148335170974,
+      "learning_rate": 1.1505349546575728e-05,
+      "loss": 0.1931,
+      "step": 9210
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8807830721570364,
+      "learning_rate": 1.1503721325263778e-05,
+      "loss": 0.2041,
+      "step": 9211
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.131425982837395,
+      "learning_rate": 1.15020930631633e-05,
+      "loss": 0.1908,
+      "step": 9212
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.016980414378633,
+      "learning_rate": 1.1500464760318462e-05,
+      "loss": 0.1705,
+      "step": 9213
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3114196960927185,
+      "learning_rate": 1.1498836416773433e-05,
+      "loss": 0.2042,
+      "step": 9214
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9548647672267575,
+      "learning_rate": 1.1497208032572385e-05,
+      "loss": 0.1688,
+      "step": 9215
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4053452700177105,
+      "learning_rate": 1.149557960775948e-05,
+      "loss": 0.2194,
+      "step": 9216
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9455345695137635,
+      "learning_rate": 1.1493951142378896e-05,
+      "loss": 0.174,
+      "step": 9217
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.914664529750119,
+      "learning_rate": 1.1492322636474802e-05,
+      "loss": 0.1867,
+      "step": 9218
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9728464313455532,
+      "learning_rate": 1.1490694090091375e-05,
+      "loss": 0.1737,
+      "step": 9219
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8613771294384492,
+      "learning_rate": 1.1489065503272785e-05,
+      "loss": 0.1906,
+      "step": 9220
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9827950275426468,
+      "learning_rate": 1.1487436876063205e-05,
+      "loss": 0.2,
+      "step": 9221
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.253797589125273,
+      "learning_rate": 1.148580820850682e-05,
+      "loss": 0.1748,
+      "step": 9222
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.92996072269765,
+      "learning_rate": 1.1484179500647802e-05,
+      "loss": 0.1813,
+      "step": 9223
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9294885667105391,
+      "learning_rate": 1.1482550752530332e-05,
+      "loss": 0.1911,
+      "step": 9224
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4144234683841561,
+      "learning_rate": 1.148092196419859e-05,
+      "loss": 0.2001,
+      "step": 9225
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8613645922670157,
+      "learning_rate": 1.1479293135696755e-05,
+      "loss": 0.2162,
+      "step": 9226
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0450389190608478,
+      "learning_rate": 1.1477664267069009e-05,
+      "loss": 0.1856,
+      "step": 9227
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8930640924927837,
+      "learning_rate": 1.1476035358359539e-05,
+      "loss": 0.2097,
+      "step": 9228
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.592902215469274,
+      "learning_rate": 1.1474406409612524e-05,
+      "loss": 0.1921,
+      "step": 9229
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7904460656923809,
+      "learning_rate": 1.1472777420872154e-05,
+      "loss": 0.1891,
+      "step": 9230
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8015309344030134,
+      "learning_rate": 1.147114839218261e-05,
+      "loss": 0.1601,
+      "step": 9231
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9705118301556879,
+      "learning_rate": 1.1469519323588085e-05,
+      "loss": 0.1639,
+      "step": 9232
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5770038604029533,
+      "learning_rate": 1.1467890215132767e-05,
+      "loss": 0.1919,
+      "step": 9233
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1379908559648302,
+      "learning_rate": 1.146626106686084e-05,
+      "loss": 0.1739,
+      "step": 9234
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7509321612595246,
+      "learning_rate": 1.1464631878816502e-05,
+      "loss": 0.1993,
+      "step": 9235
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1126365561109344,
+      "learning_rate": 1.1463002651043942e-05,
+      "loss": 0.1867,
+      "step": 9236
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9436631209359791,
+      "learning_rate": 1.146137338358735e-05,
+      "loss": 0.1892,
+      "step": 9237
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.293770946685034,
+      "learning_rate": 1.1459744076490924e-05,
+      "loss": 0.1957,
+      "step": 9238
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9276318520781688,
+      "learning_rate": 1.1458114729798855e-05,
+      "loss": 0.1872,
+      "step": 9239
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6795751167978585,
+      "learning_rate": 1.1456485343555344e-05,
+      "loss": 0.1803,
+      "step": 9240
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9264723014647175,
+      "learning_rate": 1.1454855917804586e-05,
+      "loss": 0.1718,
+      "step": 9241
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.457193211728748,
+      "learning_rate": 1.145322645259078e-05,
+      "loss": 0.1898,
+      "step": 9242
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9019050296705318,
+      "learning_rate": 1.1451596947958122e-05,
+      "loss": 0.2239,
+      "step": 9243
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.865206930755326,
+      "learning_rate": 1.1449967403950812e-05,
+      "loss": 0.2017,
+      "step": 9244
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.132181096344825,
+      "learning_rate": 1.1448337820613061e-05,
+      "loss": 0.1885,
+      "step": 9245
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.151784318954177,
+      "learning_rate": 1.1446708197989061e-05,
+      "loss": 0.2025,
+      "step": 9246
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0029738228138243,
+      "learning_rate": 1.144507853612302e-05,
+      "loss": 0.1855,
+      "step": 9247
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0962947125468563,
+      "learning_rate": 1.1443448835059141e-05,
+      "loss": 0.1944,
+      "step": 9248
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8477426463884647,
+      "learning_rate": 1.144181909484163e-05,
+      "loss": 0.1753,
+      "step": 9249
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8817457321386621,
+      "learning_rate": 1.1440189315514698e-05,
+      "loss": 0.1827,
+      "step": 9250
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9198362005845908,
+      "learning_rate": 1.143855949712255e-05,
+      "loss": 0.1758,
+      "step": 9251
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8501168011702994,
+      "learning_rate": 1.143692963970939e-05,
+      "loss": 0.1951,
+      "step": 9252
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1709863082016263,
+      "learning_rate": 1.1435299743319434e-05,
+      "loss": 0.1998,
+      "step": 9253
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9096201448211303,
+      "learning_rate": 1.1433669807996894e-05,
+      "loss": 0.17,
+      "step": 9254
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1348872412188835,
+      "learning_rate": 1.1432039833785979e-05,
+      "loss": 0.1884,
+      "step": 9255
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1259494677848945,
+      "learning_rate": 1.1430409820730902e-05,
+      "loss": 0.1881,
+      "step": 9256
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1273228094294094,
+      "learning_rate": 1.1428779768875874e-05,
+      "loss": 0.1934,
+      "step": 9257
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1033316887349531,
+      "learning_rate": 1.1427149678265119e-05,
+      "loss": 0.2156,
+      "step": 9258
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.008896724459771,
+      "learning_rate": 1.1425519548942847e-05,
+      "loss": 0.2092,
+      "step": 9259
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.263669512820033,
+      "learning_rate": 1.1423889380953277e-05,
+      "loss": 0.1778,
+      "step": 9260
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9381984779225716,
+      "learning_rate": 1.1422259174340624e-05,
+      "loss": 0.1963,
+      "step": 9261
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5088037717808391,
+      "learning_rate": 1.1420628929149114e-05,
+      "loss": 0.2038,
+      "step": 9262
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0717096467954101,
+      "learning_rate": 1.1418998645422963e-05,
+      "loss": 0.1927,
+      "step": 9263
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.193684945593441,
+      "learning_rate": 1.1417368323206395e-05,
+      "loss": 0.2046,
+      "step": 9264
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0533682703561398,
+      "learning_rate": 1.141573796254363e-05,
+      "loss": 0.1697,
+      "step": 9265
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1840786054105483,
+      "learning_rate": 1.141410756347889e-05,
+      "loss": 0.1822,
+      "step": 9266
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.749805516924357,
+      "learning_rate": 1.1412477126056405e-05,
+      "loss": 0.1979,
+      "step": 9267
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.253852439701606,
+      "learning_rate": 1.14108466503204e-05,
+      "loss": 0.2176,
+      "step": 9268
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1173364128331582,
+      "learning_rate": 1.1409216136315097e-05,
+      "loss": 0.1778,
+      "step": 9269
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.035032031791532,
+      "learning_rate": 1.1407585584084726e-05,
+      "loss": 0.2161,
+      "step": 9270
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2405865114634105,
+      "learning_rate": 1.1405954993673522e-05,
+      "loss": 0.1771,
+      "step": 9271
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9961618122285781,
+      "learning_rate": 1.1404324365125708e-05,
+      "loss": 0.193,
+      "step": 9272
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6814547632964016,
+      "learning_rate": 1.1402693698485512e-05,
+      "loss": 0.1832,
+      "step": 9273
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2294370918587325,
+      "learning_rate": 1.1401062993797171e-05,
+      "loss": 0.1903,
+      "step": 9274
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.415889465472221,
+      "learning_rate": 1.1399432251104918e-05,
+      "loss": 0.2137,
+      "step": 9275
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0334241329352223,
+      "learning_rate": 1.1397801470452987e-05,
+      "loss": 0.196,
+      "step": 9276
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9642912610388481,
+      "learning_rate": 1.1396170651885613e-05,
+      "loss": 0.182,
+      "step": 9277
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2648969637697525,
+      "learning_rate": 1.139453979544703e-05,
+      "loss": 0.1908,
+      "step": 9278
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8346050383865007,
+      "learning_rate": 1.1392908901181474e-05,
+      "loss": 0.176,
+      "step": 9279
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3413644643229863,
+      "learning_rate": 1.139127796913319e-05,
+      "loss": 0.1911,
+      "step": 9280
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9490728998996395,
+      "learning_rate": 1.138964699934641e-05,
+      "loss": 0.1796,
+      "step": 9281
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0848424621668546,
+      "learning_rate": 1.1388015991865377e-05,
+      "loss": 0.1773,
+      "step": 9282
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8101725755414195,
+      "learning_rate": 1.138638494673433e-05,
+      "loss": 0.2012,
+      "step": 9283
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3028282362221764,
+      "learning_rate": 1.1384753863997516e-05,
+      "loss": 0.1876,
+      "step": 9284
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1140020330220612,
+      "learning_rate": 1.1383122743699173e-05,
+      "loss": 0.2109,
+      "step": 9285
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1084320608542635,
+      "learning_rate": 1.1381491585883548e-05,
+      "loss": 0.1971,
+      "step": 9286
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3256072713621745,
+      "learning_rate": 1.1379860390594888e-05,
+      "loss": 0.2103,
+      "step": 9287
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9168074800231482,
+      "learning_rate": 1.1378229157877432e-05,
+      "loss": 0.1789,
+      "step": 9288
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8194316208116729,
+      "learning_rate": 1.1376597887775438e-05,
+      "loss": 0.1948,
+      "step": 9289
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.98526556813502,
+      "learning_rate": 1.1374966580333147e-05,
+      "loss": 0.1984,
+      "step": 9290
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.061805734811115,
+      "learning_rate": 1.1373335235594809e-05,
+      "loss": 0.2147,
+      "step": 9291
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.952970495944534,
+      "learning_rate": 1.1371703853604672e-05,
+      "loss": 0.1879,
+      "step": 9292
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0049105878186886,
+      "learning_rate": 1.1370072434406993e-05,
+      "loss": 0.2125,
+      "step": 9293
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.180288824955952,
+      "learning_rate": 1.1368440978046022e-05,
+      "loss": 0.1977,
+      "step": 9294
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9273278457059216,
+      "learning_rate": 1.1366809484566015e-05,
+      "loss": 0.1786,
+      "step": 9295
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7298925974661218,
+      "learning_rate": 1.136517795401122e-05,
+      "loss": 0.18,
+      "step": 9296
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9384320818110587,
+      "learning_rate": 1.1363546386425895e-05,
+      "loss": 0.184,
+      "step": 9297
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7859695205526979,
+      "learning_rate": 1.13619147818543e-05,
+      "loss": 0.167,
+      "step": 9298
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9421177588324476,
+      "learning_rate": 1.1360283140340688e-05,
+      "loss": 0.1688,
+      "step": 9299
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2079617951803552,
+      "learning_rate": 1.1358651461929319e-05,
+      "loss": 0.1966,
+      "step": 9300
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.817762563013867,
+      "learning_rate": 1.1357019746664453e-05,
+      "loss": 0.1757,
+      "step": 9301
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1866910568283346,
+      "learning_rate": 1.135538799459035e-05,
+      "loss": 0.199,
+      "step": 9302
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9677330185793297,
+      "learning_rate": 1.1353756205751272e-05,
+      "loss": 0.2217,
+      "step": 9303
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7153143992431847,
+      "learning_rate": 1.1352124380191479e-05,
+      "loss": 0.1836,
+      "step": 9304
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9525697878726538,
+      "learning_rate": 1.1350492517955234e-05,
+      "loss": 0.1999,
+      "step": 9305
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9008507657210286,
+      "learning_rate": 1.1348860619086808e-05,
+      "loss": 0.1982,
+      "step": 9306
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9362961925883473,
+      "learning_rate": 1.134722868363046e-05,
+      "loss": 0.1963,
+      "step": 9307
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8781050347037737,
+      "learning_rate": 1.1345596711630456e-05,
+      "loss": 0.1846,
+      "step": 9308
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9572232189569388,
+      "learning_rate": 1.1343964703131065e-05,
+      "loss": 0.1866,
+      "step": 9309
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3214437628723248,
+      "learning_rate": 1.1342332658176556e-05,
+      "loss": 0.1705,
+      "step": 9310
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7267957828656673,
+      "learning_rate": 1.1340700576811198e-05,
+      "loss": 0.1847,
+      "step": 9311
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6346922718412682,
+      "learning_rate": 1.1339068459079262e-05,
+      "loss": 0.1748,
+      "step": 9312
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7569900182507847,
+      "learning_rate": 1.1337436305025019e-05,
+      "loss": 0.177,
+      "step": 9313
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.955594887499051,
+      "learning_rate": 1.1335804114692737e-05,
+      "loss": 0.2005,
+      "step": 9314
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9007643934387923,
+      "learning_rate": 1.1334171888126698e-05,
+      "loss": 0.192,
+      "step": 9315
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1458182580341056,
+      "learning_rate": 1.1332539625371166e-05,
+      "loss": 0.179,
+      "step": 9316
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.346045821307898,
+      "learning_rate": 1.1330907326470426e-05,
+      "loss": 0.1999,
+      "step": 9317
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1606367702267062,
+      "learning_rate": 1.1329274991468747e-05,
+      "loss": 0.1847,
+      "step": 9318
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1289045966659277,
+      "learning_rate": 1.1327642620410408e-05,
+      "loss": 0.1875,
+      "step": 9319
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8345586647548614,
+      "learning_rate": 1.1326010213339688e-05,
+      "loss": 0.1749,
+      "step": 9320
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4714817836575558,
+      "learning_rate": 1.132437777030087e-05,
+      "loss": 0.2186,
+      "step": 9321
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7114676751705088,
+      "learning_rate": 1.1322745291338226e-05,
+      "loss": 0.1805,
+      "step": 9322
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1326550840352263,
+      "learning_rate": 1.1321112776496042e-05,
+      "loss": 0.1794,
+      "step": 9323
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8232979729290985,
+      "learning_rate": 1.1319480225818602e-05,
+      "loss": 0.1973,
+      "step": 9324
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9378338038834931,
+      "learning_rate": 1.1317847639350186e-05,
+      "loss": 0.1777,
+      "step": 9325
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.927375096532382,
+      "learning_rate": 1.1316215017135076e-05,
+      "loss": 0.203,
+      "step": 9326
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0456569263599929,
+      "learning_rate": 1.1314582359217558e-05,
+      "loss": 0.2025,
+      "step": 9327
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2830682932161301,
+      "learning_rate": 1.1312949665641923e-05,
+      "loss": 0.1987,
+      "step": 9328
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.80481172348749,
+      "learning_rate": 1.1311316936452452e-05,
+      "loss": 0.1925,
+      "step": 9329
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9271681235067764,
+      "learning_rate": 1.1309684171693435e-05,
+      "loss": 0.1907,
+      "step": 9330
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8884154679752989,
+      "learning_rate": 1.1308051371409162e-05,
+      "loss": 0.1738,
+      "step": 9331
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2122619537768777,
+      "learning_rate": 1.1306418535643922e-05,
+      "loss": 0.1828,
+      "step": 9332
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0197330364239467,
+      "learning_rate": 1.1304785664442003e-05,
+      "loss": 0.1931,
+      "step": 9333
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9782015436799009,
+      "learning_rate": 1.1303152757847702e-05,
+      "loss": 0.2035,
+      "step": 9334
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9159995398951478,
+      "learning_rate": 1.1301519815905309e-05,
+      "loss": 0.1963,
+      "step": 9335
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9347505438490339,
+      "learning_rate": 1.1299886838659114e-05,
+      "loss": 0.1951,
+      "step": 9336
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0526098163364204,
+      "learning_rate": 1.1298253826153415e-05,
+      "loss": 0.1929,
+      "step": 9337
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.193015880699807,
+      "learning_rate": 1.1296620778432512e-05,
+      "loss": 0.1911,
+      "step": 9338
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4693323716969673,
+      "learning_rate": 1.1294987695540695e-05,
+      "loss": 0.1605,
+      "step": 9339
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2914714608353897,
+      "learning_rate": 1.1293354577522264e-05,
+      "loss": 0.1874,
+      "step": 9340
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8955534658899678,
+      "learning_rate": 1.1291721424421518e-05,
+      "loss": 0.1762,
+      "step": 9341
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9457021966906232,
+      "learning_rate": 1.1290088236282752e-05,
+      "loss": 0.1945,
+      "step": 9342
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9463505398747077,
+      "learning_rate": 1.1288455013150275e-05,
+      "loss": 0.1811,
+      "step": 9343
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8537062086454555,
+      "learning_rate": 1.1286821755068375e-05,
+      "loss": 0.1776,
+      "step": 9344
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1395762597930732,
+      "learning_rate": 1.128518846208137e-05,
+      "loss": 0.1853,
+      "step": 9345
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8332298531890091,
+      "learning_rate": 1.128355513423355e-05,
+      "loss": 0.1726,
+      "step": 9346
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.075619778357324,
+      "learning_rate": 1.1281921771569229e-05,
+      "loss": 0.1671,
+      "step": 9347
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1719243199926213,
+      "learning_rate": 1.1280288374132704e-05,
+      "loss": 0.1979,
+      "step": 9348
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1759049357909859,
+      "learning_rate": 1.1278654941968285e-05,
+      "loss": 0.1915,
+      "step": 9349
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8400856096808293,
+      "learning_rate": 1.127702147512028e-05,
+      "loss": 0.1709,
+      "step": 9350
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1444023227495415,
+      "learning_rate": 1.1275387973632994e-05,
+      "loss": 0.2173,
+      "step": 9351
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0254301032804167,
+      "learning_rate": 1.1273754437550738e-05,
+      "loss": 0.1826,
+      "step": 9352
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.30260397154843,
+      "learning_rate": 1.1272120866917821e-05,
+      "loss": 0.1936,
+      "step": 9353
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8536684491265273,
+      "learning_rate": 1.1270487261778554e-05,
+      "loss": 0.1925,
+      "step": 9354
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9385710112540513,
+      "learning_rate": 1.1268853622177248e-05,
+      "loss": 0.1765,
+      "step": 9355
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.90856099719088,
+      "learning_rate": 1.1267219948158215e-05,
+      "loss": 0.201,
+      "step": 9356
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9892179146824301,
+      "learning_rate": 1.1265586239765772e-05,
+      "loss": 0.2016,
+      "step": 9357
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2092280855303295,
+      "learning_rate": 1.1263952497044225e-05,
+      "loss": 0.1823,
+      "step": 9358
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1156350393611687,
+      "learning_rate": 1.1262318720037902e-05,
+      "loss": 0.1772,
+      "step": 9359
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9048227767853564,
+      "learning_rate": 1.1260684908791109e-05,
+      "loss": 0.1869,
+      "step": 9360
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0731084504573483,
+      "learning_rate": 1.1259051063348167e-05,
+      "loss": 0.1601,
+      "step": 9361
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4848992993314354,
+      "learning_rate": 1.1257417183753391e-05,
+      "loss": 0.1813,
+      "step": 9362
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.987609588215139,
+      "learning_rate": 1.1255783270051105e-05,
+      "loss": 0.1609,
+      "step": 9363
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1217420741641009,
+      "learning_rate": 1.125414932228563e-05,
+      "loss": 0.184,
+      "step": 9364
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0911673067257806,
+      "learning_rate": 1.1252515340501282e-05,
+      "loss": 0.1864,
+      "step": 9365
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2569941304967243,
+      "learning_rate": 1.1250881324742382e-05,
+      "loss": 0.2255,
+      "step": 9366
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9116639119858985,
+      "learning_rate": 1.1249247275053256e-05,
+      "loss": 0.1844,
+      "step": 9367
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.083187498682193,
+      "learning_rate": 1.1247613191478231e-05,
+      "loss": 0.181,
+      "step": 9368
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2340382247496755,
+      "learning_rate": 1.1245979074061623e-05,
+      "loss": 0.1796,
+      "step": 9369
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2310194776736283,
+      "learning_rate": 1.1244344922847765e-05,
+      "loss": 0.2008,
+      "step": 9370
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0514258589881065,
+      "learning_rate": 1.1242710737880979e-05,
+      "loss": 0.2235,
+      "step": 9371
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0733065946367741,
+      "learning_rate": 1.1241076519205595e-05,
+      "loss": 0.1793,
+      "step": 9372
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6731214337021334,
+      "learning_rate": 1.123944226686594e-05,
+      "loss": 0.2054,
+      "step": 9373
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0373545401706548,
+      "learning_rate": 1.1237807980906346e-05,
+      "loss": 0.1899,
+      "step": 9374
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8567521447141507,
+      "learning_rate": 1.1236173661371139e-05,
+      "loss": 0.1926,
+      "step": 9375
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6535204425524637,
+      "learning_rate": 1.123453930830465e-05,
+      "loss": 0.1908,
+      "step": 9376
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6849632397852876,
+      "learning_rate": 1.1232904921751216e-05,
+      "loss": 0.2385,
+      "step": 9377
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9553081200938266,
+      "learning_rate": 1.1231270501755162e-05,
+      "loss": 0.1924,
+      "step": 9378
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9561839368777103,
+      "learning_rate": 1.1229636048360828e-05,
+      "loss": 0.2014,
+      "step": 9379
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9692146046944835,
+      "learning_rate": 1.1228001561612547e-05,
+      "loss": 0.1691,
+      "step": 9380
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0657126735953355,
+      "learning_rate": 1.1226367041554655e-05,
+      "loss": 0.1835,
+      "step": 9381
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9752036307321181,
+      "learning_rate": 1.1224732488231487e-05,
+      "loss": 0.1828,
+      "step": 9382
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9659786017577724,
+      "learning_rate": 1.1223097901687382e-05,
+      "loss": 0.1909,
+      "step": 9383
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8918016838091298,
+      "learning_rate": 1.1221463281966673e-05,
+      "loss": 0.1661,
+      "step": 9384
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8987191636079584,
+      "learning_rate": 1.1219828629113707e-05,
+      "loss": 0.1765,
+      "step": 9385
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.45173704218604,
+      "learning_rate": 1.1218193943172821e-05,
+      "loss": 0.2047,
+      "step": 9386
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6950586457712218,
+      "learning_rate": 1.1216559224188355e-05,
+      "loss": 0.2014,
+      "step": 9387
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0394831725837366,
+      "learning_rate": 1.1214924472204651e-05,
+      "loss": 0.1983,
+      "step": 9388
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2175066037402202,
+      "learning_rate": 1.1213289687266052e-05,
+      "loss": 0.1872,
+      "step": 9389
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3022976481634068,
+      "learning_rate": 1.1211654869416901e-05,
+      "loss": 0.2173,
+      "step": 9390
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.027564229058217,
+      "learning_rate": 1.1210020018701546e-05,
+      "loss": 0.1745,
+      "step": 9391
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1945177307533967,
+      "learning_rate": 1.1208385135164329e-05,
+      "loss": 0.2103,
+      "step": 9392
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8846131409121079,
+      "learning_rate": 1.120675021884959e-05,
+      "loss": 0.1483,
+      "step": 9393
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.012946214427254,
+      "learning_rate": 1.1205115269801695e-05,
+      "loss": 0.1726,
+      "step": 9394
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4632348655698826,
+      "learning_rate": 1.1203480288064974e-05,
+      "loss": 0.1836,
+      "step": 9395
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2870539131782863,
+      "learning_rate": 1.1201845273683782e-05,
+      "loss": 0.1799,
+      "step": 9396
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8808388430584856,
+      "learning_rate": 1.1200210226702469e-05,
+      "loss": 0.2151,
+      "step": 9397
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9740500347235727,
+      "learning_rate": 1.1198575147165384e-05,
+      "loss": 0.177,
+      "step": 9398
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5208257341292488,
+      "learning_rate": 1.1196940035116884e-05,
+      "loss": 0.1716,
+      "step": 9399
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.035575690228776,
+      "learning_rate": 1.1195304890601317e-05,
+      "loss": 0.2063,
+      "step": 9400
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.012321287351004,
+      "learning_rate": 1.1193669713663039e-05,
+      "loss": 0.1733,
+      "step": 9401
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3756052531505762,
+      "learning_rate": 1.1192034504346397e-05,
+      "loss": 0.1951,
+      "step": 9402
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4966464979842127,
+      "learning_rate": 1.1190399262695757e-05,
+      "loss": 0.1959,
+      "step": 9403
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9375503280340816,
+      "learning_rate": 1.1188763988755467e-05,
+      "loss": 0.183,
+      "step": 9404
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1260862585486986,
+      "learning_rate": 1.1187128682569888e-05,
+      "loss": 0.1945,
+      "step": 9405
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9728000538819264,
+      "learning_rate": 1.1185493344183375e-05,
+      "loss": 0.1859,
+      "step": 9406
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0834736784120405,
+      "learning_rate": 1.1183857973640289e-05,
+      "loss": 0.1855,
+      "step": 9407
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9348768976215369,
+      "learning_rate": 1.118222257098499e-05,
+      "loss": 0.1856,
+      "step": 9408
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9706361044266855,
+      "learning_rate": 1.1180587136261835e-05,
+      "loss": 0.1889,
+      "step": 9409
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9726258225414485,
+      "learning_rate": 1.117895166951519e-05,
+      "loss": 0.1734,
+      "step": 9410
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0166944674563385,
+      "learning_rate": 1.1177316170789412e-05,
+      "loss": 0.1909,
+      "step": 9411
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8494335568980105,
+      "learning_rate": 1.1175680640128867e-05,
+      "loss": 0.2172,
+      "step": 9412
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9503780665887606,
+      "learning_rate": 1.117404507757792e-05,
+      "loss": 0.1835,
+      "step": 9413
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0304138091835113,
+      "learning_rate": 1.1172409483180929e-05,
+      "loss": 0.1669,
+      "step": 9414
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3386853123357887,
+      "learning_rate": 1.1170773856982268e-05,
+      "loss": 0.1824,
+      "step": 9415
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.192617945116487,
+      "learning_rate": 1.11691381990263e-05,
+      "loss": 0.1875,
+      "step": 9416
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.044363310431338,
+      "learning_rate": 1.1167502509357393e-05,
+      "loss": 0.1805,
+      "step": 9417
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8158167248839474,
+      "learning_rate": 1.1165866788019912e-05,
+      "loss": 0.1921,
+      "step": 9418
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4395607267781292,
+      "learning_rate": 1.1164231035058228e-05,
+      "loss": 0.2025,
+      "step": 9419
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.015506286236725,
+      "learning_rate": 1.1162595250516715e-05,
+      "loss": 0.1666,
+      "step": 9420
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5064817417365441,
+      "learning_rate": 1.116095943443974e-05,
+      "loss": 0.1904,
+      "step": 9421
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7158616883607665,
+      "learning_rate": 1.1159323586871673e-05,
+      "loss": 0.1879,
+      "step": 9422
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8696293588736635,
+      "learning_rate": 1.1157687707856888e-05,
+      "loss": 0.1676,
+      "step": 9423
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2717209787363684,
+      "learning_rate": 1.1156051797439757e-05,
+      "loss": 0.177,
+      "step": 9424
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0687536577180041,
+      "learning_rate": 1.1154415855664657e-05,
+      "loss": 0.2175,
+      "step": 9425
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.487045634947814,
+      "learning_rate": 1.1152779882575964e-05,
+      "loss": 0.1979,
+      "step": 9426
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.247667420377959,
+      "learning_rate": 1.115114387821805e-05,
+      "loss": 0.2063,
+      "step": 9427
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9232023822380149,
+      "learning_rate": 1.1149507842635293e-05,
+      "loss": 0.193,
+      "step": 9428
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0482885930779677,
+      "learning_rate": 1.1147871775872072e-05,
+      "loss": 0.1661,
+      "step": 9429
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9521862144910216,
+      "learning_rate": 1.1146235677972765e-05,
+      "loss": 0.1883,
+      "step": 9430
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8533427250808602,
+      "learning_rate": 1.1144599548981749e-05,
+      "loss": 0.2021,
+      "step": 9431
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0419278456030308,
+      "learning_rate": 1.1142963388943405e-05,
+      "loss": 0.1666,
+      "step": 9432
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7751955248876855,
+      "learning_rate": 1.1141327197902114e-05,
+      "loss": 0.163,
+      "step": 9433
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8975275086310701,
+      "learning_rate": 1.113969097590226e-05,
+      "loss": 0.1881,
+      "step": 9434
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.92585182171884,
+      "learning_rate": 1.1138054722988223e-05,
+      "loss": 0.2046,
+      "step": 9435
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7624586845530239,
+      "learning_rate": 1.1136418439204388e-05,
+      "loss": 0.1841,
+      "step": 9436
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.557712694484254,
+      "learning_rate": 1.1134782124595136e-05,
+      "loss": 0.1889,
+      "step": 9437
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0712559355250177,
+      "learning_rate": 1.113314577920486e-05,
+      "loss": 0.204,
+      "step": 9438
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9869076526938929,
+      "learning_rate": 1.1131509403077936e-05,
+      "loss": 0.1832,
+      "step": 9439
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4004004264938446,
+      "learning_rate": 1.1129872996258757e-05,
+      "loss": 0.2028,
+      "step": 9440
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8442967787105335,
+      "learning_rate": 1.1128236558791708e-05,
+      "loss": 0.1941,
+      "step": 9441
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9274354478195815,
+      "learning_rate": 1.112660009072118e-05,
+      "loss": 0.1768,
+      "step": 9442
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.130228477397798,
+      "learning_rate": 1.1124963592091563e-05,
+      "loss": 0.1773,
+      "step": 9443
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9728083311241017,
+      "learning_rate": 1.1123327062947243e-05,
+      "loss": 0.2137,
+      "step": 9444
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9970451783684262,
+      "learning_rate": 1.1121690503332613e-05,
+      "loss": 0.1762,
+      "step": 9445
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8959882886040591,
+      "learning_rate": 1.1120053913292066e-05,
+      "loss": 0.1735,
+      "step": 9446
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1404202161163575,
+      "learning_rate": 1.1118417292869992e-05,
+      "loss": 0.2025,
+      "step": 9447
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9352040327694172,
+      "learning_rate": 1.1116780642110785e-05,
+      "loss": 0.1955,
+      "step": 9448
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3641165860043831,
+      "learning_rate": 1.1115143961058843e-05,
+      "loss": 0.182,
+      "step": 9449
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0398020609796403,
+      "learning_rate": 1.1113507249758553e-05,
+      "loss": 0.1765,
+      "step": 9450
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6875637408914828,
+      "learning_rate": 1.111187050825432e-05,
+      "loss": 0.1832,
+      "step": 9451
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0012695602418653,
+      "learning_rate": 1.1110233736590535e-05,
+      "loss": 0.1798,
+      "step": 9452
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.137091287242088,
+      "learning_rate": 1.1108596934811598e-05,
+      "loss": 0.2062,
+      "step": 9453
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.073121214044263,
+      "learning_rate": 1.1106960102961906e-05,
+      "loss": 0.1909,
+      "step": 9454
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.919479504533419,
+      "learning_rate": 1.1105323241085856e-05,
+      "loss": 0.1874,
+      "step": 9455
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0339051802221684,
+      "learning_rate": 1.1103686349227856e-05,
+      "loss": 0.1888,
+      "step": 9456
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9920965654107798,
+      "learning_rate": 1.11020494274323e-05,
+      "loss": 0.1766,
+      "step": 9457
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.091572426986709,
+      "learning_rate": 1.1100412475743589e-05,
+      "loss": 0.198,
+      "step": 9458
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9002278553615146,
+      "learning_rate": 1.1098775494206126e-05,
+      "loss": 0.1809,
+      "step": 9459
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9662421903169365,
+      "learning_rate": 1.109713848286432e-05,
+      "loss": 0.1869,
+      "step": 9460
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0745650451858144,
+      "learning_rate": 1.1095501441762568e-05,
+      "loss": 0.1761,
+      "step": 9461
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6982312184914814,
+      "learning_rate": 1.109386437094528e-05,
+      "loss": 0.1811,
+      "step": 9462
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8952950825627658,
+      "learning_rate": 1.1092227270456857e-05,
+      "loss": 0.1971,
+      "step": 9463
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1015031040313652,
+      "learning_rate": 1.1090590140341709e-05,
+      "loss": 0.2004,
+      "step": 9464
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8780746732320548,
+      "learning_rate": 1.1088952980644242e-05,
+      "loss": 0.1639,
+      "step": 9465
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0166800530835955,
+      "learning_rate": 1.1087315791408864e-05,
+      "loss": 0.1938,
+      "step": 9466
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.893028751235322,
+      "learning_rate": 1.1085678572679978e-05,
+      "loss": 0.1791,
+      "step": 9467
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1214063958853724,
+      "learning_rate": 1.1084041324502006e-05,
+      "loss": 0.1951,
+      "step": 9468
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9458754700189468,
+      "learning_rate": 1.108240404691935e-05,
+      "loss": 0.1844,
+      "step": 9469
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4094387971337061,
+      "learning_rate": 1.1080766739976424e-05,
+      "loss": 0.2075,
+      "step": 9470
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9267791702679756,
+      "learning_rate": 1.1079129403717639e-05,
+      "loss": 0.1891,
+      "step": 9471
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.480391044904194,
+      "learning_rate": 1.1077492038187403e-05,
+      "loss": 0.2007,
+      "step": 9472
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2714801697325204,
+      "learning_rate": 1.107585464343014e-05,
+      "loss": 0.1809,
+      "step": 9473
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0151298523185912,
+      "learning_rate": 1.1074217219490258e-05,
+      "loss": 0.1926,
+      "step": 9474
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4766124566730703,
+      "learning_rate": 1.1072579766412172e-05,
+      "loss": 0.2004,
+      "step": 9475
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0682720487733304,
+      "learning_rate": 1.10709422842403e-05,
+      "loss": 0.1746,
+      "step": 9476
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0623812204356664,
+      "learning_rate": 1.1069304773019058e-05,
+      "loss": 0.2052,
+      "step": 9477
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.98325888769799,
+      "learning_rate": 1.1067667232792864e-05,
+      "loss": 0.1969,
+      "step": 9478
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.249394726179517,
+      "learning_rate": 1.1066029663606138e-05,
+      "loss": 0.1765,
+      "step": 9479
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1215393168986563,
+      "learning_rate": 1.1064392065503294e-05,
+      "loss": 0.2001,
+      "step": 9480
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9950671645201333,
+      "learning_rate": 1.1062754438528758e-05,
+      "loss": 0.1866,
+      "step": 9481
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4252768997024146,
+      "learning_rate": 1.1061116782726947e-05,
+      "loss": 0.2047,
+      "step": 9482
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6916831687929794,
+      "learning_rate": 1.1059479098142281e-05,
+      "loss": 0.1886,
+      "step": 9483
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.298294711202152,
+      "learning_rate": 1.1057841384819185e-05,
+      "loss": 0.1986,
+      "step": 9484
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9317317237057114,
+      "learning_rate": 1.1056203642802081e-05,
+      "loss": 0.1926,
+      "step": 9485
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8458105659558617,
+      "learning_rate": 1.1054565872135397e-05,
+      "loss": 0.189,
+      "step": 9486
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2018879195347907,
+      "learning_rate": 1.1052928072863552e-05,
+      "loss": 0.1777,
+      "step": 9487
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.894558397914112,
+      "learning_rate": 1.1051290245030975e-05,
+      "loss": 0.1987,
+      "step": 9488
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1740428506485563,
+      "learning_rate": 1.1049652388682088e-05,
+      "loss": 0.1954,
+      "step": 9489
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8405196629069694,
+      "learning_rate": 1.1048014503861321e-05,
+      "loss": 0.1678,
+      "step": 9490
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.025401225748899,
+      "learning_rate": 1.1046376590613103e-05,
+      "loss": 0.1766,
+      "step": 9491
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2048673707118842,
+      "learning_rate": 1.104473864898186e-05,
+      "loss": 0.2014,
+      "step": 9492
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2694804452467194,
+      "learning_rate": 1.1043100679012025e-05,
+      "loss": 0.1935,
+      "step": 9493
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8965625555710108,
+      "learning_rate": 1.104146268074802e-05,
+      "loss": 0.1854,
+      "step": 9494
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9878495481703905,
+      "learning_rate": 1.1039824654234286e-05,
+      "loss": 0.1748,
+      "step": 9495
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3167225063683237,
+      "learning_rate": 1.1038186599515247e-05,
+      "loss": 0.2079,
+      "step": 9496
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.148398782309424,
+      "learning_rate": 1.1036548516635339e-05,
+      "loss": 0.1933,
+      "step": 9497
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9509034949490187,
+      "learning_rate": 1.1034910405638992e-05,
+      "loss": 0.1888,
+      "step": 9498
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0496257154764759,
+      "learning_rate": 1.1033272266570645e-05,
+      "loss": 0.1887,
+      "step": 9499
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.035097016626563,
+      "learning_rate": 1.1031634099474727e-05,
+      "loss": 0.1916,
+      "step": 9500
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1664136233350906,
+      "learning_rate": 1.1029995904395676e-05,
+      "loss": 0.2019,
+      "step": 9501
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.132179506013568,
+      "learning_rate": 1.1028357681377928e-05,
+      "loss": 0.179,
+      "step": 9502
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9801111290950743,
+      "learning_rate": 1.1026719430465919e-05,
+      "loss": 0.2018,
+      "step": 9503
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.116829993968936,
+      "learning_rate": 1.1025081151704089e-05,
+      "loss": 0.171,
+      "step": 9504
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9850572439318865,
+      "learning_rate": 1.1023442845136874e-05,
+      "loss": 0.2012,
+      "step": 9505
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1544031189586472,
+      "learning_rate": 1.1021804510808715e-05,
+      "loss": 0.1966,
+      "step": 9506
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1640617416122763,
+      "learning_rate": 1.102016614876405e-05,
+      "loss": 0.1639,
+      "step": 9507
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7324786741544342,
+      "learning_rate": 1.1018527759047319e-05,
+      "loss": 0.1758,
+      "step": 9508
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9896374147248223,
+      "learning_rate": 1.1016889341702968e-05,
+      "loss": 0.1674,
+      "step": 9509
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0471666833332736,
+      "learning_rate": 1.1015250896775436e-05,
+      "loss": 0.1844,
+      "step": 9510
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9148938137694808,
+      "learning_rate": 1.1013612424309163e-05,
+      "loss": 0.1826,
+      "step": 9511
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0928563897735315,
+      "learning_rate": 1.1011973924348599e-05,
+      "loss": 0.2285,
+      "step": 9512
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2076467126569248,
+      "learning_rate": 1.1010335396938183e-05,
+      "loss": 0.2117,
+      "step": 9513
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7580399000340794,
+      "learning_rate": 1.1008696842122364e-05,
+      "loss": 0.1968,
+      "step": 9514
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9902090228799503,
+      "learning_rate": 1.1007058259945584e-05,
+      "loss": 0.1701,
+      "step": 9515
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8934607040812377,
+      "learning_rate": 1.1005419650452294e-05,
+      "loss": 0.2102,
+      "step": 9516
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5274316465729678,
+      "learning_rate": 1.1003781013686939e-05,
+      "loss": 0.192,
+      "step": 9517
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1934406110291276,
+      "learning_rate": 1.1002142349693967e-05,
+      "loss": 0.1596,
+      "step": 9518
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8946649186923104,
+      "learning_rate": 1.1000503658517827e-05,
+      "loss": 0.1874,
+      "step": 9519
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2912572338837025,
+      "learning_rate": 1.0998864940202967e-05,
+      "loss": 0.2044,
+      "step": 9520
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1351873568499273,
+      "learning_rate": 1.0997226194793842e-05,
+      "loss": 0.1764,
+      "step": 9521
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1338616587599282,
+      "learning_rate": 1.09955874223349e-05,
+      "loss": 0.1884,
+      "step": 9522
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9169417601362231,
+      "learning_rate": 1.099394862287059e-05,
+      "loss": 0.1832,
+      "step": 9523
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8400343124098275,
+      "learning_rate": 1.099230979644537e-05,
+      "loss": 0.1825,
+      "step": 9524
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2284716183910207,
+      "learning_rate": 1.0990670943103688e-05,
+      "loss": 0.1828,
+      "step": 9525
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2249751854792137,
+      "learning_rate": 1.0989032062890004e-05,
+      "loss": 0.1851,
+      "step": 9526
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1007619155064332,
+      "learning_rate": 1.0987393155848767e-05,
+      "loss": 0.1899,
+      "step": 9527
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3302949936435415,
+      "learning_rate": 1.0985754222024437e-05,
+      "loss": 0.1778,
+      "step": 9528
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7935309274750862,
+      "learning_rate": 1.0984115261461466e-05,
+      "loss": 0.1679,
+      "step": 9529
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8740483493541972,
+      "learning_rate": 1.0982476274204314e-05,
+      "loss": 0.1912,
+      "step": 9530
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9544592534379195,
+      "learning_rate": 1.0980837260297437e-05,
+      "loss": 0.1756,
+      "step": 9531
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4647054361443923,
+      "learning_rate": 1.0979198219785296e-05,
+      "loss": 0.1668,
+      "step": 9532
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0763420008421867,
+      "learning_rate": 1.0977559152712347e-05,
+      "loss": 0.1942,
+      "step": 9533
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3867599307183982,
+      "learning_rate": 1.0975920059123051e-05,
+      "loss": 0.1943,
+      "step": 9534
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4807281378852024,
+      "learning_rate": 1.0974280939061867e-05,
+      "loss": 0.1769,
+      "step": 9535
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7477775660421317,
+      "learning_rate": 1.0972641792573258e-05,
+      "loss": 0.1705,
+      "step": 9536
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0163715440655499,
+      "learning_rate": 1.0971002619701682e-05,
+      "loss": 0.1845,
+      "step": 9537
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0746521485497653,
+      "learning_rate": 1.096936342049161e-05,
+      "loss": 0.2001,
+      "step": 9538
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8781367171305492,
+      "learning_rate": 1.0967724194987498e-05,
+      "loss": 0.1904,
+      "step": 9539
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.765734891397578,
+      "learning_rate": 1.0966084943233818e-05,
+      "loss": 0.1884,
+      "step": 9540
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.13823561612025,
+      "learning_rate": 1.0964445665275023e-05,
+      "loss": 0.1873,
+      "step": 9541
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9539136937544149,
+      "learning_rate": 1.0962806361155585e-05,
+      "loss": 0.1784,
+      "step": 9542
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0034189361586376,
+      "learning_rate": 1.0961167030919973e-05,
+      "loss": 0.1868,
+      "step": 9543
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2143927037480549,
+      "learning_rate": 1.095952767461265e-05,
+      "loss": 0.1821,
+      "step": 9544
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.54044962525446,
+      "learning_rate": 1.0957888292278084e-05,
+      "loss": 0.1871,
+      "step": 9545
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9402634558591103,
+      "learning_rate": 1.0956248883960744e-05,
+      "loss": 0.1947,
+      "step": 9546
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9362298716995432,
+      "learning_rate": 1.0954609449705097e-05,
+      "loss": 0.1812,
+      "step": 9547
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1257224446570795,
+      "learning_rate": 1.095296998955562e-05,
+      "loss": 0.1759,
+      "step": 9548
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7555764622968162,
+      "learning_rate": 1.0951330503556776e-05,
+      "loss": 0.2133,
+      "step": 9549
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8595309424350507,
+      "learning_rate": 1.0949690991753036e-05,
+      "loss": 0.1849,
+      "step": 9550
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2556944727905175,
+      "learning_rate": 1.0948051454188877e-05,
+      "loss": 0.1932,
+      "step": 9551
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0799262374997667,
+      "learning_rate": 1.094641189090877e-05,
+      "loss": 0.1774,
+      "step": 9552
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2042809302790212,
+      "learning_rate": 1.0944772301957185e-05,
+      "loss": 0.1714,
+      "step": 9553
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8870239839138336,
+      "learning_rate": 1.0943132687378597e-05,
+      "loss": 0.2022,
+      "step": 9554
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7968798872828126,
+      "learning_rate": 1.0941493047217482e-05,
+      "loss": 0.1526,
+      "step": 9555
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9582730119266584,
+      "learning_rate": 1.0939853381518315e-05,
+      "loss": 0.1779,
+      "step": 9556
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8933870974205196,
+      "learning_rate": 1.0938213690325572e-05,
+      "loss": 0.1911,
+      "step": 9557
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8760223525297648,
+      "learning_rate": 1.093657397368373e-05,
+      "loss": 0.1833,
+      "step": 9558
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8660757127061413,
+      "learning_rate": 1.0934934231637267e-05,
+      "loss": 0.2105,
+      "step": 9559
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0198198870011186,
+      "learning_rate": 1.0933294464230657e-05,
+      "loss": 0.1763,
+      "step": 9560
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8730812713789345,
+      "learning_rate": 1.0931654671508384e-05,
+      "loss": 0.203,
+      "step": 9561
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8240497451748958,
+      "learning_rate": 1.0930014853514925e-05,
+      "loss": 0.1809,
+      "step": 9562
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3428577217343693,
+      "learning_rate": 1.0928375010294762e-05,
+      "loss": 0.189,
+      "step": 9563
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9102713113626955,
+      "learning_rate": 1.092673514189237e-05,
+      "loss": 0.194,
+      "step": 9564
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8800419886753066,
+      "learning_rate": 1.092509524835224e-05,
+      "loss": 0.1863,
+      "step": 9565
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8915238012081395,
+      "learning_rate": 1.0923455329718849e-05,
+      "loss": 0.1711,
+      "step": 9566
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8247002058985641,
+      "learning_rate": 1.0921815386036679e-05,
+      "loss": 0.1808,
+      "step": 9567
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9421256090529048,
+      "learning_rate": 1.0920175417350214e-05,
+      "loss": 0.1706,
+      "step": 9568
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5488011925565883,
+      "learning_rate": 1.091853542370394e-05,
+      "loss": 0.1864,
+      "step": 9569
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0941063553037418,
+      "learning_rate": 1.0916895405142339e-05,
+      "loss": 0.1973,
+      "step": 9570
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.933533004750533,
+      "learning_rate": 1.09152553617099e-05,
+      "loss": 0.1759,
+      "step": 9571
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.116679573487381,
+      "learning_rate": 1.0913615293451105e-05,
+      "loss": 0.1764,
+      "step": 9572
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3262860299047816,
+      "learning_rate": 1.0911975200410445e-05,
+      "loss": 0.2024,
+      "step": 9573
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0543549657524613,
+      "learning_rate": 1.0910335082632406e-05,
+      "loss": 0.2182,
+      "step": 9574
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1044120478533699,
+      "learning_rate": 1.0908694940161477e-05,
+      "loss": 0.1796,
+      "step": 9575
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8634574691411527,
+      "learning_rate": 1.0907054773042148e-05,
+      "loss": 0.1866,
+      "step": 9576
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8953783284154978,
+      "learning_rate": 1.0905414581318902e-05,
+      "loss": 0.1815,
+      "step": 9577
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0168979420911626,
+      "learning_rate": 1.090377436503624e-05,
+      "loss": 0.1585,
+      "step": 9578
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3193192319875138,
+      "learning_rate": 1.0902134124238644e-05,
+      "loss": 0.1978,
+      "step": 9579
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9542646773015464,
+      "learning_rate": 1.090049385897061e-05,
+      "loss": 0.2019,
+      "step": 9580
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2211461636864696,
+      "learning_rate": 1.089885356927663e-05,
+      "loss": 0.1991,
+      "step": 9581
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4789441101008454,
+      "learning_rate": 1.0897213255201193e-05,
+      "loss": 0.1985,
+      "step": 9582
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2913650894369069,
+      "learning_rate": 1.0895572916788799e-05,
+      "loss": 0.1891,
+      "step": 9583
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8862058384341132,
+      "learning_rate": 1.089393255408394e-05,
+      "loss": 0.1904,
+      "step": 9584
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9609850270437007,
+      "learning_rate": 1.0892292167131107e-05,
+      "loss": 0.2003,
+      "step": 9585
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.068301573582356,
+      "learning_rate": 1.08906517559748e-05,
+      "loss": 0.1921,
+      "step": 9586
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9769987088158,
+      "learning_rate": 1.0889011320659513e-05,
+      "loss": 0.1679,
+      "step": 9587
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3124652154756535,
+      "learning_rate": 1.0887370861229744e-05,
+      "loss": 0.1971,
+      "step": 9588
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2458187192822827,
+      "learning_rate": 1.0885730377729993e-05,
+      "loss": 0.1789,
+      "step": 9589
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9722915506603313,
+      "learning_rate": 1.0884089870204751e-05,
+      "loss": 0.1861,
+      "step": 9590
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0798640423811054,
+      "learning_rate": 1.0882449338698521e-05,
+      "loss": 0.1671,
+      "step": 9591
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9778314506386806,
+      "learning_rate": 1.0880808783255808e-05,
+      "loss": 0.185,
+      "step": 9592
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1909601832238668,
+      "learning_rate": 1.0879168203921105e-05,
+      "loss": 0.1663,
+      "step": 9593
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.354804048495785,
+      "learning_rate": 1.0877527600738913e-05,
+      "loss": 0.2059,
+      "step": 9594
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2460384170334757,
+      "learning_rate": 1.0875886973753735e-05,
+      "loss": 0.1811,
+      "step": 9595
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4932185049543425,
+      "learning_rate": 1.0874246323010074e-05,
+      "loss": 0.2045,
+      "step": 9596
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.298711194215064,
+      "learning_rate": 1.0872605648552435e-05,
+      "loss": 0.1746,
+      "step": 9597
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.322696201426117,
+      "learning_rate": 1.0870964950425315e-05,
+      "loss": 0.1721,
+      "step": 9598
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0958001072962307,
+      "learning_rate": 1.0869324228673222e-05,
+      "loss": 0.1864,
+      "step": 9599
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.75974717231408,
+      "learning_rate": 1.086768348334066e-05,
+      "loss": 0.204,
+      "step": 9600
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.966607763752332,
+      "learning_rate": 1.0866042714472136e-05,
+      "loss": 0.2106,
+      "step": 9601
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0667990702956014,
+      "learning_rate": 1.0864401922112155e-05,
+      "loss": 0.1905,
+      "step": 9602
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9604517721025619,
+      "learning_rate": 1.0862761106305222e-05,
+      "loss": 0.1987,
+      "step": 9603
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0185280849062157,
+      "learning_rate": 1.0861120267095846e-05,
+      "loss": 0.2039,
+      "step": 9604
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1082943502475084,
+      "learning_rate": 1.0859479404528532e-05,
+      "loss": 0.1873,
+      "step": 9605
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1452453729289234,
+      "learning_rate": 1.0857838518647794e-05,
+      "loss": 0.2176,
+      "step": 9606
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.148710275207547,
+      "learning_rate": 1.0856197609498135e-05,
+      "loss": 0.1744,
+      "step": 9607
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.209646023184379,
+      "learning_rate": 1.0854556677124066e-05,
+      "loss": 0.1627,
+      "step": 9608
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1650388011788848,
+      "learning_rate": 1.08529157215701e-05,
+      "loss": 0.1956,
+      "step": 9609
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.157072205528815,
+      "learning_rate": 1.085127474288075e-05,
+      "loss": 0.1939,
+      "step": 9610
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.648764821811985,
+      "learning_rate": 1.0849633741100522e-05,
+      "loss": 0.1991,
+      "step": 9611
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0288598778364175,
+      "learning_rate": 1.084799271627393e-05,
+      "loss": 0.2053,
+      "step": 9612
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1278542729437095,
+      "learning_rate": 1.0846351668445489e-05,
+      "loss": 0.1999,
+      "step": 9613
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.457359559733764,
+      "learning_rate": 1.084471059765971e-05,
+      "loss": 0.2182,
+      "step": 9614
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.185454539256383,
+      "learning_rate": 1.0843069503961112e-05,
+      "loss": 0.1964,
+      "step": 9615
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1001727164465225,
+      "learning_rate": 1.0841428387394204e-05,
+      "loss": 0.1753,
+      "step": 9616
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.468981106103501,
+      "learning_rate": 1.0839787248003499e-05,
+      "loss": 0.2171,
+      "step": 9617
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2066098512363195,
+      "learning_rate": 1.0838146085833523e-05,
+      "loss": 0.1775,
+      "step": 9618
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3883135625315148,
+      "learning_rate": 1.0836504900928786e-05,
+      "loss": 0.1777,
+      "step": 9619
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1582429998119048,
+      "learning_rate": 1.0834863693333805e-05,
+      "loss": 0.1803,
+      "step": 9620
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.076354165273277,
+      "learning_rate": 1.08332224630931e-05,
+      "loss": 0.1923,
+      "step": 9621
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9869213870487541,
+      "learning_rate": 1.083158121025119e-05,
+      "loss": 0.1957,
+      "step": 9622
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4017559780579356,
+      "learning_rate": 1.082993993485259e-05,
+      "loss": 0.2127,
+      "step": 9623
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9471222448724727,
+      "learning_rate": 1.0828298636941826e-05,
+      "loss": 0.1712,
+      "step": 9624
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3935679546422635,
+      "learning_rate": 1.0826657316563412e-05,
+      "loss": 0.1822,
+      "step": 9625
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1490294573199775,
+      "learning_rate": 1.082501597376187e-05,
+      "loss": 0.19,
+      "step": 9626
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1764376635965907,
+      "learning_rate": 1.0823374608581727e-05,
+      "loss": 0.1984,
+      "step": 9627
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6176490098619052,
+      "learning_rate": 1.0821733221067499e-05,
+      "loss": 0.1924,
+      "step": 9628
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1701847629385655,
+      "learning_rate": 1.082009181126371e-05,
+      "loss": 0.1885,
+      "step": 9629
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4873077019656005,
+      "learning_rate": 1.0818450379214887e-05,
+      "loss": 0.1864,
+      "step": 9630
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0821228362386008,
+      "learning_rate": 1.081680892496555e-05,
+      "loss": 0.1884,
+      "step": 9631
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0979545228674528,
+      "learning_rate": 1.0815167448560225e-05,
+      "loss": 0.1843,
+      "step": 9632
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2455347475804985,
+      "learning_rate": 1.0813525950043435e-05,
+      "loss": 0.1813,
+      "step": 9633
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.978559329533385,
+      "learning_rate": 1.0811884429459708e-05,
+      "loss": 0.1883,
+      "step": 9634
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9707061312525472,
+      "learning_rate": 1.0810242886853572e-05,
+      "loss": 0.1923,
+      "step": 9635
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4835241358193412,
+      "learning_rate": 1.0808601322269553e-05,
+      "loss": 0.1855,
+      "step": 9636
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0247533098985855,
+      "learning_rate": 1.0806959735752174e-05,
+      "loss": 0.1938,
+      "step": 9637
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9173282273112893,
+      "learning_rate": 1.0805318127345968e-05,
+      "loss": 0.1907,
+      "step": 9638
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2942874387981254,
+      "learning_rate": 1.0803676497095463e-05,
+      "loss": 0.1993,
+      "step": 9639
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.103891088845872,
+      "learning_rate": 1.0802034845045189e-05,
+      "loss": 0.1806,
+      "step": 9640
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8294224003058488,
+      "learning_rate": 1.0800393171239672e-05,
+      "loss": 0.1798,
+      "step": 9641
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6864174989057856,
+      "learning_rate": 1.0798751475723446e-05,
+      "loss": 0.1904,
+      "step": 9642
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3092171275756168,
+      "learning_rate": 1.0797109758541038e-05,
+      "loss": 0.201,
+      "step": 9643
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9974374522277218,
+      "learning_rate": 1.0795468019736988e-05,
+      "loss": 0.2134,
+      "step": 9644
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2165807216297406,
+      "learning_rate": 1.079382625935582e-05,
+      "loss": 0.175,
+      "step": 9645
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9718849350489902,
+      "learning_rate": 1.0792184477442072e-05,
+      "loss": 0.1899,
+      "step": 9646
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4698470200530558,
+      "learning_rate": 1.079054267404027e-05,
+      "loss": 0.1963,
+      "step": 9647
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1466382693672115,
+      "learning_rate": 1.078890084919496e-05,
+      "loss": 0.181,
+      "step": 9648
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.25333634486384,
+      "learning_rate": 1.0787259002950665e-05,
+      "loss": 0.2133,
+      "step": 9649
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.012534747933004,
+      "learning_rate": 1.0785617135351927e-05,
+      "loss": 0.1924,
+      "step": 9650
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1474468277559202,
+      "learning_rate": 1.0783975246443281e-05,
+      "loss": 0.209,
+      "step": 9651
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9345414739937334,
+      "learning_rate": 1.078233333626926e-05,
+      "loss": 0.1717,
+      "step": 9652
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5527860447119741,
+      "learning_rate": 1.0780691404874404e-05,
+      "loss": 0.1913,
+      "step": 9653
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9515322340897505,
+      "learning_rate": 1.077904945230325e-05,
+      "loss": 0.1982,
+      "step": 9654
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9235823339102696,
+      "learning_rate": 1.0777407478600334e-05,
+      "loss": 0.2162,
+      "step": 9655
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7539908144802698,
+      "learning_rate": 1.0775765483810199e-05,
+      "loss": 0.1876,
+      "step": 9656
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9304968353210401,
+      "learning_rate": 1.0774123467977379e-05,
+      "loss": 0.1866,
+      "step": 9657
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2604049295529116,
+      "learning_rate": 1.077248143114642e-05,
+      "loss": 0.1925,
+      "step": 9658
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9924552531913771,
+      "learning_rate": 1.0770839373361854e-05,
+      "loss": 0.1633,
+      "step": 9659
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.364045262513066,
+      "learning_rate": 1.0769197294668228e-05,
+      "loss": 0.1889,
+      "step": 9660
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9236530495839659,
+      "learning_rate": 1.0767555195110082e-05,
+      "loss": 0.204,
+      "step": 9661
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0476753106491055,
+      "learning_rate": 1.0765913074731957e-05,
+      "loss": 0.175,
+      "step": 9662
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0026805236194896,
+      "learning_rate": 1.07642709335784e-05,
+      "loss": 0.1936,
+      "step": 9663
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9422222095805597,
+      "learning_rate": 1.0762628771693948e-05,
+      "loss": 0.2034,
+      "step": 9664
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9433156465199231,
+      "learning_rate": 1.0760986589123145e-05,
+      "loss": 0.2009,
+      "step": 9665
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2291693414472664,
+      "learning_rate": 1.0759344385910541e-05,
+      "loss": 0.1703,
+      "step": 9666
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0785063042674297,
+      "learning_rate": 1.0757702162100679e-05,
+      "loss": 0.1848,
+      "step": 9667
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.949233765859445,
+      "learning_rate": 1.0756059917738102e-05,
+      "loss": 0.1871,
+      "step": 9668
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0232043255187737,
+      "learning_rate": 1.0754417652867357e-05,
+      "loss": 0.1785,
+      "step": 9669
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9992934331116861,
+      "learning_rate": 1.0752775367532988e-05,
+      "loss": 0.2028,
+      "step": 9670
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4295122213944318,
+      "learning_rate": 1.0751133061779545e-05,
+      "loss": 0.1792,
+      "step": 9671
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.18761065388684,
+      "learning_rate": 1.074949073565158e-05,
+      "loss": 0.1577,
+      "step": 9672
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7470259053038004,
+      "learning_rate": 1.0747848389193633e-05,
+      "loss": 0.1993,
+      "step": 9673
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9904275822503751,
+      "learning_rate": 1.0746206022450256e-05,
+      "loss": 0.1931,
+      "step": 9674
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7834337706150454,
+      "learning_rate": 1.0744563635466e-05,
+      "loss": 0.1788,
+      "step": 9675
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7895005401380205,
+      "learning_rate": 1.0742921228285412e-05,
+      "loss": 0.1897,
+      "step": 9676
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.283230030469254,
+      "learning_rate": 1.0741278800953045e-05,
+      "loss": 0.2225,
+      "step": 9677
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.12162867780618,
+      "learning_rate": 1.0739636353513446e-05,
+      "loss": 0.1907,
+      "step": 9678
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0680028023000427,
+      "learning_rate": 1.0737993886011171e-05,
+      "loss": 0.2474,
+      "step": 9679
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8863078026342777,
+      "learning_rate": 1.0736351398490772e-05,
+      "loss": 0.1842,
+      "step": 9680
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1023494855573912,
+      "learning_rate": 1.0734708890996797e-05,
+      "loss": 0.1818,
+      "step": 9681
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9156047317575732,
+      "learning_rate": 1.0733066363573803e-05,
+      "loss": 0.1983,
+      "step": 9682
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.105766073750055,
+      "learning_rate": 1.073142381626634e-05,
+      "loss": 0.1703,
+      "step": 9683
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2012940202707465,
+      "learning_rate": 1.0729781249118966e-05,
+      "loss": 0.2002,
+      "step": 9684
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.162556841011124,
+      "learning_rate": 1.0728138662176237e-05,
+      "loss": 0.2084,
+      "step": 9685
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9256523144338665,
+      "learning_rate": 1.0726496055482705e-05,
+      "loss": 0.2,
+      "step": 9686
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9945108882251645,
+      "learning_rate": 1.0724853429082923e-05,
+      "loss": 0.187,
+      "step": 9687
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8823018088973252,
+      "learning_rate": 1.0723210783021454e-05,
+      "loss": 0.209,
+      "step": 9688
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0012462162192681,
+      "learning_rate": 1.072156811734285e-05,
+      "loss": 0.1664,
+      "step": 9689
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9969096264391915,
+      "learning_rate": 1.0719925432091671e-05,
+      "loss": 0.207,
+      "step": 9690
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8109754142384766,
+      "learning_rate": 1.0718282727312475e-05,
+      "loss": 0.1923,
+      "step": 9691
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2853994775292084,
+      "learning_rate": 1.0716640003049818e-05,
+      "loss": 0.1762,
+      "step": 9692
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1245990847029432,
+      "learning_rate": 1.0714997259348261e-05,
+      "loss": 0.1988,
+      "step": 9693
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8821098044855635,
+      "learning_rate": 1.0713354496252364e-05,
+      "loss": 0.1747,
+      "step": 9694
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3418193495044213,
+      "learning_rate": 1.0711711713806684e-05,
+      "loss": 0.2024,
+      "step": 9695
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.121844576126626,
+      "learning_rate": 1.0710068912055784e-05,
+      "loss": 0.1763,
+      "step": 9696
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.469846337002135,
+      "learning_rate": 1.0708426091044224e-05,
+      "loss": 0.1865,
+      "step": 9697
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.127361301705819,
+      "learning_rate": 1.0706783250816568e-05,
+      "loss": 0.1752,
+      "step": 9698
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.371137664811588,
+      "learning_rate": 1.0705140391417377e-05,
+      "loss": 0.1794,
+      "step": 9699
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9201500615344242,
+      "learning_rate": 1.070349751289121e-05,
+      "loss": 0.1792,
+      "step": 9700
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9205893271689509,
+      "learning_rate": 1.0701854615282635e-05,
+      "loss": 0.1859,
+      "step": 9701
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3951584715649386,
+      "learning_rate": 1.0700211698636214e-05,
+      "loss": 0.1859,
+      "step": 9702
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2693974342116394,
+      "learning_rate": 1.069856876299651e-05,
+      "loss": 0.1972,
+      "step": 9703
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9274877290317544,
+      "learning_rate": 1.0696925808408092e-05,
+      "loss": 0.1689,
+      "step": 9704
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2619010784277056,
+      "learning_rate": 1.0695282834915517e-05,
+      "loss": 0.1971,
+      "step": 9705
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1154770717828937,
+      "learning_rate": 1.069363984256336e-05,
+      "loss": 0.195,
+      "step": 9706
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8760721779934151,
+      "learning_rate": 1.0691996831396181e-05,
+      "loss": 0.1914,
+      "step": 9707
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1540281001625392,
+      "learning_rate": 1.0690353801458551e-05,
+      "loss": 0.2161,
+      "step": 9708
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0251076150045788,
+      "learning_rate": 1.0688710752795033e-05,
+      "loss": 0.1751,
+      "step": 9709
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8244986961316821,
+      "learning_rate": 1.0687067685450199e-05,
+      "loss": 0.1642,
+      "step": 9710
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.814014152897162,
+      "learning_rate": 1.0685424599468615e-05,
+      "loss": 0.181,
+      "step": 9711
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2640957549475418,
+      "learning_rate": 1.068378149489485e-05,
+      "loss": 0.1815,
+      "step": 9712
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8867530504394332,
+      "learning_rate": 1.068213837177347e-05,
+      "loss": 0.1818,
+      "step": 9713
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8642166809082094,
+      "learning_rate": 1.068049523014905e-05,
+      "loss": 0.1707,
+      "step": 9714
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.013203360223355,
+      "learning_rate": 1.067885207006616e-05,
+      "loss": 0.1836,
+      "step": 9715
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2534834176344625,
+      "learning_rate": 1.0677208891569366e-05,
+      "loss": 0.1988,
+      "step": 9716
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3915590820409383,
+      "learning_rate": 1.0675565694703248e-05,
+      "loss": 0.1985,
+      "step": 9717
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9889276453661713,
+      "learning_rate": 1.0673922479512366e-05,
+      "loss": 0.1814,
+      "step": 9718
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1748920823909086,
+      "learning_rate": 1.0672279246041301e-05,
+      "loss": 0.1878,
+      "step": 9719
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4392217921353778,
+      "learning_rate": 1.0670635994334626e-05,
+      "loss": 0.1757,
+      "step": 9720
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7471315556478759,
+      "learning_rate": 1.066899272443691e-05,
+      "loss": 0.1693,
+      "step": 9721
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1259472114454028,
+      "learning_rate": 1.0667349436392727e-05,
+      "loss": 0.1794,
+      "step": 9722
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1501304500378,
+      "learning_rate": 1.0665706130246654e-05,
+      "loss": 0.2131,
+      "step": 9723
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0239004877819975,
+      "learning_rate": 1.0664062806043266e-05,
+      "loss": 0.1724,
+      "step": 9724
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0414264832020326,
+      "learning_rate": 1.0662419463827136e-05,
+      "loss": 0.201,
+      "step": 9725
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3522003124483102,
+      "learning_rate": 1.066077610364284e-05,
+      "loss": 0.1844,
+      "step": 9726
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1639536779282773,
+      "learning_rate": 1.0659132725534958e-05,
+      "loss": 0.1931,
+      "step": 9727
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9436860780724251,
+      "learning_rate": 1.065748932954806e-05,
+      "loss": 0.1929,
+      "step": 9728
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.273126733844216,
+      "learning_rate": 1.0655845915726728e-05,
+      "loss": 0.1848,
+      "step": 9729
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2985323865422416,
+      "learning_rate": 1.065420248411554e-05,
+      "loss": 0.1935,
+      "step": 9730
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.039552057235806,
+      "learning_rate": 1.0652559034759069e-05,
+      "loss": 0.195,
+      "step": 9731
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9724179390813914,
+      "learning_rate": 1.0650915567701897e-05,
+      "loss": 0.1773,
+      "step": 9732
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.31261878897922,
+      "learning_rate": 1.0649272082988609e-05,
+      "loss": 0.2152,
+      "step": 9733
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0956625730563676,
+      "learning_rate": 1.0647628580663775e-05,
+      "loss": 0.1902,
+      "step": 9734
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9542691101221298,
+      "learning_rate": 1.0645985060771978e-05,
+      "loss": 0.1743,
+      "step": 9735
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1992225194268242,
+      "learning_rate": 1.0644341523357802e-05,
+      "loss": 0.1901,
+      "step": 9736
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8999673120258931,
+      "learning_rate": 1.0642697968465827e-05,
+      "loss": 0.1845,
+      "step": 9737
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0087462698811394,
+      "learning_rate": 1.0641054396140631e-05,
+      "loss": 0.1856,
+      "step": 9738
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8581094685422574,
+      "learning_rate": 1.06394108064268e-05,
+      "loss": 0.2202,
+      "step": 9739
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1336111707300922,
+      "learning_rate": 1.0637767199368911e-05,
+      "loss": 0.1673,
+      "step": 9740
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.814053132649836,
+      "learning_rate": 1.0636123575011555e-05,
+      "loss": 0.1984,
+      "step": 9741
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2560613817757404,
+      "learning_rate": 1.063447993339931e-05,
+      "loss": 0.1903,
+      "step": 9742
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1206558307710335,
+      "learning_rate": 1.0632836274576761e-05,
+      "loss": 0.1904,
+      "step": 9743
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7950750967837613,
+      "learning_rate": 1.0631192598588493e-05,
+      "loss": 0.1781,
+      "step": 9744
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7829207386631539,
+      "learning_rate": 1.062954890547909e-05,
+      "loss": 0.1681,
+      "step": 9745
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9252872638635755,
+      "learning_rate": 1.0627905195293135e-05,
+      "loss": 0.1822,
+      "step": 9746
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1785591768775445,
+      "learning_rate": 1.0626261468075218e-05,
+      "loss": 0.1863,
+      "step": 9747
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6476250713917159,
+      "learning_rate": 1.0624617723869921e-05,
+      "loss": 0.191,
+      "step": 9748
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0225638341945646,
+      "learning_rate": 1.0622973962721836e-05,
+      "loss": 0.1719,
+      "step": 9749
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2016612896236973,
+      "learning_rate": 1.0621330184675547e-05,
+      "loss": 0.202,
+      "step": 9750
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9244176686589038,
+      "learning_rate": 1.061968638977564e-05,
+      "loss": 0.192,
+      "step": 9751
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.185460765223053,
+      "learning_rate": 1.0618042578066707e-05,
+      "loss": 0.1725,
+      "step": 9752
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0514921292392578,
+      "learning_rate": 1.0616398749593331e-05,
+      "loss": 0.2072,
+      "step": 9753
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1355402752532482,
+      "learning_rate": 1.0614754904400105e-05,
+      "loss": 0.1859,
+      "step": 9754
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8676669752768543,
+      "learning_rate": 1.0613111042531618e-05,
+      "loss": 0.1521,
+      "step": 9755
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9336257746652679,
+      "learning_rate": 1.061146716403246e-05,
+      "loss": 0.1879,
+      "step": 9756
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2460212844758167,
+      "learning_rate": 1.0609823268947219e-05,
+      "loss": 0.2039,
+      "step": 9757
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.30362751558046,
+      "learning_rate": 1.0608179357320487e-05,
+      "loss": 0.2029,
+      "step": 9758
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1747846489666585,
+      "learning_rate": 1.0606535429196858e-05,
+      "loss": 0.2088,
+      "step": 9759
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9912826968645417,
+      "learning_rate": 1.060489148462092e-05,
+      "loss": 0.203,
+      "step": 9760
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.554812461103568,
+      "learning_rate": 1.0603247523637268e-05,
+      "loss": 0.1879,
+      "step": 9761
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1857395296609805,
+      "learning_rate": 1.0601603546290491e-05,
+      "loss": 0.1882,
+      "step": 9762
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0603146410098505,
+      "learning_rate": 1.0599959552625186e-05,
+      "loss": 0.1918,
+      "step": 9763
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.896561227828809,
+      "learning_rate": 1.0598315542685941e-05,
+      "loss": 0.1835,
+      "step": 9764
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9037701781341697,
+      "learning_rate": 1.0596671516517356e-05,
+      "loss": 0.1953,
+      "step": 9765
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0873238212270744,
+      "learning_rate": 1.059502747416402e-05,
+      "loss": 0.1801,
+      "step": 9766
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7932512597856124,
+      "learning_rate": 1.059338341567053e-05,
+      "loss": 0.1721,
+      "step": 9767
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9210138857139931,
+      "learning_rate": 1.0591739341081485e-05,
+      "loss": 0.1906,
+      "step": 9768
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.224460682832701,
+      "learning_rate": 1.0590095250441473e-05,
+      "loss": 0.181,
+      "step": 9769
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9226760646689801,
+      "learning_rate": 1.0588451143795093e-05,
+      "loss": 0.181,
+      "step": 9770
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8404512315313271,
+      "learning_rate": 1.0586807021186946e-05,
+      "loss": 0.1965,
+      "step": 9771
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7582090020659907,
+      "learning_rate": 1.0585162882661624e-05,
+      "loss": 0.1869,
+      "step": 9772
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8639559963409217,
+      "learning_rate": 1.0583518728263726e-05,
+      "loss": 0.1993,
+      "step": 9773
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8684827736925508,
+      "learning_rate": 1.058187455803785e-05,
+      "loss": 0.1904,
+      "step": 9774
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.543355549440279,
+      "learning_rate": 1.0580230372028593e-05,
+      "loss": 0.1883,
+      "step": 9775
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1117833572882903,
+      "learning_rate": 1.0578586170280554e-05,
+      "loss": 0.1982,
+      "step": 9776
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9471809713270811,
+      "learning_rate": 1.0576941952838334e-05,
+      "loss": 0.2013,
+      "step": 9777
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2033368512094136,
+      "learning_rate": 1.0575297719746533e-05,
+      "loss": 0.1684,
+      "step": 9778
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0747091930537156,
+      "learning_rate": 1.0573653471049745e-05,
+      "loss": 0.1869,
+      "step": 9779
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9710104972552251,
+      "learning_rate": 1.0572009206792575e-05,
+      "loss": 0.1758,
+      "step": 9780
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2635121948298544,
+      "learning_rate": 1.0570364927019623e-05,
+      "loss": 0.1872,
+      "step": 9781
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7568544286952503,
+      "learning_rate": 1.0568720631775491e-05,
+      "loss": 0.1767,
+      "step": 9782
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8659534324188721,
+      "learning_rate": 1.0567076321104776e-05,
+      "loss": 0.1758,
+      "step": 9783
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0559891149766256,
+      "learning_rate": 1.0565431995052089e-05,
+      "loss": 0.1908,
+      "step": 9784
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.057283250817933,
+      "learning_rate": 1.0563787653662025e-05,
+      "loss": 0.2087,
+      "step": 9785
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8860916224318578,
+      "learning_rate": 1.0562143296979188e-05,
+      "loss": 0.1736,
+      "step": 9786
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.859785432898717,
+      "learning_rate": 1.0560498925048186e-05,
+      "loss": 0.1838,
+      "step": 9787
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.385780047711686,
+      "learning_rate": 1.0558854537913614e-05,
+      "loss": 0.1905,
+      "step": 9788
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9099122713281608,
+      "learning_rate": 1.0557210135620084e-05,
+      "loss": 0.1958,
+      "step": 9789
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0201497510398987,
+      "learning_rate": 1.0555565718212198e-05,
+      "loss": 0.1751,
+      "step": 9790
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0445946424152273,
+      "learning_rate": 1.0553921285734559e-05,
+      "loss": 0.1961,
+      "step": 9791
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7918993501345337,
+      "learning_rate": 1.0552276838231773e-05,
+      "loss": 0.166,
+      "step": 9792
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9163235676527428,
+      "learning_rate": 1.0550632375748448e-05,
+      "loss": 0.1755,
+      "step": 9793
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0610958812797693,
+      "learning_rate": 1.0548987898329188e-05,
+      "loss": 0.2183,
+      "step": 9794
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1659132117607722,
+      "learning_rate": 1.0547343406018602e-05,
+      "loss": 0.2157,
+      "step": 9795
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.828123363788774,
+      "learning_rate": 1.0545698898861293e-05,
+      "loss": 0.1848,
+      "step": 9796
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8330494134593698,
+      "learning_rate": 1.0544054376901872e-05,
+      "loss": 0.1919,
+      "step": 9797
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9986865906064578,
+      "learning_rate": 1.0542409840184946e-05,
+      "loss": 0.1824,
+      "step": 9798
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.431778286746509,
+      "learning_rate": 1.0540765288755124e-05,
+      "loss": 0.171,
+      "step": 9799
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3598909405812594,
+      "learning_rate": 1.053912072265701e-05,
+      "loss": 0.174,
+      "step": 9800
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8439246989389632,
+      "learning_rate": 1.0537476141935215e-05,
+      "loss": 0.1882,
+      "step": 9801
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1056060930525182,
+      "learning_rate": 1.053583154663435e-05,
+      "loss": 0.1848,
+      "step": 9802
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8568831388430186,
+      "learning_rate": 1.0534186936799024e-05,
+      "loss": 0.19,
+      "step": 9803
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.881329670015039,
+      "learning_rate": 1.053254231247385e-05,
+      "loss": 0.1883,
+      "step": 9804
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8986303767656837,
+      "learning_rate": 1.0530897673703431e-05,
+      "loss": 0.1952,
+      "step": 9805
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.053070481720899,
+      "learning_rate": 1.0529253020532386e-05,
+      "loss": 0.2114,
+      "step": 9806
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8058192743448597,
+      "learning_rate": 1.0527608353005324e-05,
+      "loss": 0.1872,
+      "step": 9807
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0561154218791682,
+      "learning_rate": 1.0525963671166852e-05,
+      "loss": 0.1821,
+      "step": 9808
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.950268019639858,
+      "learning_rate": 1.0524318975061589e-05,
+      "loss": 0.1885,
+      "step": 9809
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0711119481698064,
+      "learning_rate": 1.0522674264734141e-05,
+      "loss": 0.1874,
+      "step": 9810
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4547965340786408,
+      "learning_rate": 1.0521029540229126e-05,
+      "loss": 0.1782,
+      "step": 9811
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8882553205810025,
+      "learning_rate": 1.0519384801591155e-05,
+      "loss": 0.1625,
+      "step": 9812
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9386707906152143,
+      "learning_rate": 1.0517740048864843e-05,
+      "loss": 0.2197,
+      "step": 9813
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.481738485759149,
+      "learning_rate": 1.05160952820948e-05,
+      "loss": 0.1855,
+      "step": 9814
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3158177722380133,
+      "learning_rate": 1.0514450501325646e-05,
+      "loss": 0.1794,
+      "step": 9815
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8984572182708918,
+      "learning_rate": 1.0512805706601994e-05,
+      "loss": 0.173,
+      "step": 9816
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2354073521381275,
+      "learning_rate": 1.0511160897968456e-05,
+      "loss": 0.1829,
+      "step": 9817
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2088642384857178,
+      "learning_rate": 1.0509516075469648e-05,
+      "loss": 0.1863,
+      "step": 9818
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9717684108168861,
+      "learning_rate": 1.0507871239150192e-05,
+      "loss": 0.1976,
+      "step": 9819
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1378371016917381,
+      "learning_rate": 1.0506226389054697e-05,
+      "loss": 0.1918,
+      "step": 9820
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1337727830864175,
+      "learning_rate": 1.0504581525227784e-05,
+      "loss": 0.1701,
+      "step": 9821
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9734523957238299,
+      "learning_rate": 1.0502936647714068e-05,
+      "loss": 0.1675,
+      "step": 9822
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8085856071065838,
+      "learning_rate": 1.0501291756558166e-05,
+      "loss": 0.2105,
+      "step": 9823
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3614299113261774,
+      "learning_rate": 1.0499646851804698e-05,
+      "loss": 0.1949,
+      "step": 9824
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0225480331788612,
+      "learning_rate": 1.049800193349828e-05,
+      "loss": 0.1803,
+      "step": 9825
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1063594496079652,
+      "learning_rate": 1.0496357001683535e-05,
+      "loss": 0.2012,
+      "step": 9826
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8650630444598459,
+      "learning_rate": 1.0494712056405077e-05,
+      "loss": 0.1975,
+      "step": 9827
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2071697382974003,
+      "learning_rate": 1.0493067097707521e-05,
+      "loss": 0.173,
+      "step": 9828
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8907759542377793,
+      "learning_rate": 1.0491422125635497e-05,
+      "loss": 0.1606,
+      "step": 9829
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1407485010753793,
+      "learning_rate": 1.0489777140233619e-05,
+      "loss": 0.2085,
+      "step": 9830
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2018164434360048,
+      "learning_rate": 1.048813214154651e-05,
+      "loss": 0.1777,
+      "step": 9831
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8739314910355729,
+      "learning_rate": 1.0486487129618787e-05,
+      "loss": 0.1964,
+      "step": 9832
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2347078940200111,
+      "learning_rate": 1.0484842104495077e-05,
+      "loss": 0.1826,
+      "step": 9833
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9880699813876799,
+      "learning_rate": 1.0483197066219994e-05,
+      "loss": 0.1739,
+      "step": 9834
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5484746841825274,
+      "learning_rate": 1.0481552014838164e-05,
+      "loss": 0.2093,
+      "step": 9835
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8828003341975913,
+      "learning_rate": 1.0479906950394205e-05,
+      "loss": 0.174,
+      "step": 9836
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8714267237009181,
+      "learning_rate": 1.0478261872932747e-05,
+      "loss": 0.1865,
+      "step": 9837
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9416663230472567,
+      "learning_rate": 1.0476616782498408e-05,
+      "loss": 0.1954,
+      "step": 9838
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8907966044512139,
+      "learning_rate": 1.0474971679135812e-05,
+      "loss": 0.18,
+      "step": 9839
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0428955435248508,
+      "learning_rate": 1.0473326562889583e-05,
+      "loss": 0.2001,
+      "step": 9840
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8931635815421982,
+      "learning_rate": 1.047168143380434e-05,
+      "loss": 0.2256,
+      "step": 9841
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.230855516835648,
+      "learning_rate": 1.0470036291924716e-05,
+      "loss": 0.1949,
+      "step": 9842
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9159697293692711,
+      "learning_rate": 1.046839113729533e-05,
+      "loss": 0.1648,
+      "step": 9843
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0587328554788622,
+      "learning_rate": 1.0466745969960808e-05,
+      "loss": 0.1693,
+      "step": 9844
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0236897380683918,
+      "learning_rate": 1.0465100789965774e-05,
+      "loss": 0.1741,
+      "step": 9845
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8364371860688766,
+      "learning_rate": 1.0463455597354857e-05,
+      "loss": 0.1745,
+      "step": 9846
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2590298353872946,
+      "learning_rate": 1.0461810392172678e-05,
+      "loss": 0.1973,
+      "step": 9847
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.55764474316235,
+      "learning_rate": 1.046016517446387e-05,
+      "loss": 0.2304,
+      "step": 9848
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3950511436078774,
+      "learning_rate": 1.0458519944273051e-05,
+      "loss": 0.2017,
+      "step": 9849
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.075205875836515,
+      "learning_rate": 1.0456874701644857e-05,
+      "loss": 0.1854,
+      "step": 9850
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1534536502690333,
+      "learning_rate": 1.0455229446623909e-05,
+      "loss": 0.188,
+      "step": 9851
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.105654838126189,
+      "learning_rate": 1.0453584179254837e-05,
+      "loss": 0.2011,
+      "step": 9852
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2759741028043035,
+      "learning_rate": 1.0451938899582263e-05,
+      "loss": 0.2026,
+      "step": 9853
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3135892552693105,
+      "learning_rate": 1.0450293607650828e-05,
+      "loss": 0.1803,
+      "step": 9854
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4361511876030737,
+      "learning_rate": 1.044864830350515e-05,
+      "loss": 0.1854,
+      "step": 9855
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7257367038897319,
+      "learning_rate": 1.0447002987189863e-05,
+      "loss": 0.1817,
+      "step": 9856
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0461617761421993,
+      "learning_rate": 1.0445357658749596e-05,
+      "loss": 0.1761,
+      "step": 9857
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9720243915830117,
+      "learning_rate": 1.0443712318228973e-05,
+      "loss": 0.1935,
+      "step": 9858
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.109979195338577,
+      "learning_rate": 1.044206696567263e-05,
+      "loss": 0.184,
+      "step": 9859
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.952114468244844,
+      "learning_rate": 1.0440421601125196e-05,
+      "loss": 0.1996,
+      "step": 9860
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.832335857365281,
+      "learning_rate": 1.0438776224631305e-05,
+      "loss": 0.1811,
+      "step": 9861
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.04682268647991,
+      "learning_rate": 1.043713083623558e-05,
+      "loss": 0.1811,
+      "step": 9862
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.700625011153412,
+      "learning_rate": 1.0435485435982655e-05,
+      "loss": 0.1839,
+      "step": 9863
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8833745933497529,
+      "learning_rate": 1.0433840023917166e-05,
+      "loss": 0.2041,
+      "step": 9864
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0260558036137535,
+      "learning_rate": 1.043219460008374e-05,
+      "loss": 0.1931,
+      "step": 9865
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.158223230321419,
+      "learning_rate": 1.0430549164527012e-05,
+      "loss": 0.1954,
+      "step": 9866
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.840651535607204,
+      "learning_rate": 1.0428903717291614e-05,
+      "loss": 0.1628,
+      "step": 9867
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8215057895386293,
+      "learning_rate": 1.0427258258422179e-05,
+      "loss": 0.1621,
+      "step": 9868
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9297959300807954,
+      "learning_rate": 1.042561278796334e-05,
+      "loss": 0.2035,
+      "step": 9869
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0375582587651102,
+      "learning_rate": 1.0423967305959727e-05,
+      "loss": 0.2245,
+      "step": 9870
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8712326630395105,
+      "learning_rate": 1.0422321812455977e-05,
+      "loss": 0.1941,
+      "step": 9871
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2359447371119525,
+      "learning_rate": 1.0420676307496727e-05,
+      "loss": 0.1969,
+      "step": 9872
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9577458563174123,
+      "learning_rate": 1.0419030791126608e-05,
+      "loss": 0.2073,
+      "step": 9873
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9015402021284342,
+      "learning_rate": 1.0417385263390253e-05,
+      "loss": 0.1743,
+      "step": 9874
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8146898383568317,
+      "learning_rate": 1.0415739724332301e-05,
+      "loss": 0.1706,
+      "step": 9875
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7583665868297698,
+      "learning_rate": 1.0414094173997382e-05,
+      "loss": 0.1644,
+      "step": 9876
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7536084868690823,
+      "learning_rate": 1.0412448612430139e-05,
+      "loss": 0.1932,
+      "step": 9877
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8179173485214218,
+      "learning_rate": 1.0410803039675203e-05,
+      "loss": 0.1711,
+      "step": 9878
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7474447497140159,
+      "learning_rate": 1.0409157455777212e-05,
+      "loss": 0.1742,
+      "step": 9879
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3573782407500192,
+      "learning_rate": 1.0407511860780798e-05,
+      "loss": 0.2061,
+      "step": 9880
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1693626671955326,
+      "learning_rate": 1.0405866254730607e-05,
+      "loss": 0.1809,
+      "step": 9881
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.121973631958896,
+      "learning_rate": 1.0404220637671269e-05,
+      "loss": 0.2006,
+      "step": 9882
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9838023394438137,
+      "learning_rate": 1.0402575009647427e-05,
+      "loss": 0.184,
+      "step": 9883
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8474963997760555,
+      "learning_rate": 1.040092937070371e-05,
+      "loss": 0.1793,
+      "step": 9884
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.436673796146176,
+      "learning_rate": 1.0399283720884761e-05,
+      "loss": 0.1649,
+      "step": 9885
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9966040144415655,
+      "learning_rate": 1.039763806023522e-05,
+      "loss": 0.1769,
+      "step": 9886
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6373730020474027,
+      "learning_rate": 1.0395992388799725e-05,
+      "loss": 0.1877,
+      "step": 9887
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.883731825593106,
+      "learning_rate": 1.0394346706622915e-05,
+      "loss": 0.1512,
+      "step": 9888
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.316877146392383,
+      "learning_rate": 1.0392701013749424e-05,
+      "loss": 0.211,
+      "step": 9889
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9359088500850488,
+      "learning_rate": 1.0391055310223899e-05,
+      "loss": 0.1746,
+      "step": 9890
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0825036335565394,
+      "learning_rate": 1.0389409596090975e-05,
+      "loss": 0.1844,
+      "step": 9891
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8675212267076646,
+      "learning_rate": 1.0387763871395298e-05,
+      "loss": 0.1681,
+      "step": 9892
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.252490369140081,
+      "learning_rate": 1.0386118136181498e-05,
+      "loss": 0.1764,
+      "step": 9893
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8267953666195103,
+      "learning_rate": 1.0384472390494225e-05,
+      "loss": 0.1704,
+      "step": 9894
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1896296090305618,
+      "learning_rate": 1.0382826634378115e-05,
+      "loss": 0.1895,
+      "step": 9895
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8499202270123422,
+      "learning_rate": 1.0381180867877813e-05,
+      "loss": 0.1709,
+      "step": 9896
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6666646990052674,
+      "learning_rate": 1.037953509103796e-05,
+      "loss": 0.1928,
+      "step": 9897
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2907050050197117,
+      "learning_rate": 1.037788930390319e-05,
+      "loss": 0.1838,
+      "step": 9898
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9548489444956723,
+      "learning_rate": 1.0376243506518157e-05,
+      "loss": 0.1835,
+      "step": 9899
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2427355712972166,
+      "learning_rate": 1.0374597698927496e-05,
+      "loss": 0.1887,
+      "step": 9900
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0923649933816408,
+      "learning_rate": 1.0372951881175854e-05,
+      "loss": 0.2256,
+      "step": 9901
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4007054267078818,
+      "learning_rate": 1.0371306053307866e-05,
+      "loss": 0.1744,
+      "step": 9902
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8349455748931837,
+      "learning_rate": 1.0369660215368182e-05,
+      "loss": 0.1902,
+      "step": 9903
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.160864993415897,
+      "learning_rate": 1.0368014367401447e-05,
+      "loss": 0.1748,
+      "step": 9904
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.100514323505627,
+      "learning_rate": 1.0366368509452298e-05,
+      "loss": 0.1795,
+      "step": 9905
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3266089140201034,
+      "learning_rate": 1.0364722641565381e-05,
+      "loss": 0.1917,
+      "step": 9906
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6110809904091798,
+      "learning_rate": 1.0363076763785345e-05,
+      "loss": 0.1889,
+      "step": 9907
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9414698726238333,
+      "learning_rate": 1.0361430876156831e-05,
+      "loss": 0.1756,
+      "step": 9908
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2947285498046557,
+      "learning_rate": 1.0359784978724483e-05,
+      "loss": 0.2214,
+      "step": 9909
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.014018441865324,
+      "learning_rate": 1.0358139071532949e-05,
+      "loss": 0.1785,
+      "step": 9910
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1017582642136696,
+      "learning_rate": 1.0356493154626868e-05,
+      "loss": 0.1778,
+      "step": 9911
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5116573330006802,
+      "learning_rate": 1.0354847228050895e-05,
+      "loss": 0.1851,
+      "step": 9912
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1288381460495436,
+      "learning_rate": 1.0353201291849668e-05,
+      "loss": 0.2002,
+      "step": 9913
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8981263323620646,
+      "learning_rate": 1.0351555346067836e-05,
+      "loss": 0.1919,
+      "step": 9914
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9193426516734726,
+      "learning_rate": 1.0349909390750046e-05,
+      "loss": 0.1826,
+      "step": 9915
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9227102536487921,
+      "learning_rate": 1.0348263425940945e-05,
+      "loss": 0.1691,
+      "step": 9916
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1041781716681,
+      "learning_rate": 1.034661745168518e-05,
+      "loss": 0.1606,
+      "step": 9917
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2902489109492576,
+      "learning_rate": 1.0344971468027397e-05,
+      "loss": 0.1863,
+      "step": 9918
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9399133950621475,
+      "learning_rate": 1.034332547501224e-05,
+      "loss": 0.1891,
+      "step": 9919
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2562837223830832,
+      "learning_rate": 1.034167947268436e-05,
+      "loss": 0.1842,
+      "step": 9920
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8729064282047535,
+      "learning_rate": 1.0340033461088408e-05,
+      "loss": 0.1671,
+      "step": 9921
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8439636188899999,
+      "learning_rate": 1.0338387440269029e-05,
+      "loss": 0.1943,
+      "step": 9922
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3023244038905142,
+      "learning_rate": 1.0336741410270872e-05,
+      "loss": 0.1724,
+      "step": 9923
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8552753076216781,
+      "learning_rate": 1.0335095371138582e-05,
+      "loss": 0.1997,
+      "step": 9924
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9254039051965883,
+      "learning_rate": 1.0333449322916812e-05,
+      "loss": 0.2074,
+      "step": 9925
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9900810976488515,
+      "learning_rate": 1.0331803265650212e-05,
+      "loss": 0.1736,
+      "step": 9926
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2135277999895162,
+      "learning_rate": 1.0330157199383428e-05,
+      "loss": 0.1773,
+      "step": 9927
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1341813241407668,
+      "learning_rate": 1.0328511124161111e-05,
+      "loss": 0.1741,
+      "step": 9928
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0080976235152903,
+      "learning_rate": 1.0326865040027914e-05,
+      "loss": 0.1836,
+      "step": 9929
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9764000986623692,
+      "learning_rate": 1.0325218947028483e-05,
+      "loss": 0.2,
+      "step": 9930
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0610741734776916,
+      "learning_rate": 1.032357284520747e-05,
+      "loss": 0.2032,
+      "step": 9931
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.168931745220606,
+      "learning_rate": 1.0321926734609525e-05,
+      "loss": 0.1873,
+      "step": 9932
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2369314036165708,
+      "learning_rate": 1.0320280615279297e-05,
+      "loss": 0.1899,
+      "step": 9933
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7069155450925078,
+      "learning_rate": 1.031863448726144e-05,
+      "loss": 0.1916,
+      "step": 9934
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1104411774982685,
+      "learning_rate": 1.0316988350600608e-05,
+      "loss": 0.1847,
+      "step": 9935
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.295191707646736,
+      "learning_rate": 1.0315342205341448e-05,
+      "loss": 0.2246,
+      "step": 9936
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9595382596028376,
+      "learning_rate": 1.031369605152861e-05,
+      "loss": 0.1805,
+      "step": 9937
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3007792169988164,
+      "learning_rate": 1.031204988920675e-05,
+      "loss": 0.1858,
+      "step": 9938
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3047610591721563,
+      "learning_rate": 1.031040371842052e-05,
+      "loss": 0.1952,
+      "step": 9939
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.094211112565609,
+      "learning_rate": 1.0308757539214573e-05,
+      "loss": 0.1874,
+      "step": 9940
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.831448138141734,
+      "learning_rate": 1.0307111351633556e-05,
+      "loss": 0.1948,
+      "step": 9941
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1926966402724417,
+      "learning_rate": 1.030546515572213e-05,
+      "loss": 0.1982,
+      "step": 9942
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4148961719797137,
+      "learning_rate": 1.0303818951524941e-05,
+      "loss": 0.1655,
+      "step": 9943
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1868873442922836,
+      "learning_rate": 1.0302172739086651e-05,
+      "loss": 0.193,
+      "step": 9944
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0994668253712192,
+      "learning_rate": 1.0300526518451906e-05,
+      "loss": 0.2119,
+      "step": 9945
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7749159270480126,
+      "learning_rate": 1.0298880289665359e-05,
+      "loss": 0.1655,
+      "step": 9946
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2181962641100348,
+      "learning_rate": 1.029723405277167e-05,
+      "loss": 0.1956,
+      "step": 9947
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3402104542718751,
+      "learning_rate": 1.029558780781549e-05,
+      "loss": 0.1954,
+      "step": 9948
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0498104781581756,
+      "learning_rate": 1.0293941554841475e-05,
+      "loss": 0.1621,
+      "step": 9949
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8955502875619714,
+      "learning_rate": 1.0292295293894279e-05,
+      "loss": 0.1815,
+      "step": 9950
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4261451905653821,
+      "learning_rate": 1.0290649025018553e-05,
+      "loss": 0.1885,
+      "step": 9951
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1050297705764032,
+      "learning_rate": 1.0289002748258961e-05,
+      "loss": 0.1985,
+      "step": 9952
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9725277146089406,
+      "learning_rate": 1.0287356463660152e-05,
+      "loss": 0.1828,
+      "step": 9953
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2985263149088095,
+      "learning_rate": 1.0285710171266778e-05,
+      "loss": 0.1933,
+      "step": 9954
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8226489022264653,
+      "learning_rate": 1.0284063871123504e-05,
+      "loss": 0.1784,
+      "step": 9955
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5280430717726388,
+      "learning_rate": 1.0282417563274982e-05,
+      "loss": 0.1858,
+      "step": 9956
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4197035876160895,
+      "learning_rate": 1.0280771247765865e-05,
+      "loss": 0.1906,
+      "step": 9957
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1898992638116423,
+      "learning_rate": 1.0279124924640813e-05,
+      "loss": 0.1845,
+      "step": 9958
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.301609927287,
+      "learning_rate": 1.027747859394448e-05,
+      "loss": 0.1846,
+      "step": 9959
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.03226421368799,
+      "learning_rate": 1.0275832255721527e-05,
+      "loss": 0.1687,
+      "step": 9960
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1475592463345812,
+      "learning_rate": 1.0274185910016608e-05,
+      "loss": 0.2076,
+      "step": 9961
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.857957466398941,
+      "learning_rate": 1.0272539556874381e-05,
+      "loss": 0.1826,
+      "step": 9962
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8451749170398627,
+      "learning_rate": 1.0270893196339499e-05,
+      "loss": 0.1783,
+      "step": 9963
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0211113370636866,
+      "learning_rate": 1.026924682845663e-05,
+      "loss": 0.1852,
+      "step": 9964
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3551642322630637,
+      "learning_rate": 1.0267600453270422e-05,
+      "loss": 0.1822,
+      "step": 9965
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9282330658954546,
+      "learning_rate": 1.0265954070825536e-05,
+      "loss": 0.1837,
+      "step": 9966
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.555358636297375,
+      "learning_rate": 1.0264307681166634e-05,
+      "loss": 0.1748,
+      "step": 9967
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5972130118339416,
+      "learning_rate": 1.0262661284338367e-05,
+      "loss": 0.1762,
+      "step": 9968
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.001103276739541,
+      "learning_rate": 1.02610148803854e-05,
+      "loss": 0.1952,
+      "step": 9969
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1031313216687382,
+      "learning_rate": 1.025936846935239e-05,
+      "loss": 0.2099,
+      "step": 9970
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2314792299673958,
+      "learning_rate": 1.0257722051283998e-05,
+      "loss": 0.1816,
+      "step": 9971
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.061943164826423,
+      "learning_rate": 1.0256075626224876e-05,
+      "loss": 0.1715,
+      "step": 9972
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9307765106478149,
+      "learning_rate": 1.0254429194219694e-05,
+      "loss": 0.1957,
+      "step": 9973
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9256565304503027,
+      "learning_rate": 1.02527827553131e-05,
+      "loss": 0.2083,
+      "step": 9974
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4219555269572766,
+      "learning_rate": 1.0251136309549764e-05,
+      "loss": 0.1665,
+      "step": 9975
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1394062648630405,
+      "learning_rate": 1.0249489856974335e-05,
+      "loss": 0.1743,
+      "step": 9976
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8884799606782805,
+      "learning_rate": 1.0247843397631485e-05,
+      "loss": 0.2081,
+      "step": 9977
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0280662741511235,
+      "learning_rate": 1.0246196931565869e-05,
+      "loss": 0.1822,
+      "step": 9978
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1389930656547747,
+      "learning_rate": 1.0244550458822145e-05,
+      "loss": 0.1817,
+      "step": 9979
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8778846451344755,
+      "learning_rate": 1.0242903979444976e-05,
+      "loss": 0.1585,
+      "step": 9980
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9142638715956067,
+      "learning_rate": 1.0241257493479022e-05,
+      "loss": 0.2058,
+      "step": 9981
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.18773644917128,
+      "learning_rate": 1.0239611000968948e-05,
+      "loss": 0.2521,
+      "step": 9982
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0356077541986568,
+      "learning_rate": 1.023796450195941e-05,
+      "loss": 0.1737,
+      "step": 9983
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1727486639573697,
+      "learning_rate": 1.0236317996495074e-05,
+      "loss": 0.1765,
+      "step": 9984
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9419796126948489,
+      "learning_rate": 1.0234671484620595e-05,
+      "loss": 0.1902,
+      "step": 9985
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7801012010400616,
+      "learning_rate": 1.023302496638064e-05,
+      "loss": 0.1929,
+      "step": 9986
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.013663718762519,
+      "learning_rate": 1.023137844181987e-05,
+      "loss": 0.1936,
+      "step": 9987
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.386473113559732,
+      "learning_rate": 1.022973191098295e-05,
+      "loss": 0.1757,
+      "step": 9988
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9898436861099714,
+      "learning_rate": 1.0228085373914534e-05,
+      "loss": 0.1833,
+      "step": 9989
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8509169739451241,
+      "learning_rate": 1.022643883065929e-05,
+      "loss": 0.1759,
+      "step": 9990
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8572529282667672,
+      "learning_rate": 1.0224792281261883e-05,
+      "loss": 0.1923,
+      "step": 9991
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.52936893410723,
+      "learning_rate": 1.0223145725766972e-05,
+      "loss": 0.215,
+      "step": 9992
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7402502741709267,
+      "learning_rate": 1.022149916421922e-05,
+      "loss": 0.1867,
+      "step": 9993
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4300609361053354,
+      "learning_rate": 1.0219852596663287e-05,
+      "loss": 0.1933,
+      "step": 9994
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0778884603823187,
+      "learning_rate": 1.0218206023143843e-05,
+      "loss": 0.1901,
+      "step": 9995
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0155397155807473,
+      "learning_rate": 1.0216559443705549e-05,
+      "loss": 0.1654,
+      "step": 9996
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2284066582722302,
+      "learning_rate": 1.0214912858393069e-05,
+      "loss": 0.182,
+      "step": 9997
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0145416710485446,
+      "learning_rate": 1.0213266267251063e-05,
+      "loss": 0.1879,
+      "step": 9998
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.19001894105872,
+      "learning_rate": 1.0211619670324196e-05,
+      "loss": 0.1792,
+      "step": 9999
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.970080459098825,
+      "learning_rate": 1.0209973067657138e-05,
+      "loss": 0.1814,
+      "step": 10000
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0951474234469627,
+      "learning_rate": 1.0208326459294544e-05,
+      "loss": 0.189,
+      "step": 10001
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8388440522398718,
+      "learning_rate": 1.0206679845281086e-05,
+      "loss": 0.2007,
+      "step": 10002
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0514825151087455,
+      "learning_rate": 1.0205033225661425e-05,
+      "loss": 0.1738,
+      "step": 10003
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9795950508541672,
+      "learning_rate": 1.0203386600480225e-05,
+      "loss": 0.1931,
+      "step": 10004
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.335975532830077,
+      "learning_rate": 1.0201739969782154e-05,
+      "loss": 0.1905,
+      "step": 10005
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9444491493237841,
+      "learning_rate": 1.0200093333611877e-05,
+      "loss": 0.202,
+      "step": 10006
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8810868878895413,
+      "learning_rate": 1.0198446692014052e-05,
+      "loss": 0.1995,
+      "step": 10007
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9449853646973742,
+      "learning_rate": 1.019680004503335e-05,
+      "loss": 0.1927,
+      "step": 10008
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.285770560807395,
+      "learning_rate": 1.0195153392714439e-05,
+      "loss": 0.2011,
+      "step": 10009
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4409911905410049,
+      "learning_rate": 1.019350673510198e-05,
+      "loss": 0.1609,
+      "step": 10010
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1450798008075653,
+      "learning_rate": 1.0191860072240638e-05,
+      "loss": 0.1909,
+      "step": 10011
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2757295834893168,
+      "learning_rate": 1.019021340417508e-05,
+      "loss": 0.1918,
+      "step": 10012
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9584243701572146,
+      "learning_rate": 1.0188566730949977e-05,
+      "loss": 0.1855,
+      "step": 10013
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1818591305464108,
+      "learning_rate": 1.0186920052609988e-05,
+      "loss": 0.1811,
+      "step": 10014
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3379051075950539,
+      "learning_rate": 1.0185273369199781e-05,
+      "loss": 0.2011,
+      "step": 10015
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7753559912686392,
+      "learning_rate": 1.0183626680764023e-05,
+      "loss": 0.1962,
+      "step": 10016
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9930898004506734,
+      "learning_rate": 1.0181979987347383e-05,
+      "loss": 0.178,
+      "step": 10017
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9906395873333926,
+      "learning_rate": 1.0180333288994526e-05,
+      "loss": 0.1679,
+      "step": 10018
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1841086133641656,
+      "learning_rate": 1.0178686585750117e-05,
+      "loss": 0.2037,
+      "step": 10019
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9662665276637551,
+      "learning_rate": 1.0177039877658825e-05,
+      "loss": 0.1944,
+      "step": 10020
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.092277086848292,
+      "learning_rate": 1.0175393164765315e-05,
+      "loss": 0.1799,
+      "step": 10021
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9675325735824981,
+      "learning_rate": 1.0173746447114257e-05,
+      "loss": 0.1645,
+      "step": 10022
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1363807412961018,
+      "learning_rate": 1.017209972475032e-05,
+      "loss": 0.1859,
+      "step": 10023
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1286346321271528,
+      "learning_rate": 1.0170452997718161e-05,
+      "loss": 0.1827,
+      "step": 10024
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0493532996425716,
+      "learning_rate": 1.0168806266062459e-05,
+      "loss": 0.1824,
+      "step": 10025
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5678128545444037,
+      "learning_rate": 1.0167159529827876e-05,
+      "loss": 0.1938,
+      "step": 10026
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9157850060664193,
+      "learning_rate": 1.0165512789059084e-05,
+      "loss": 0.2014,
+      "step": 10027
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1309156120661967,
+      "learning_rate": 1.0163866043800748e-05,
+      "loss": 0.2063,
+      "step": 10028
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.25804053030035,
+      "learning_rate": 1.0162219294097531e-05,
+      "loss": 0.1726,
+      "step": 10029
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3146522587286733,
+      "learning_rate": 1.0160572539994111e-05,
+      "loss": 0.1736,
+      "step": 10030
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.140177136353688,
+      "learning_rate": 1.015892578153515e-05,
+      "loss": 0.1942,
+      "step": 10031
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9510605440053564,
+      "learning_rate": 1.015727901876532e-05,
+      "loss": 0.1731,
+      "step": 10032
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9945450978518271,
+      "learning_rate": 1.0155632251729289e-05,
+      "loss": 0.2,
+      "step": 10033
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.894224695698637,
+      "learning_rate": 1.015398548047172e-05,
+      "loss": 0.1977,
+      "step": 10034
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.142454133947765,
+      "learning_rate": 1.0152338705037288e-05,
+      "loss": 0.1869,
+      "step": 10035
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.976074582797507,
+      "learning_rate": 1.0150691925470661e-05,
+      "loss": 0.1678,
+      "step": 10036
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.110742325414474,
+      "learning_rate": 1.0149045141816507e-05,
+      "loss": 0.1991,
+      "step": 10037
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5073131663612216,
+      "learning_rate": 1.0147398354119493e-05,
+      "loss": 0.2022,
+      "step": 10038
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1136933742481399,
+      "learning_rate": 1.0145751562424293e-05,
+      "loss": 0.1838,
+      "step": 10039
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8396311826471067,
+      "learning_rate": 1.0144104766775574e-05,
+      "loss": 0.1648,
+      "step": 10040
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5889142219711996,
+      "learning_rate": 1.0142457967218004e-05,
+      "loss": 0.1527,
+      "step": 10041
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5526245240475558,
+      "learning_rate": 1.0140811163796251e-05,
+      "loss": 0.1834,
+      "step": 10042
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4435183341553837,
+      "learning_rate": 1.0139164356554991e-05,
+      "loss": 0.18,
+      "step": 10043
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8552306532750134,
+      "learning_rate": 1.0137517545538889e-05,
+      "loss": 0.2122,
+      "step": 10044
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0819198266646604,
+      "learning_rate": 1.0135870730792614e-05,
+      "loss": 0.1571,
+      "step": 10045
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2645867726469782,
+      "learning_rate": 1.0134223912360841e-05,
+      "loss": 0.1843,
+      "step": 10046
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9493612997745103,
+      "learning_rate": 1.013257709028823e-05,
+      "loss": 0.1805,
+      "step": 10047
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9482335058184064,
+      "learning_rate": 1.0130930264619464e-05,
+      "loss": 0.1744,
+      "step": 10048
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0100789781159738,
+      "learning_rate": 1.0129283435399209e-05,
+      "loss": 0.2131,
+      "step": 10049
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.210160492224197,
+      "learning_rate": 1.0127636602672129e-05,
+      "loss": 0.1801,
+      "step": 10050
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0121260518070379,
+      "learning_rate": 1.01259897664829e-05,
+      "loss": 0.1851,
+      "step": 10051
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0048530916729317,
+      "learning_rate": 1.0124342926876191e-05,
+      "loss": 0.193,
+      "step": 10052
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5781470869665752,
+      "learning_rate": 1.0122696083896675e-05,
+      "loss": 0.1884,
+      "step": 10053
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3276050210602262,
+      "learning_rate": 1.012104923758902e-05,
+      "loss": 0.178,
+      "step": 10054
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7500154198232776,
+      "learning_rate": 1.0119402387997896e-05,
+      "loss": 0.1751,
+      "step": 10055
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1186441647606769,
+      "learning_rate": 1.0117755535167976e-05,
+      "loss": 0.1866,
+      "step": 10056
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0804238144692095,
+      "learning_rate": 1.0116108679143932e-05,
+      "loss": 0.2126,
+      "step": 10057
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8679185570128787,
+      "learning_rate": 1.0114461819970435e-05,
+      "loss": 0.1635,
+      "step": 10058
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.012499264568714,
+      "learning_rate": 1.0112814957692151e-05,
+      "loss": 0.1922,
+      "step": 10059
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2950140590211783,
+      "learning_rate": 1.0111168092353755e-05,
+      "loss": 0.1999,
+      "step": 10060
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4073424409836586,
+      "learning_rate": 1.010952122399992e-05,
+      "loss": 0.1851,
+      "step": 10061
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9568381123647998,
+      "learning_rate": 1.0107874352675318e-05,
+      "loss": 0.1821,
+      "step": 10062
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.262395675575379,
+      "learning_rate": 1.0106227478424616e-05,
+      "loss": 0.2034,
+      "step": 10063
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9072944589352504,
+      "learning_rate": 1.0104580601292484e-05,
+      "loss": 0.1892,
+      "step": 10064
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9383135405528775,
+      "learning_rate": 1.01029337213236e-05,
+      "loss": 0.1723,
+      "step": 10065
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0190912617277366,
+      "learning_rate": 1.0101286838562634e-05,
+      "loss": 0.1561,
+      "step": 10066
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0876522637918589,
+      "learning_rate": 1.0099639953054256e-05,
+      "loss": 0.2164,
+      "step": 10067
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1657087864947426,
+      "learning_rate": 1.0097993064843138e-05,
+      "loss": 0.1862,
+      "step": 10068
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.273406700391757,
+      "learning_rate": 1.0096346173973951e-05,
+      "loss": 0.2154,
+      "step": 10069
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1349999528339405,
+      "learning_rate": 1.0094699280491371e-05,
+      "loss": 0.1691,
+      "step": 10070
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8161335476433862,
+      "learning_rate": 1.009305238444007e-05,
+      "loss": 0.1859,
+      "step": 10071
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1468136125885018,
+      "learning_rate": 1.0091405485864714e-05,
+      "loss": 0.2021,
+      "step": 10072
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5859126246115993,
+      "learning_rate": 1.008975858480998e-05,
+      "loss": 0.1969,
+      "step": 10073
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.07772020244796,
+      "learning_rate": 1.0088111681320539e-05,
+      "loss": 0.17,
+      "step": 10074
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.367254327526161,
+      "learning_rate": 1.0086464775441064e-05,
+      "loss": 0.1844,
+      "step": 10075
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5431365797665315,
+      "learning_rate": 1.008481786721623e-05,
+      "loss": 0.2131,
+      "step": 10076
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2134258991202571,
+      "learning_rate": 1.0083170956690702e-05,
+      "loss": 0.1739,
+      "step": 10077
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1734516351523943,
+      "learning_rate": 1.008152404390916e-05,
+      "loss": 0.2067,
+      "step": 10078
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3435819692760993,
+      "learning_rate": 1.0079877128916274e-05,
+      "loss": 0.1844,
+      "step": 10079
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1947122193617454,
+      "learning_rate": 1.0078230211756714e-05,
+      "loss": 0.2014,
+      "step": 10080
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1594082609931562,
+      "learning_rate": 1.0076583292475157e-05,
+      "loss": 0.1865,
+      "step": 10081
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0904335945044985,
+      "learning_rate": 1.007493637111627e-05,
+      "loss": 0.1975,
+      "step": 10082
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9743470552569026,
+      "learning_rate": 1.0073289447724735e-05,
+      "loss": 0.202,
+      "step": 10083
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9636925739935388,
+      "learning_rate": 1.0071642522345217e-05,
+      "loss": 0.1914,
+      "step": 10084
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1185462840749154,
+      "learning_rate": 1.0069995595022393e-05,
+      "loss": 0.1728,
+      "step": 10085
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2161215893688821,
+      "learning_rate": 1.006834866580093e-05,
+      "loss": 0.1995,
+      "step": 10086
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5347187922548848,
+      "learning_rate": 1.006670173472551e-05,
+      "loss": 0.195,
+      "step": 10087
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8850629354087978,
+      "learning_rate": 1.00650548018408e-05,
+      "loss": 0.1853,
+      "step": 10088
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2386088666481518,
+      "learning_rate": 1.0063407867191478e-05,
+      "loss": 0.2176,
+      "step": 10089
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.797387745492623,
+      "learning_rate": 1.006176093082221e-05,
+      "loss": 0.1852,
+      "step": 10090
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.150179340161504,
+      "learning_rate": 1.0060113992777674e-05,
+      "loss": 0.1699,
+      "step": 10091
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0130362875047425,
+      "learning_rate": 1.0058467053102544e-05,
+      "loss": 0.2271,
+      "step": 10092
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.491247780734237,
+      "learning_rate": 1.0056820111841495e-05,
+      "loss": 0.1868,
+      "step": 10093
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9510571415665885,
+      "learning_rate": 1.0055173169039192e-05,
+      "loss": 0.2148,
+      "step": 10094
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4031686577019125,
+      "learning_rate": 1.0053526224740313e-05,
+      "loss": 0.1746,
+      "step": 10095
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.783505650323615,
+      "learning_rate": 1.0051879278989536e-05,
+      "loss": 0.1739,
+      "step": 10096
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7477764660744681,
+      "learning_rate": 1.0050232331831528e-05,
+      "loss": 0.1789,
+      "step": 10097
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8995038115576395,
+      "learning_rate": 1.0048585383310967e-05,
+      "loss": 0.2128,
+      "step": 10098
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8440069820748402,
+      "learning_rate": 1.0046938433472522e-05,
+      "loss": 0.1614,
+      "step": 10099
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0316875891218211,
+      "learning_rate": 1.0045291482360871e-05,
+      "loss": 0.1923,
+      "step": 10100
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3853684601406886,
+      "learning_rate": 1.0043644530020686e-05,
+      "loss": 0.1897,
+      "step": 10101
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0621006982820664,
+      "learning_rate": 1.0041997576496643e-05,
+      "loss": 0.1831,
+      "step": 10102
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0929803339188837,
+      "learning_rate": 1.004035062183341e-05,
+      "loss": 0.1871,
+      "step": 10103
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5969371607412692,
+      "learning_rate": 1.0038703666075665e-05,
+      "loss": 0.1802,
+      "step": 10104
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2339497000753874,
+      "learning_rate": 1.003705670926808e-05,
+      "loss": 0.1894,
+      "step": 10105
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.408984839071695,
+      "learning_rate": 1.0035409751455332e-05,
+      "loss": 0.1927,
+      "step": 10106
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9577016281366301,
+      "learning_rate": 1.0033762792682092e-05,
+      "loss": 0.1802,
+      "step": 10107
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0593468957036303,
+      "learning_rate": 1.0032115832993032e-05,
+      "loss": 0.1736,
+      "step": 10108
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1893297903859361,
+      "learning_rate": 1.003046887243283e-05,
+      "loss": 0.1927,
+      "step": 10109
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7841198800976856,
+      "learning_rate": 1.0028821911046158e-05,
+      "loss": 0.1733,
+      "step": 10110
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9635994815974888,
+      "learning_rate": 1.0027174948877692e-05,
+      "loss": 0.1769,
+      "step": 10111
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4976506517758972,
+      "learning_rate": 1.0025527985972102e-05,
+      "loss": 0.1903,
+      "step": 10112
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9258277904885904,
+      "learning_rate": 1.0023881022374062e-05,
+      "loss": 0.181,
+      "step": 10113
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1322482761758057,
+      "learning_rate": 1.0022234058128251e-05,
+      "loss": 0.185,
+      "step": 10114
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.697952043975562,
+      "learning_rate": 1.0020587093279339e-05,
+      "loss": 0.1945,
+      "step": 10115
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0221196595404018,
+      "learning_rate": 1.0018940127872001e-05,
+      "loss": 0.173,
+      "step": 10116
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.525650165385635,
+      "learning_rate": 1.001729316195091e-05,
+      "loss": 0.196,
+      "step": 10117
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9281390262973999,
+      "learning_rate": 1.001564619556074e-05,
+      "loss": 0.1758,
+      "step": 10118
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5021408002983778,
+      "learning_rate": 1.001399922874617e-05,
+      "loss": 0.2149,
+      "step": 10119
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.338027793563955,
+      "learning_rate": 1.0012352261551868e-05,
+      "loss": 0.1987,
+      "step": 10120
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.034500217978134,
+      "learning_rate": 1.001070529402251e-05,
+      "loss": 0.1678,
+      "step": 10121
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9757847773691674,
+      "learning_rate": 1.0009058326202768e-05,
+      "loss": 0.1821,
+      "step": 10122
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9283672375689633,
+      "learning_rate": 1.000741135813732e-05,
+      "loss": 0.2065,
+      "step": 10123
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0360793799716828,
+      "learning_rate": 1.000576438987084e-05,
+      "loss": 0.188,
+      "step": 10124
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9095239168298347,
+      "learning_rate": 1.0004117421448e-05,
+      "loss": 0.1953,
+      "step": 10125
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0856311188231922,
+      "learning_rate": 1.0002470452913473e-05,
+      "loss": 0.184,
+      "step": 10126
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0522283892013424,
+      "learning_rate": 1.0000823484311937e-05,
+      "loss": 0.1797,
+      "step": 10127
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8729040385760526,
+      "learning_rate": 9.999176515688066e-06,
+      "loss": 0.1783,
+      "step": 10128
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5443301439658077,
+      "learning_rate": 9.997529547086527e-06,
+      "loss": 0.1928,
+      "step": 10129
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0450069976328333,
+      "learning_rate": 9.995882578552002e-06,
+      "loss": 0.1769,
+      "step": 10130
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.594677740582822,
+      "learning_rate": 9.99423561012916e-06,
+      "loss": 0.1859,
+      "step": 10131
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9298369881994603,
+      "learning_rate": 9.992588641862682e-06,
+      "loss": 0.183,
+      "step": 10132
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1359824341091553,
+      "learning_rate": 9.990941673797234e-06,
+      "loss": 0.2131,
+      "step": 10133
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1726229751635113,
+      "learning_rate": 9.989294705977494e-06,
+      "loss": 0.2187,
+      "step": 10134
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0697105168059207,
+      "learning_rate": 9.987647738448134e-06,
+      "loss": 0.1792,
+      "step": 10135
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0446155555654533,
+      "learning_rate": 9.986000771253835e-06,
+      "loss": 0.1842,
+      "step": 10136
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5252893644381011,
+      "learning_rate": 9.984353804439264e-06,
+      "loss": 0.2099,
+      "step": 10137
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.991401095277735,
+      "learning_rate": 9.982706838049094e-06,
+      "loss": 0.1812,
+      "step": 10138
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9456998564950463,
+      "learning_rate": 9.981059872128004e-06,
+      "loss": 0.1749,
+      "step": 10139
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.141972747906058,
+      "learning_rate": 9.979412906720663e-06,
+      "loss": 0.1747,
+      "step": 10140
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9011930047903997,
+      "learning_rate": 9.977765941871754e-06,
+      "loss": 0.1748,
+      "step": 10141
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.135559019401279,
+      "learning_rate": 9.976118977625941e-06,
+      "loss": 0.1878,
+      "step": 10142
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4727396649384408,
+      "learning_rate": 9.974472014027903e-06,
+      "loss": 0.1737,
+      "step": 10143
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8580717686639748,
+      "learning_rate": 9.97282505112231e-06,
+      "loss": 0.21,
+      "step": 10144
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.09660598116125,
+      "learning_rate": 9.971178088953845e-06,
+      "loss": 0.1681,
+      "step": 10145
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8839707672338685,
+      "learning_rate": 9.969531127567172e-06,
+      "loss": 0.1792,
+      "step": 10146
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.970215974749392,
+      "learning_rate": 9.96788416700697e-06,
+      "loss": 0.177,
+      "step": 10147
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.26868888668654,
+      "learning_rate": 9.96623720731791e-06,
+      "loss": 0.1784,
+      "step": 10148
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.656729429434053,
+      "learning_rate": 9.964590248544671e-06,
+      "loss": 0.2014,
+      "step": 10149
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.93392320181895,
+      "learning_rate": 9.96294329073192e-06,
+      "loss": 0.1887,
+      "step": 10150
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8589463763497676,
+      "learning_rate": 9.961296333924338e-06,
+      "loss": 0.1678,
+      "step": 10151
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0833683363223623,
+      "learning_rate": 9.959649378166593e-06,
+      "loss": 0.1936,
+      "step": 10152
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9254068495035073,
+      "learning_rate": 9.95800242350336e-06,
+      "loss": 0.1791,
+      "step": 10153
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2301348039871198,
+      "learning_rate": 9.95635546997932e-06,
+      "loss": 0.1916,
+      "step": 10154
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8856938637562976,
+      "learning_rate": 9.95470851763913e-06,
+      "loss": 0.1649,
+      "step": 10155
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.540285019478901,
+      "learning_rate": 9.953061566527481e-06,
+      "loss": 0.1785,
+      "step": 10156
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9488229772162609,
+      "learning_rate": 9.951414616689037e-06,
+      "loss": 0.1971,
+      "step": 10157
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4195672508913053,
+      "learning_rate": 9.949767668168477e-06,
+      "loss": 0.1716,
+      "step": 10158
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9390328203362874,
+      "learning_rate": 9.948120721010467e-06,
+      "loss": 0.1847,
+      "step": 10159
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9321828642929739,
+      "learning_rate": 9.94647377525969e-06,
+      "loss": 0.1757,
+      "step": 10160
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4279965604940652,
+      "learning_rate": 9.94482683096081e-06,
+      "loss": 0.201,
+      "step": 10161
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0650861327001324,
+      "learning_rate": 9.943179888158512e-06,
+      "loss": 0.1886,
+      "step": 10162
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.259419552091082,
+      "learning_rate": 9.941532946897456e-06,
+      "loss": 0.1867,
+      "step": 10163
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0238322413879728,
+      "learning_rate": 9.93988600722233e-06,
+      "loss": 0.1822,
+      "step": 10164
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9864417236171614,
+      "learning_rate": 9.938239069177792e-06,
+      "loss": 0.1855,
+      "step": 10165
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8899000559027274,
+      "learning_rate": 9.936592132808526e-06,
+      "loss": 0.1858,
+      "step": 10166
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0665160725544522,
+      "learning_rate": 9.9349451981592e-06,
+      "loss": 0.1728,
+      "step": 10167
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0003805091844538,
+      "learning_rate": 9.933298265274493e-06,
+      "loss": 0.1756,
+      "step": 10168
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7563077112028692,
+      "learning_rate": 9.93165133419907e-06,
+      "loss": 0.1651,
+      "step": 10169
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9976909236680422,
+      "learning_rate": 9.93000440497761e-06,
+      "loss": 0.2197,
+      "step": 10170
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0635056769523115,
+      "learning_rate": 9.928357477654783e-06,
+      "loss": 0.1755,
+      "step": 10171
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9026074106361105,
+      "learning_rate": 9.926710552275268e-06,
+      "loss": 0.1995,
+      "step": 10172
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8673749012582641,
+      "learning_rate": 9.925063628883731e-06,
+      "loss": 0.1817,
+      "step": 10173
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9998411376931166,
+      "learning_rate": 9.923416707524845e-06,
+      "loss": 0.205,
+      "step": 10174
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.334736412034857,
+      "learning_rate": 9.921769788243291e-06,
+      "loss": 0.1721,
+      "step": 10175
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.853685111169905,
+      "learning_rate": 9.92012287108373e-06,
+      "loss": 0.1825,
+      "step": 10176
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5061052824171897,
+      "learning_rate": 9.918475956090845e-06,
+      "loss": 0.1907,
+      "step": 10177
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.649723100823968,
+      "learning_rate": 9.9168290433093e-06,
+      "loss": 0.1581,
+      "step": 10178
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9177755922432216,
+      "learning_rate": 9.915182132783773e-06,
+      "loss": 0.1698,
+      "step": 10179
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.171004729759711,
+      "learning_rate": 9.913535224558936e-06,
+      "loss": 0.1862,
+      "step": 10180
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1815972737473774,
+      "learning_rate": 9.911888318679463e-06,
+      "loss": 0.2017,
+      "step": 10181
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9401379104944637,
+      "learning_rate": 9.910241415190022e-06,
+      "loss": 0.1664,
+      "step": 10182
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.086417801036447,
+      "learning_rate": 9.908594514135288e-06,
+      "loss": 0.2047,
+      "step": 10183
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1246348855935053,
+      "learning_rate": 9.906947615559932e-06,
+      "loss": 0.182,
+      "step": 10184
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5114210520282993,
+      "learning_rate": 9.90530071950863e-06,
+      "loss": 0.2235,
+      "step": 10185
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7888776269891093,
+      "learning_rate": 9.903653826026049e-06,
+      "loss": 0.1705,
+      "step": 10186
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3258985075992884,
+      "learning_rate": 9.902006935156863e-06,
+      "loss": 0.1849,
+      "step": 10187
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.954590283479778,
+      "learning_rate": 9.900360046945746e-06,
+      "loss": 0.184,
+      "step": 10188
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9657271524800654,
+      "learning_rate": 9.89871316143737e-06,
+      "loss": 0.1725,
+      "step": 10189
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.910098204363335,
+      "learning_rate": 9.897066278676405e-06,
+      "loss": 0.2054,
+      "step": 10190
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.980882661006259,
+      "learning_rate": 9.89541939870752e-06,
+      "loss": 0.169,
+      "step": 10191
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3256139950041805,
+      "learning_rate": 9.893772521575391e-06,
+      "loss": 0.1984,
+      "step": 10192
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2102877874991573,
+      "learning_rate": 9.892125647324686e-06,
+      "loss": 0.1972,
+      "step": 10193
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2624640998421532,
+      "learning_rate": 9.890478776000084e-06,
+      "loss": 0.1975,
+      "step": 10194
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6185245911520465,
+      "learning_rate": 9.888831907646246e-06,
+      "loss": 0.1702,
+      "step": 10195
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3108016567585763,
+      "learning_rate": 9.887185042307852e-06,
+      "loss": 0.1929,
+      "step": 10196
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1164020898665021,
+      "learning_rate": 9.885538180029568e-06,
+      "loss": 0.1819,
+      "step": 10197
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3538769640540862,
+      "learning_rate": 9.883891320856071e-06,
+      "loss": 0.2121,
+      "step": 10198
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.724159505917361,
+      "learning_rate": 9.882244464832026e-06,
+      "loss": 0.1805,
+      "step": 10199
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2116009207840774,
+      "learning_rate": 9.880597612002106e-06,
+      "loss": 0.1754,
+      "step": 10200
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0692218955637767,
+      "learning_rate": 9.878950762410981e-06,
+      "loss": 0.1719,
+      "step": 10201
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1384005316524284,
+      "learning_rate": 9.877303916103328e-06,
+      "loss": 0.2003,
+      "step": 10202
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7804404480331262,
+      "learning_rate": 9.87565707312381e-06,
+      "loss": 0.1939,
+      "step": 10203
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1253730710614271,
+      "learning_rate": 9.874010233517103e-06,
+      "loss": 0.1829,
+      "step": 10204
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0966039460231016,
+      "learning_rate": 9.872363397327873e-06,
+      "loss": 0.1832,
+      "step": 10205
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9516794217543744,
+      "learning_rate": 9.870716564600796e-06,
+      "loss": 0.1649,
+      "step": 10206
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4001621379783895,
+      "learning_rate": 9.869069735380539e-06,
+      "loss": 0.2184,
+      "step": 10207
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9562249033797356,
+      "learning_rate": 9.86742290971177e-06,
+      "loss": 0.1564,
+      "step": 10208
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1783580116509145,
+      "learning_rate": 9.865776087639166e-06,
+      "loss": 0.1931,
+      "step": 10209
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9452027253631481,
+      "learning_rate": 9.864129269207388e-06,
+      "loss": 0.1644,
+      "step": 10210
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7783206095949607,
+      "learning_rate": 9.862482454461116e-06,
+      "loss": 0.1727,
+      "step": 10211
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2040093366723683,
+      "learning_rate": 9.860835643445012e-06,
+      "loss": 0.1946,
+      "step": 10212
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0919123989202537,
+      "learning_rate": 9.85918883620375e-06,
+      "loss": 0.1927,
+      "step": 10213
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2210304943276447,
+      "learning_rate": 9.857542032781998e-06,
+      "loss": 0.1658,
+      "step": 10214
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.403600247600213,
+      "learning_rate": 9.855895233224431e-06,
+      "loss": 0.1764,
+      "step": 10215
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6703789557349276,
+      "learning_rate": 9.854248437575709e-06,
+      "loss": 0.1786,
+      "step": 10216
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2725040333955278,
+      "learning_rate": 9.852601645880509e-06,
+      "loss": 0.1944,
+      "step": 10217
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2734027828974808,
+      "learning_rate": 9.850954858183496e-06,
+      "loss": 0.1755,
+      "step": 10218
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5556163072912323,
+      "learning_rate": 9.84930807452934e-06,
+      "loss": 0.2188,
+      "step": 10219
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8223797292168257,
+      "learning_rate": 9.847661294962712e-06,
+      "loss": 0.1837,
+      "step": 10220
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9725737606009888,
+      "learning_rate": 9.846014519528284e-06,
+      "loss": 0.1765,
+      "step": 10221
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9305960044566103,
+      "learning_rate": 9.844367748270715e-06,
+      "loss": 0.1787,
+      "step": 10222
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9975408276525014,
+      "learning_rate": 9.842720981234682e-06,
+      "loss": 0.1891,
+      "step": 10223
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8328704239311866,
+      "learning_rate": 9.841074218464852e-06,
+      "loss": 0.2045,
+      "step": 10224
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8011822959442132,
+      "learning_rate": 9.839427460005892e-06,
+      "loss": 0.1606,
+      "step": 10225
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1040336320983504,
+      "learning_rate": 9.83778070590247e-06,
+      "loss": 0.1841,
+      "step": 10226
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6536014897279836,
+      "learning_rate": 9.836133956199256e-06,
+      "loss": 0.2025,
+      "step": 10227
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1487464384278228,
+      "learning_rate": 9.834487210940921e-06,
+      "loss": 0.1803,
+      "step": 10228
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.167115620085271,
+      "learning_rate": 9.832840470172125e-06,
+      "loss": 0.169,
+      "step": 10229
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1368296793900705,
+      "learning_rate": 9.831193733937546e-06,
+      "loss": 0.1966,
+      "step": 10230
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.248366853893879,
+      "learning_rate": 9.829547002281842e-06,
+      "loss": 0.1962,
+      "step": 10231
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2296604069185608,
+      "learning_rate": 9.827900275249686e-06,
+      "loss": 0.1698,
+      "step": 10232
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0772480915238765,
+      "learning_rate": 9.826253552885744e-06,
+      "loss": 0.1702,
+      "step": 10233
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1237664542437968,
+      "learning_rate": 9.824606835234689e-06,
+      "loss": 0.1807,
+      "step": 10234
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1427706105579902,
+      "learning_rate": 9.822960122341178e-06,
+      "loss": 0.162,
+      "step": 10235
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1936009566155101,
+      "learning_rate": 9.821313414249885e-06,
+      "loss": 0.1701,
+      "step": 10236
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.858311831063239,
+      "learning_rate": 9.819666711005475e-06,
+      "loss": 0.1671,
+      "step": 10237
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9974203917111358,
+      "learning_rate": 9.818020012652619e-06,
+      "loss": 0.1792,
+      "step": 10238
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.048164162370106,
+      "learning_rate": 9.816373319235978e-06,
+      "loss": 0.1773,
+      "step": 10239
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1626993038494857,
+      "learning_rate": 9.81472663080022e-06,
+      "loss": 0.2068,
+      "step": 10240
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1444195275217797,
+      "learning_rate": 9.813079947390014e-06,
+      "loss": 0.176,
+      "step": 10241
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.142794849853422,
+      "learning_rate": 9.811433269050028e-06,
+      "loss": 0.1936,
+      "step": 10242
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.73344991210697,
+      "learning_rate": 9.809786595824922e-06,
+      "loss": 0.1756,
+      "step": 10243
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9409173951911256,
+      "learning_rate": 9.808139927759363e-06,
+      "loss": 0.1816,
+      "step": 10244
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0067367796209783,
+      "learning_rate": 9.806493264898025e-06,
+      "loss": 0.1731,
+      "step": 10245
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0266266779009294,
+      "learning_rate": 9.804846607285564e-06,
+      "loss": 0.1842,
+      "step": 10246
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1987923485278358,
+      "learning_rate": 9.803199954966653e-06,
+      "loss": 0.1718,
+      "step": 10247
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6993453303103312,
+      "learning_rate": 9.801553307985951e-06,
+      "loss": 0.1612,
+      "step": 10248
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1169837375895257,
+      "learning_rate": 9.799906666388129e-06,
+      "loss": 0.2179,
+      "step": 10249
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2233213875080469,
+      "learning_rate": 9.798260030217846e-06,
+      "loss": 0.1806,
+      "step": 10250
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8535516549948462,
+      "learning_rate": 9.796613399519777e-06,
+      "loss": 0.1985,
+      "step": 10251
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.042384382904542,
+      "learning_rate": 9.794966774338576e-06,
+      "loss": 0.1733,
+      "step": 10252
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2915414484488394,
+      "learning_rate": 9.793320154718916e-06,
+      "loss": 0.2085,
+      "step": 10253
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0400857471966334,
+      "learning_rate": 9.791673540705455e-06,
+      "loss": 0.187,
+      "step": 10254
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0354907674864533,
+      "learning_rate": 9.790026932342867e-06,
+      "loss": 0.1905,
+      "step": 10255
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.156361371396123,
+      "learning_rate": 9.788380329675804e-06,
+      "loss": 0.1737,
+      "step": 10256
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.130435978970632,
+      "learning_rate": 9.78673373274894e-06,
+      "loss": 0.1822,
+      "step": 10257
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.005638120384302,
+      "learning_rate": 9.785087141606933e-06,
+      "loss": 0.1739,
+      "step": 10258
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1380722013661921,
+      "learning_rate": 9.783440556294453e-06,
+      "loss": 0.1632,
+      "step": 10259
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9394729324116381,
+      "learning_rate": 9.781793976856162e-06,
+      "loss": 0.1935,
+      "step": 10260
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0403813908247483,
+      "learning_rate": 9.780147403336715e-06,
+      "loss": 0.1996,
+      "step": 10261
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0395031212258359,
+      "learning_rate": 9.778500835780787e-06,
+      "loss": 0.1701,
+      "step": 10262
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4688923849207338,
+      "learning_rate": 9.776854274233033e-06,
+      "loss": 0.216,
+      "step": 10263
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9303142358262178,
+      "learning_rate": 9.775207718738122e-06,
+      "loss": 0.1751,
+      "step": 10264
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9180433229028818,
+      "learning_rate": 9.773561169340711e-06,
+      "loss": 0.1701,
+      "step": 10265
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.3860890328689495,
+      "learning_rate": 9.771914626085469e-06,
+      "loss": 0.182,
+      "step": 10266
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9648158711671928,
+      "learning_rate": 9.770268089017053e-06,
+      "loss": 0.1796,
+      "step": 10267
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2075022464405123,
+      "learning_rate": 9.768621558180132e-06,
+      "loss": 0.1843,
+      "step": 10268
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0368552296049038,
+      "learning_rate": 9.766975033619361e-06,
+      "loss": 0.2178,
+      "step": 10269
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9632831147646977,
+      "learning_rate": 9.765328515379407e-06,
+      "loss": 0.192,
+      "step": 10270
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0815652516872132,
+      "learning_rate": 9.763682003504928e-06,
+      "loss": 0.2028,
+      "step": 10271
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0663126245248298,
+      "learning_rate": 9.762035498040594e-06,
+      "loss": 0.1806,
+      "step": 10272
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0079893330839749,
+      "learning_rate": 9.760388999031052e-06,
+      "loss": 0.1645,
+      "step": 10273
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8802844478953326,
+      "learning_rate": 9.758742506520981e-06,
+      "loss": 0.2002,
+      "step": 10274
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0872710320559473,
+      "learning_rate": 9.757096020555026e-06,
+      "loss": 0.1872,
+      "step": 10275
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3758322155268397,
+      "learning_rate": 9.755449541177858e-06,
+      "loss": 0.1975,
+      "step": 10276
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8157296114345972,
+      "learning_rate": 9.753803068434138e-06,
+      "loss": 0.1757,
+      "step": 10277
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8569053998828822,
+      "learning_rate": 9.752156602368518e-06,
+      "loss": 0.1921,
+      "step": 10278
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.086718255160991,
+      "learning_rate": 9.750510143025667e-06,
+      "loss": 0.183,
+      "step": 10279
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8547794002110998,
+      "learning_rate": 9.74886369045024e-06,
+      "loss": 0.1925,
+      "step": 10280
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9958419750913367,
+      "learning_rate": 9.747217244686904e-06,
+      "loss": 0.1836,
+      "step": 10281
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9492194733753996,
+      "learning_rate": 9.745570805780312e-06,
+      "loss": 0.1913,
+      "step": 10282
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0321429448228026,
+      "learning_rate": 9.743924373775125e-06,
+      "loss": 0.1981,
+      "step": 10283
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2903368325471654,
+      "learning_rate": 9.742277948716004e-06,
+      "loss": 0.1851,
+      "step": 10284
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9680758698483248,
+      "learning_rate": 9.740631530647611e-06,
+      "loss": 0.1756,
+      "step": 10285
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1418268403416956,
+      "learning_rate": 9.7389851196146e-06,
+      "loss": 0.1651,
+      "step": 10286
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0744824585428467,
+      "learning_rate": 9.737338715661635e-06,
+      "loss": 0.1815,
+      "step": 10287
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9499836138535693,
+      "learning_rate": 9.735692318833368e-06,
+      "loss": 0.1643,
+      "step": 10288
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.033330852504397,
+      "learning_rate": 9.734045929174465e-06,
+      "loss": 0.1745,
+      "step": 10289
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9317581030504224,
+      "learning_rate": 9.732399546729578e-06,
+      "loss": 0.181,
+      "step": 10290
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.936147738733036,
+      "learning_rate": 9.730753171543374e-06,
+      "loss": 0.1947,
+      "step": 10291
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4341538894034087,
+      "learning_rate": 9.729106803660501e-06,
+      "loss": 0.1622,
+      "step": 10292
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0503544146668324,
+      "learning_rate": 9.727460443125622e-06,
+      "loss": 0.1868,
+      "step": 10293
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9132717870250818,
+      "learning_rate": 9.725814089983398e-06,
+      "loss": 0.1837,
+      "step": 10294
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2074972029352926,
+      "learning_rate": 9.724167744278475e-06,
+      "loss": 0.1876,
+      "step": 10295
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.166491148740018,
+      "learning_rate": 9.722521406055521e-06,
+      "loss": 0.2105,
+      "step": 10296
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1476688881483905,
+      "learning_rate": 9.720875075359188e-06,
+      "loss": 0.201,
+      "step": 10297
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0005395496337184,
+      "learning_rate": 9.71922875223414e-06,
+      "loss": 0.1653,
+      "step": 10298
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.045700555276377,
+      "learning_rate": 9.717582436725021e-06,
+      "loss": 0.1933,
+      "step": 10299
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1122857347006354,
+      "learning_rate": 9.715936128876501e-06,
+      "loss": 0.2016,
+      "step": 10300
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8552952996571777,
+      "learning_rate": 9.714289828733223e-06,
+      "loss": 0.1983,
+      "step": 10301
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9366334544382264,
+      "learning_rate": 9.712643536339853e-06,
+      "loss": 0.1673,
+      "step": 10302
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8599452455983897,
+      "learning_rate": 9.71099725174104e-06,
+      "loss": 0.177,
+      "step": 10303
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4254728191793864,
+      "learning_rate": 9.709350974981449e-06,
+      "loss": 0.1746,
+      "step": 10304
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2572506240427659,
+      "learning_rate": 9.707704706105724e-06,
+      "loss": 0.2239,
+      "step": 10305
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8240594846353297,
+      "learning_rate": 9.706058445158527e-06,
+      "loss": 0.162,
+      "step": 10306
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3421053745298535,
+      "learning_rate": 9.704412192184511e-06,
+      "loss": 0.1903,
+      "step": 10307
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.707243750628533,
+      "learning_rate": 9.702765947228333e-06,
+      "loss": 0.1605,
+      "step": 10308
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4382999861183035,
+      "learning_rate": 9.701119710334641e-06,
+      "loss": 0.2078,
+      "step": 10309
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9283704283606212,
+      "learning_rate": 9.699473481548097e-06,
+      "loss": 0.2208,
+      "step": 10310
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9548298747791736,
+      "learning_rate": 9.69782726091335e-06,
+      "loss": 0.185,
+      "step": 10311
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2026902730032376,
+      "learning_rate": 9.69618104847506e-06,
+      "loss": 0.2108,
+      "step": 10312
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2489949232747564,
+      "learning_rate": 9.694534844277876e-06,
+      "loss": 0.1893,
+      "step": 10313
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9972214462379292,
+      "learning_rate": 9.692888648366447e-06,
+      "loss": 0.21,
+      "step": 10314
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1213588878823344,
+      "learning_rate": 9.691242460785433e-06,
+      "loss": 0.1822,
+      "step": 10315
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9237740150032636,
+      "learning_rate": 9.689596281579481e-06,
+      "loss": 0.1859,
+      "step": 10316
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0459244960203853,
+      "learning_rate": 9.687950110793254e-06,
+      "loss": 0.1757,
+      "step": 10317
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9527148976547666,
+      "learning_rate": 9.686303948471393e-06,
+      "loss": 0.1902,
+      "step": 10318
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.399354081487811,
+      "learning_rate": 9.684657794658557e-06,
+      "loss": 0.1742,
+      "step": 10319
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2006200307552037,
+      "learning_rate": 9.683011649399393e-06,
+      "loss": 0.1801,
+      "step": 10320
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8273043295743692,
+      "learning_rate": 9.681365512738561e-06,
+      "loss": 0.2077,
+      "step": 10321
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9695511459683863,
+      "learning_rate": 9.679719384720705e-06,
+      "loss": 0.1791,
+      "step": 10322
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4961861053556897,
+      "learning_rate": 9.678073265390478e-06,
+      "loss": 0.1728,
+      "step": 10323
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1475693874767454,
+      "learning_rate": 9.676427154792532e-06,
+      "loss": 0.1927,
+      "step": 10324
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1376021528687463,
+      "learning_rate": 9.67478105297152e-06,
+      "loss": 0.1805,
+      "step": 10325
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4852948187095072,
+      "learning_rate": 9.673134959972087e-06,
+      "loss": 0.1786,
+      "step": 10326
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3038353556527462,
+      "learning_rate": 9.671488875838892e-06,
+      "loss": 0.1658,
+      "step": 10327
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2454552539516948,
+      "learning_rate": 9.669842800616573e-06,
+      "loss": 0.1997,
+      "step": 10328
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1513530264433791,
+      "learning_rate": 9.66819673434979e-06,
+      "loss": 0.176,
+      "step": 10329
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0618655964419974,
+      "learning_rate": 9.666550677083193e-06,
+      "loss": 0.1856,
+      "step": 10330
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8207853176957545,
+      "learning_rate": 9.664904628861423e-06,
+      "loss": 0.165,
+      "step": 10331
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4835936664284424,
+      "learning_rate": 9.663258589729133e-06,
+      "loss": 0.1818,
+      "step": 10332
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.148356495795242,
+      "learning_rate": 9.661612559730974e-06,
+      "loss": 0.1886,
+      "step": 10333
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.977431153226433,
+      "learning_rate": 9.659966538911597e-06,
+      "loss": 0.1764,
+      "step": 10334
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3184864457409595,
+      "learning_rate": 9.658320527315642e-06,
+      "loss": 0.1875,
+      "step": 10335
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0881433063606767,
+      "learning_rate": 9.656674524987764e-06,
+      "loss": 0.1858,
+      "step": 10336
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9988224231568393,
+      "learning_rate": 9.655028531972607e-06,
+      "loss": 0.2006,
+      "step": 10337
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0873306012414692,
+      "learning_rate": 9.653382548314824e-06,
+      "loss": 0.1902,
+      "step": 10338
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6293296419133951,
+      "learning_rate": 9.651736574059056e-06,
+      "loss": 0.1736,
+      "step": 10339
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9422063308430143,
+      "learning_rate": 9.650090609249957e-06,
+      "loss": 0.1897,
+      "step": 10340
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.098181699439507,
+      "learning_rate": 9.648444653932166e-06,
+      "loss": 0.2025,
+      "step": 10341
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.043494769008611,
+      "learning_rate": 9.646798708150335e-06,
+      "loss": 0.1811,
+      "step": 10342
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4141343490160387,
+      "learning_rate": 9.645152771949107e-06,
+      "loss": 0.1732,
+      "step": 10343
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9833433273445842,
+      "learning_rate": 9.643506845373134e-06,
+      "loss": 0.1736,
+      "step": 10344
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4078488952573975,
+      "learning_rate": 9.641860928467054e-06,
+      "loss": 0.1871,
+      "step": 10345
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.254584100159465,
+      "learning_rate": 9.64021502127552e-06,
+      "loss": 0.2344,
+      "step": 10346
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9609613365024053,
+      "learning_rate": 9.638569123843174e-06,
+      "loss": 0.1756,
+      "step": 10347
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.863047354823533,
+      "learning_rate": 9.636923236214658e-06,
+      "loss": 0.1634,
+      "step": 10348
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9746403408105337,
+      "learning_rate": 9.635277358434622e-06,
+      "loss": 0.191,
+      "step": 10349
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9928457396463447,
+      "learning_rate": 9.633631490547705e-06,
+      "loss": 0.1771,
+      "step": 10350
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1389667439345696,
+      "learning_rate": 9.63198563259856e-06,
+      "loss": 0.1779,
+      "step": 10351
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8675214941337062,
+      "learning_rate": 9.63033978463182e-06,
+      "loss": 0.1908,
+      "step": 10352
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8474049487338804,
+      "learning_rate": 9.628693946692137e-06,
+      "loss": 0.1869,
+      "step": 10353
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9039438034001152,
+      "learning_rate": 9.62704811882415e-06,
+      "loss": 0.198,
+      "step": 10354
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2251513400924823,
+      "learning_rate": 9.625402301072508e-06,
+      "loss": 0.1837,
+      "step": 10355
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7946568640432952,
+      "learning_rate": 9.623756493481845e-06,
+      "loss": 0.169,
+      "step": 10356
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0808439840321342,
+      "learning_rate": 9.622110696096812e-06,
+      "loss": 0.2229,
+      "step": 10357
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1578573480507568,
+      "learning_rate": 9.620464908962044e-06,
+      "loss": 0.2092,
+      "step": 10358
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2304337056025627,
+      "learning_rate": 9.618819132122188e-06,
+      "loss": 0.1985,
+      "step": 10359
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9198798137925053,
+      "learning_rate": 9.617173365621885e-06,
+      "loss": 0.2165,
+      "step": 10360
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3088003472110072,
+      "learning_rate": 9.615527609505778e-06,
+      "loss": 0.2178,
+      "step": 10361
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1875781433128594,
+      "learning_rate": 9.613881863818504e-06,
+      "loss": 0.2038,
+      "step": 10362
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7732911230002019,
+      "learning_rate": 9.612236128604707e-06,
+      "loss": 0.1597,
+      "step": 10363
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8891492803152726,
+      "learning_rate": 9.610590403909028e-06,
+      "loss": 0.1673,
+      "step": 10364
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4303661276989428,
+      "learning_rate": 9.608944689776104e-06,
+      "loss": 0.2045,
+      "step": 10365
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1698005064544776,
+      "learning_rate": 9.607298986250578e-06,
+      "loss": 0.1826,
+      "step": 10366
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9063023615216049,
+      "learning_rate": 9.605653293377088e-06,
+      "loss": 0.1959,
+      "step": 10367
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9339563685741792,
+      "learning_rate": 9.60400761120028e-06,
+      "loss": 0.1899,
+      "step": 10368
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9848466830667983,
+      "learning_rate": 9.60236193976478e-06,
+      "loss": 0.1742,
+      "step": 10369
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9967791172095308,
+      "learning_rate": 9.600716279115244e-06,
+      "loss": 0.182,
+      "step": 10370
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1842568174382346,
+      "learning_rate": 9.599070629296295e-06,
+      "loss": 0.1643,
+      "step": 10371
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6531054022617186,
+      "learning_rate": 9.597424990352578e-06,
+      "loss": 0.2011,
+      "step": 10372
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7400670830777677,
+      "learning_rate": 9.595779362328731e-06,
+      "loss": 0.156,
+      "step": 10373
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9571451657640488,
+      "learning_rate": 9.594133745269396e-06,
+      "loss": 0.1706,
+      "step": 10374
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1804292443357665,
+      "learning_rate": 9.5924881392192e-06,
+      "loss": 0.1502,
+      "step": 10375
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3734033967550605,
+      "learning_rate": 9.59084254422279e-06,
+      "loss": 0.1662,
+      "step": 10376
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1294070061228543,
+      "learning_rate": 9.589196960324797e-06,
+      "loss": 0.2031,
+      "step": 10377
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1415644059540622,
+      "learning_rate": 9.587551387569863e-06,
+      "loss": 0.1832,
+      "step": 10378
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9893916259689024,
+      "learning_rate": 9.585905826002618e-06,
+      "loss": 0.1943,
+      "step": 10379
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9395348895068868,
+      "learning_rate": 9.584260275667702e-06,
+      "loss": 0.1786,
+      "step": 10380
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9656296619115283,
+      "learning_rate": 9.582614736609746e-06,
+      "loss": 0.1878,
+      "step": 10381
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9396833635719322,
+      "learning_rate": 9.580969208873396e-06,
+      "loss": 0.1971,
+      "step": 10382
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.850830522329186,
+      "learning_rate": 9.579323692503278e-06,
+      "loss": 0.1804,
+      "step": 10383
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9138329218506234,
+      "learning_rate": 9.577678187544024e-06,
+      "loss": 0.1811,
+      "step": 10384
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.895827065989639,
+      "learning_rate": 9.576032694040278e-06,
+      "loss": 0.19,
+      "step": 10385
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9258072443873068,
+      "learning_rate": 9.574387212036664e-06,
+      "loss": 0.1925,
+      "step": 10386
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8878547805058776,
+      "learning_rate": 9.572741741577826e-06,
+      "loss": 0.2078,
+      "step": 10387
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1455362422134288,
+      "learning_rate": 9.571096282708388e-06,
+      "loss": 0.1865,
+      "step": 10388
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6733140804840556,
+      "learning_rate": 9.56945083547299e-06,
+      "loss": 0.1918,
+      "step": 10389
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1029868972832708,
+      "learning_rate": 9.56780539991626e-06,
+      "loss": 0.184,
+      "step": 10390
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2399755222533273,
+      "learning_rate": 9.566159976082838e-06,
+      "loss": 0.1819,
+      "step": 10391
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.95713725716782,
+      "learning_rate": 9.564514564017345e-06,
+      "loss": 0.1873,
+      "step": 10392
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.241490567498554,
+      "learning_rate": 9.562869163764423e-06,
+      "loss": 0.2063,
+      "step": 10393
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.162189024921771,
+      "learning_rate": 9.561223775368697e-06,
+      "loss": 0.1756,
+      "step": 10394
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9746192496771294,
+      "learning_rate": 9.559578398874805e-06,
+      "loss": 0.189,
+      "step": 10395
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.997149034861936,
+      "learning_rate": 9.55793303432737e-06,
+      "loss": 0.1811,
+      "step": 10396
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8424462531241881,
+      "learning_rate": 9.55628768177103e-06,
+      "loss": 0.1777,
+      "step": 10397
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8116625462094226,
+      "learning_rate": 9.554642341250408e-06,
+      "loss": 0.1795,
+      "step": 10398
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9758897036252532,
+      "learning_rate": 9.552997012810138e-06,
+      "loss": 0.1524,
+      "step": 10399
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7726637556307456,
+      "learning_rate": 9.551351696494854e-06,
+      "loss": 0.1866,
+      "step": 10400
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4435038850679889,
+      "learning_rate": 9.549706392349175e-06,
+      "loss": 0.1796,
+      "step": 10401
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0780415380104225,
+      "learning_rate": 9.548061100417739e-06,
+      "loss": 0.1979,
+      "step": 10402
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0693027700309574,
+      "learning_rate": 9.546415820745168e-06,
+      "loss": 0.1951,
+      "step": 10403
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9908854052260064,
+      "learning_rate": 9.544770553376098e-06,
+      "loss": 0.1904,
+      "step": 10404
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6473184583869245,
+      "learning_rate": 9.543125298355147e-06,
+      "loss": 0.1923,
+      "step": 10405
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3924433832198697,
+      "learning_rate": 9.54148005572695e-06,
+      "loss": 0.1897,
+      "step": 10406
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.111984621097495,
+      "learning_rate": 9.539834825536131e-06,
+      "loss": 0.1845,
+      "step": 10407
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1519537417134282,
+      "learning_rate": 9.538189607827324e-06,
+      "loss": 0.1888,
+      "step": 10408
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.07458036048325,
+      "learning_rate": 9.536544402645144e-06,
+      "loss": 0.1847,
+      "step": 10409
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5093378512311653,
+      "learning_rate": 9.53489921003423e-06,
+      "loss": 0.1835,
+      "step": 10410
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0330251895942397,
+      "learning_rate": 9.533254030039193e-06,
+      "loss": 0.1596,
+      "step": 10411
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6626992170533568,
+      "learning_rate": 9.531608862704672e-06,
+      "loss": 0.2001,
+      "step": 10412
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.496480785595012,
+      "learning_rate": 9.529963708075284e-06,
+      "loss": 0.1882,
+      "step": 10413
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9989953762051174,
+      "learning_rate": 9.528318566195661e-06,
+      "loss": 0.1707,
+      "step": 10414
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.35912861959528,
+      "learning_rate": 9.52667343711042e-06,
+      "loss": 0.2268,
+      "step": 10415
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3312356143026387,
+      "learning_rate": 9.525028320864191e-06,
+      "loss": 0.1828,
+      "step": 10416
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9451334639642954,
+      "learning_rate": 9.523383217501596e-06,
+      "loss": 0.1897,
+      "step": 10417
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2835826239197037,
+      "learning_rate": 9.521738127067254e-06,
+      "loss": 0.1893,
+      "step": 10418
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9477317307320914,
+      "learning_rate": 9.520093049605796e-06,
+      "loss": 0.1738,
+      "step": 10419
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9035321903439242,
+      "learning_rate": 9.51844798516184e-06,
+      "loss": 0.1919,
+      "step": 10420
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7218146957856297,
+      "learning_rate": 9.516802933780011e-06,
+      "loss": 0.1565,
+      "step": 10421
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0801580770780828,
+      "learning_rate": 9.515157895504927e-06,
+      "loss": 0.192,
+      "step": 10422
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9582296167426243,
+      "learning_rate": 9.513512870381216e-06,
+      "loss": 0.1733,
+      "step": 10423
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0867462991915926,
+      "learning_rate": 9.511867858453493e-06,
+      "loss": 0.2014,
+      "step": 10424
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7490081315040854,
+      "learning_rate": 9.510222859766383e-06,
+      "loss": 0.1687,
+      "step": 10425
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9441063005416657,
+      "learning_rate": 9.508577874364503e-06,
+      "loss": 0.1992,
+      "step": 10426
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.211252729767083,
+      "learning_rate": 9.506932902292482e-06,
+      "loss": 0.182,
+      "step": 10427
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.384841114501458,
+      "learning_rate": 9.505287943594928e-06,
+      "loss": 0.175,
+      "step": 10428
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2205290375607247,
+      "learning_rate": 9.503642998316469e-06,
+      "loss": 0.1751,
+      "step": 10429
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8274716728200241,
+      "learning_rate": 9.501998066501718e-06,
+      "loss": 0.1955,
+      "step": 10430
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0254602672883348,
+      "learning_rate": 9.500353148195305e-06,
+      "loss": 0.2029,
+      "step": 10431
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7968841718287084,
+      "learning_rate": 9.498708243441834e-06,
+      "loss": 0.1899,
+      "step": 10432
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0826837399121725,
+      "learning_rate": 9.497063352285934e-06,
+      "loss": 0.1851,
+      "step": 10433
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9069755430051439,
+      "learning_rate": 9.495418474772221e-06,
+      "loss": 0.2015,
+      "step": 10434
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.271043067333602,
+      "learning_rate": 9.493773610945305e-06,
+      "loss": 0.194,
+      "step": 10435
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1972613171697342,
+      "learning_rate": 9.492128760849813e-06,
+      "loss": 0.2168,
+      "step": 10436
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0838621602061793,
+      "learning_rate": 9.490483924530353e-06,
+      "loss": 0.1731,
+      "step": 10437
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7529980716882423,
+      "learning_rate": 9.488839102031549e-06,
+      "loss": 0.1725,
+      "step": 10438
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8989904055035346,
+      "learning_rate": 9.48719429339801e-06,
+      "loss": 0.1859,
+      "step": 10439
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9278387486832459,
+      "learning_rate": 9.485549498674357e-06,
+      "loss": 0.191,
+      "step": 10440
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4556072385973486,
+      "learning_rate": 9.483904717905202e-06,
+      "loss": 0.1897,
+      "step": 10441
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9878938604682453,
+      "learning_rate": 9.48225995113516e-06,
+      "loss": 0.1697,
+      "step": 10442
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2287388076396861,
+      "learning_rate": 9.480615198408846e-06,
+      "loss": 0.1862,
+      "step": 10443
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8831594096054374,
+      "learning_rate": 9.478970459770878e-06,
+      "loss": 0.201,
+      "step": 10444
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9367663375726483,
+      "learning_rate": 9.47732573526586e-06,
+      "loss": 0.1834,
+      "step": 10445
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7777173866626321,
+      "learning_rate": 9.475681024938415e-06,
+      "loss": 0.2014,
+      "step": 10446
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.17639472112245,
+      "learning_rate": 9.474036328833148e-06,
+      "loss": 0.1915,
+      "step": 10447
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.910047139114023,
+      "learning_rate": 9.472391646994681e-06,
+      "loss": 0.1648,
+      "step": 10448
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.044503529177657,
+      "learning_rate": 9.470746979467614e-06,
+      "loss": 0.1794,
+      "step": 10449
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9275408983661686,
+      "learning_rate": 9.46910232629657e-06,
+      "loss": 0.1785,
+      "step": 10450
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0062888448818028,
+      "learning_rate": 9.467457687526156e-06,
+      "loss": 0.213,
+      "step": 10451
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0758056837608503,
+      "learning_rate": 9.465813063200978e-06,
+      "loss": 0.1883,
+      "step": 10452
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.200194759153906,
+      "learning_rate": 9.464168453365655e-06,
+      "loss": 0.1687,
+      "step": 10453
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8561814641014569,
+      "learning_rate": 9.462523858064788e-06,
+      "loss": 0.1801,
+      "step": 10454
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4085649439814016,
+      "learning_rate": 9.460879277342995e-06,
+      "loss": 0.1891,
+      "step": 10455
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.072865594506947,
+      "learning_rate": 9.459234711244881e-06,
+      "loss": 0.1901,
+      "step": 10456
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4323014820101265,
+      "learning_rate": 9.457590159815058e-06,
+      "loss": 0.1858,
+      "step": 10457
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4154831275527877,
+      "learning_rate": 9.45594562309813e-06,
+      "loss": 0.2139,
+      "step": 10458
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0001665481648387,
+      "learning_rate": 9.454301101138708e-06,
+      "loss": 0.1876,
+      "step": 10459
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9027308619514212,
+      "learning_rate": 9.452656593981398e-06,
+      "loss": 0.1831,
+      "step": 10460
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9316900980594415,
+      "learning_rate": 9.451012101670814e-06,
+      "loss": 0.1858,
+      "step": 10461
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7836745774854093,
+      "learning_rate": 9.449367624251554e-06,
+      "loss": 0.1624,
+      "step": 10462
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.230737222455105,
+      "learning_rate": 9.447723161768228e-06,
+      "loss": 0.1664,
+      "step": 10463
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8627584376712686,
+      "learning_rate": 9.446078714265441e-06,
+      "loss": 0.1978,
+      "step": 10464
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0735143941794252,
+      "learning_rate": 9.444434281787806e-06,
+      "loss": 0.1821,
+      "step": 10465
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.113359953970941,
+      "learning_rate": 9.442789864379918e-06,
+      "loss": 0.1847,
+      "step": 10466
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2387280858722214,
+      "learning_rate": 9.44114546208639e-06,
+      "loss": 0.1765,
+      "step": 10467
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8166548705579885,
+      "learning_rate": 9.439501074951817e-06,
+      "loss": 0.1841,
+      "step": 10468
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9275349732677751,
+      "learning_rate": 9.437856703020813e-06,
+      "loss": 0.1949,
+      "step": 10469
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9684713113188496,
+      "learning_rate": 9.436212346337981e-06,
+      "loss": 0.2019,
+      "step": 10470
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.752915393134308,
+      "learning_rate": 9.434568004947914e-06,
+      "loss": 0.1811,
+      "step": 10471
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0686608805981916,
+      "learning_rate": 9.432923678895225e-06,
+      "loss": 0.1788,
+      "step": 10472
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3131991830901886,
+      "learning_rate": 9.431279368224512e-06,
+      "loss": 0.1957,
+      "step": 10473
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8220619801982747,
+      "learning_rate": 9.429635072980382e-06,
+      "loss": 0.1781,
+      "step": 10474
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7944177343908115,
+      "learning_rate": 9.427990793207428e-06,
+      "loss": 0.1783,
+      "step": 10475
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8865612068771639,
+      "learning_rate": 9.426346528950258e-06,
+      "loss": 0.1681,
+      "step": 10476
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0013291908495037,
+      "learning_rate": 9.42470228025347e-06,
+      "loss": 0.1789,
+      "step": 10477
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1030495223038712,
+      "learning_rate": 9.423058047161668e-06,
+      "loss": 0.1759,
+      "step": 10478
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2086008136565232,
+      "learning_rate": 9.421413829719446e-06,
+      "loss": 0.191,
+      "step": 10479
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0555368232810498,
+      "learning_rate": 9.41976962797141e-06,
+      "loss": 0.2003,
+      "step": 10480
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8284304837222289,
+      "learning_rate": 9.418125441962151e-06,
+      "loss": 0.1916,
+      "step": 10481
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1149723921087935,
+      "learning_rate": 9.416481271736276e-06,
+      "loss": 0.178,
+      "step": 10482
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2590214185530384,
+      "learning_rate": 9.414837117338376e-06,
+      "loss": 0.2095,
+      "step": 10483
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9191664131662469,
+      "learning_rate": 9.413192978813057e-06,
+      "loss": 0.183,
+      "step": 10484
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0947807191380714,
+      "learning_rate": 9.411548856204907e-06,
+      "loss": 0.1784,
+      "step": 10485
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8485057552267062,
+      "learning_rate": 9.409904749558529e-06,
+      "loss": 0.1741,
+      "step": 10486
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2353579300556639,
+      "learning_rate": 9.408260658918522e-06,
+      "loss": 0.1711,
+      "step": 10487
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9047841484421405,
+      "learning_rate": 9.406616584329473e-06,
+      "loss": 0.1886,
+      "step": 10488
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0537218642384858,
+      "learning_rate": 9.404972525835984e-06,
+      "loss": 0.17,
+      "step": 10489
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.490393840000147,
+      "learning_rate": 9.403328483482647e-06,
+      "loss": 0.1835,
+      "step": 10490
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9113168576797979,
+      "learning_rate": 9.401684457314064e-06,
+      "loss": 0.1985,
+      "step": 10491
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8407452044189376,
+      "learning_rate": 9.400040447374818e-06,
+      "loss": 0.1784,
+      "step": 10492
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0300144394706507,
+      "learning_rate": 9.398396453709514e-06,
+      "loss": 0.1661,
+      "step": 10493
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0476008663127399,
+      "learning_rate": 9.396752476362735e-06,
+      "loss": 0.1865,
+      "step": 10494
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6707348934781523,
+      "learning_rate": 9.395108515379082e-06,
+      "loss": 0.2134,
+      "step": 10495
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4262919869218256,
+      "learning_rate": 9.393464570803142e-06,
+      "loss": 0.2053,
+      "step": 10496
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1980144717024208,
+      "learning_rate": 9.391820642679515e-06,
+      "loss": 0.1852,
+      "step": 10497
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0619071454147688,
+      "learning_rate": 9.390176731052783e-06,
+      "loss": 0.1835,
+      "step": 10498
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9156702670758646,
+      "learning_rate": 9.388532835967543e-06,
+      "loss": 0.1979,
+      "step": 10499
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1991320317620187,
+      "learning_rate": 9.386888957468383e-06,
+      "loss": 0.1743,
+      "step": 10500
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0832346660178478,
+      "learning_rate": 9.385245095599897e-06,
+      "loss": 0.1608,
+      "step": 10501
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1136661258607057,
+      "learning_rate": 9.38360125040667e-06,
+      "loss": 0.1903,
+      "step": 10502
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9346059659450856,
+      "learning_rate": 9.381957421933296e-06,
+      "loss": 0.1833,
+      "step": 10503
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0498713030587017,
+      "learning_rate": 9.380313610224364e-06,
+      "loss": 0.1831,
+      "step": 10504
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1507233448862686,
+      "learning_rate": 9.378669815324456e-06,
+      "loss": 0.1679,
+      "step": 10505
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7585534369008222,
+      "learning_rate": 9.377026037278169e-06,
+      "loss": 0.1746,
+      "step": 10506
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1627401646562339,
+      "learning_rate": 9.37538227613008e-06,
+      "loss": 0.1868,
+      "step": 10507
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9497167071824306,
+      "learning_rate": 9.373738531924787e-06,
+      "loss": 0.1733,
+      "step": 10508
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9222271876528074,
+      "learning_rate": 9.372094804706867e-06,
+      "loss": 0.1955,
+      "step": 10509
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.982110484488253,
+      "learning_rate": 9.370451094520915e-06,
+      "loss": 0.1801,
+      "step": 10510
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0602814609731621,
+      "learning_rate": 9.36880740141151e-06,
+      "loss": 0.1606,
+      "step": 10511
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2839696823036988,
+      "learning_rate": 9.367163725423242e-06,
+      "loss": 0.177,
+      "step": 10512
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2200605917065472,
+      "learning_rate": 9.365520066600691e-06,
+      "loss": 0.1706,
+      "step": 10513
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8684543483198988,
+      "learning_rate": 9.363876424988447e-06,
+      "loss": 0.19,
+      "step": 10514
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4733424165761082,
+      "learning_rate": 9.362232800631087e-06,
+      "loss": 0.1808,
+      "step": 10515
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.332430221388429,
+      "learning_rate": 9.360589193573203e-06,
+      "loss": 0.1703,
+      "step": 10516
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0983837973789459,
+      "learning_rate": 9.358945603859369e-06,
+      "loss": 0.1682,
+      "step": 10517
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9061642429714712,
+      "learning_rate": 9.357302031534178e-06,
+      "loss": 0.1752,
+      "step": 10518
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0453790873171436,
+      "learning_rate": 9.355658476642198e-06,
+      "loss": 0.191,
+      "step": 10519
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8765708013616752,
+      "learning_rate": 9.354014939228024e-06,
+      "loss": 0.1748,
+      "step": 10520
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.890012437695401,
+      "learning_rate": 9.35237141933623e-06,
+      "loss": 0.2308,
+      "step": 10521
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9096191796737242,
+      "learning_rate": 9.350727917011395e-06,
+      "loss": 0.1793,
+      "step": 10522
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5613106494417954,
+      "learning_rate": 9.349084432298106e-06,
+      "loss": 0.1811,
+      "step": 10523
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0860020915804065,
+      "learning_rate": 9.347440965240934e-06,
+      "loss": 0.2029,
+      "step": 10524
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8887090919405828,
+      "learning_rate": 9.345797515884466e-06,
+      "loss": 0.174,
+      "step": 10525
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8486337984116933,
+      "learning_rate": 9.344154084273275e-06,
+      "loss": 0.1791,
+      "step": 10526
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.006827054252546,
+      "learning_rate": 9.342510670451944e-06,
+      "loss": 0.1654,
+      "step": 10527
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9545599629514354,
+      "learning_rate": 9.340867274465047e-06,
+      "loss": 0.1948,
+      "step": 10528
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.047854515192047,
+      "learning_rate": 9.339223896357163e-06,
+      "loss": 0.1808,
+      "step": 10529
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1177400791380077,
+      "learning_rate": 9.337580536172864e-06,
+      "loss": 0.1798,
+      "step": 10530
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8060106213647832,
+      "learning_rate": 9.335937193956736e-06,
+      "loss": 0.1807,
+      "step": 10531
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.884147369338201,
+      "learning_rate": 9.334293869753346e-06,
+      "loss": 0.1824,
+      "step": 10532
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.082017160418502,
+      "learning_rate": 9.332650563607276e-06,
+      "loss": 0.205,
+      "step": 10533
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8124867757454877,
+      "learning_rate": 9.331007275563093e-06,
+      "loss": 0.1854,
+      "step": 10534
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8725530897469803,
+      "learning_rate": 9.329364005665377e-06,
+      "loss": 0.1858,
+      "step": 10535
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8761616764100726,
+      "learning_rate": 9.327720753958699e-06,
+      "loss": 0.2076,
+      "step": 10536
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0565332273224135,
+      "learning_rate": 9.326077520487637e-06,
+      "loss": 0.2011,
+      "step": 10537
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5342916385910432,
+      "learning_rate": 9.324434305296757e-06,
+      "loss": 0.1785,
+      "step": 10538
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9888427341066531,
+      "learning_rate": 9.322791108430636e-06,
+      "loss": 0.2056,
+      "step": 10539
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8702642528267842,
+      "learning_rate": 9.321147929933847e-06,
+      "loss": 0.1675,
+      "step": 10540
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2638876117441158,
+      "learning_rate": 9.319504769850953e-06,
+      "loss": 0.1782,
+      "step": 10541
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9326731432920613,
+      "learning_rate": 9.317861628226535e-06,
+      "loss": 0.1902,
+      "step": 10542
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0811358970036962,
+      "learning_rate": 9.316218505105155e-06,
+      "loss": 0.1943,
+      "step": 10543
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9911777388698443,
+      "learning_rate": 9.314575400531391e-06,
+      "loss": 0.2022,
+      "step": 10544
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1818692996375626,
+      "learning_rate": 9.312932314549804e-06,
+      "loss": 0.1933,
+      "step": 10545
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9087683719340895,
+      "learning_rate": 9.311289247204972e-06,
+      "loss": 0.187,
+      "step": 10546
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2647180730801626,
+      "learning_rate": 9.309646198541454e-06,
+      "loss": 0.1762,
+      "step": 10547
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0285760495995913,
+      "learning_rate": 9.308003168603822e-06,
+      "loss": 0.1709,
+      "step": 10548
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4024991467330072,
+      "learning_rate": 9.306360157436642e-06,
+      "loss": 0.1795,
+      "step": 10549
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9153367260426736,
+      "learning_rate": 9.304717165084486e-06,
+      "loss": 0.1668,
+      "step": 10550
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9666848999268505,
+      "learning_rate": 9.303074191591912e-06,
+      "loss": 0.2107,
+      "step": 10551
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.443604240907049,
+      "learning_rate": 9.301431237003492e-06,
+      "loss": 0.1799,
+      "step": 10552
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1412227625228761,
+      "learning_rate": 9.299788301363786e-06,
+      "loss": 0.1912,
+      "step": 10553
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.074255400598678,
+      "learning_rate": 9.298145384717369e-06,
+      "loss": 0.1648,
+      "step": 10554
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.278035111588236,
+      "learning_rate": 9.296502487108792e-06,
+      "loss": 0.2043,
+      "step": 10555
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9894725590022767,
+      "learning_rate": 9.294859608582626e-06,
+      "loss": 0.1744,
+      "step": 10556
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9126589381137777,
+      "learning_rate": 9.293216749183437e-06,
+      "loss": 0.1606,
+      "step": 10557
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9971284910584742,
+      "learning_rate": 9.291573908955777e-06,
+      "loss": 0.1872,
+      "step": 10558
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9356317290539621,
+      "learning_rate": 9.289931087944221e-06,
+      "loss": 0.1674,
+      "step": 10559
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9180737699979528,
+      "learning_rate": 9.28828828619332e-06,
+      "loss": 0.1911,
+      "step": 10560
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3102951476006044,
+      "learning_rate": 9.286645503747641e-06,
+      "loss": 0.1943,
+      "step": 10561
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0486085362353716,
+      "learning_rate": 9.28500274065174e-06,
+      "loss": 0.172,
+      "step": 10562
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0143367843255702,
+      "learning_rate": 9.283359996950187e-06,
+      "loss": 0.1923,
+      "step": 10563
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0226800184238658,
+      "learning_rate": 9.281717272687527e-06,
+      "loss": 0.1905,
+      "step": 10564
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0633593771594727,
+      "learning_rate": 9.28007456790833e-06,
+      "loss": 0.1949,
+      "step": 10565
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9299341580114174,
+      "learning_rate": 9.27843188265715e-06,
+      "loss": 0.1804,
+      "step": 10566
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.169669224468857,
+      "learning_rate": 9.276789216978549e-06,
+      "loss": 0.2227,
+      "step": 10567
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9097009835886758,
+      "learning_rate": 9.275146570917077e-06,
+      "loss": 0.1884,
+      "step": 10568
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9145517605087363,
+      "learning_rate": 9.273503944517298e-06,
+      "loss": 0.1675,
+      "step": 10569
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8695056976150293,
+      "learning_rate": 9.271861337823763e-06,
+      "loss": 0.1663,
+      "step": 10570
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.912319748951526,
+      "learning_rate": 9.270218750881036e-06,
+      "loss": 0.1891,
+      "step": 10571
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1769578406569654,
+      "learning_rate": 9.26857618373366e-06,
+      "loss": 0.2083,
+      "step": 10572
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1584981841543633,
+      "learning_rate": 9.2669336364262e-06,
+      "loss": 0.1728,
+      "step": 10573
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7834513467328376,
+      "learning_rate": 9.265291109003208e-06,
+      "loss": 0.1634,
+      "step": 10574
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5024126986383335,
+      "learning_rate": 9.263648601509231e-06,
+      "loss": 0.1773,
+      "step": 10575
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9605455781620283,
+      "learning_rate": 9.262006113988832e-06,
+      "loss": 0.1835,
+      "step": 10576
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7979609875885121,
+      "learning_rate": 9.260363646486556e-06,
+      "loss": 0.185,
+      "step": 10577
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0930640857959548,
+      "learning_rate": 9.25872119904696e-06,
+      "loss": 0.1911,
+      "step": 10578
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5730311189725499,
+      "learning_rate": 9.257078771714591e-06,
+      "loss": 0.1897,
+      "step": 10579
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.00160813983151,
+      "learning_rate": 9.255436364534005e-06,
+      "loss": 0.2044,
+      "step": 10580
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.907234305672958,
+      "learning_rate": 9.253793977549747e-06,
+      "loss": 0.1529,
+      "step": 10581
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7767435462070758,
+      "learning_rate": 9.25215161080637e-06,
+      "loss": 0.1958,
+      "step": 10582
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9855417730617082,
+      "learning_rate": 9.250509264348422e-06,
+      "loss": 0.2006,
+      "step": 10583
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8226800135543453,
+      "learning_rate": 9.248866938220456e-06,
+      "loss": 0.1796,
+      "step": 10584
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8673771835954791,
+      "learning_rate": 9.247224632467014e-06,
+      "loss": 0.1813,
+      "step": 10585
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6398020717643,
+      "learning_rate": 9.245582347132646e-06,
+      "loss": 0.1856,
+      "step": 10586
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9801202936924129,
+      "learning_rate": 9.2439400822619e-06,
+      "loss": 0.1629,
+      "step": 10587
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.09510375553041,
+      "learning_rate": 9.242297837899325e-06,
+      "loss": 0.1768,
+      "step": 10588
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.005381026821629,
+      "learning_rate": 9.240655614089459e-06,
+      "loss": 0.2058,
+      "step": 10589
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3320793772483235,
+      "learning_rate": 9.239013410876856e-06,
+      "loss": 0.1891,
+      "step": 10590
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1738236438179857,
+      "learning_rate": 9.237371228306057e-06,
+      "loss": 0.202,
+      "step": 10591
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1494644621728747,
+      "learning_rate": 9.235729066421604e-06,
+      "loss": 0.1738,
+      "step": 10592
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9827935431430217,
+      "learning_rate": 9.234086925268046e-06,
+      "loss": 0.1813,
+      "step": 10593
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1216711618319328,
+      "learning_rate": 9.232444804889921e-06,
+      "loss": 0.1901,
+      "step": 10594
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3244857919711657,
+      "learning_rate": 9.230802705331776e-06,
+      "loss": 0.184,
+      "step": 10595
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9137910573447837,
+      "learning_rate": 9.229160626638148e-06,
+      "loss": 0.1817,
+      "step": 10596
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9236684179786859,
+      "learning_rate": 9.227518568853587e-06,
+      "loss": 0.1825,
+      "step": 10597
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.135558348639019,
+      "learning_rate": 9.225876532022623e-06,
+      "loss": 0.1815,
+      "step": 10598
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9099245816546617,
+      "learning_rate": 9.224234516189803e-06,
+      "loss": 0.1802,
+      "step": 10599
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.936960906132101,
+      "learning_rate": 9.222592521399666e-06,
+      "loss": 0.1844,
+      "step": 10600
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.196191487570183,
+      "learning_rate": 9.220950547696754e-06,
+      "loss": 0.1801,
+      "step": 10601
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1492432241366455,
+      "learning_rate": 9.219308595125598e-06,
+      "loss": 0.1677,
+      "step": 10602
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2051869978190075,
+      "learning_rate": 9.217666663730744e-06,
+      "loss": 0.1814,
+      "step": 10603
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0243526918755632,
+      "learning_rate": 9.216024753556722e-06,
+      "loss": 0.1958,
+      "step": 10604
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.084683646176466,
+      "learning_rate": 9.214382864648075e-06,
+      "loss": 0.1562,
+      "step": 10605
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.080449899563939,
+      "learning_rate": 9.212740997049335e-06,
+      "loss": 0.2075,
+      "step": 10606
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2832964653669452,
+      "learning_rate": 9.211099150805046e-06,
+      "loss": 0.1761,
+      "step": 10607
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7528709193932426,
+      "learning_rate": 9.209457325959731e-06,
+      "loss": 0.1813,
+      "step": 10608
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8444403300450413,
+      "learning_rate": 9.207815522557932e-06,
+      "loss": 0.1839,
+      "step": 10609
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9972227451408476,
+      "learning_rate": 9.206173740644186e-06,
+      "loss": 0.1934,
+      "step": 10610
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.8715618480349923,
+      "learning_rate": 9.204531980263017e-06,
+      "loss": 0.191,
+      "step": 10611
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2879592708261964,
+      "learning_rate": 9.202890241458963e-06,
+      "loss": 0.1668,
+      "step": 10612
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.815992453516756,
+      "learning_rate": 9.201248524276557e-06,
+      "loss": 0.1838,
+      "step": 10613
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3697672435436143,
+      "learning_rate": 9.199606828760333e-06,
+      "loss": 0.2018,
+      "step": 10614
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0653803940723892,
+      "learning_rate": 9.197965154954815e-06,
+      "loss": 0.1773,
+      "step": 10615
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7669012746424213,
+      "learning_rate": 9.196323502904542e-06,
+      "loss": 0.1711,
+      "step": 10616
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2937877619242906,
+      "learning_rate": 9.194681872654034e-06,
+      "loss": 0.1799,
+      "step": 10617
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.490122996501818,
+      "learning_rate": 9.19304026424783e-06,
+      "loss": 0.1531,
+      "step": 10618
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3258244978870801,
+      "learning_rate": 9.191398677730449e-06,
+      "loss": 0.1871,
+      "step": 10619
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9765260196003005,
+      "learning_rate": 9.189757113146431e-06,
+      "loss": 0.1832,
+      "step": 10620
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1123303040511696,
+      "learning_rate": 9.188115570540292e-06,
+      "loss": 0.1659,
+      "step": 10621
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.132217710576086,
+      "learning_rate": 9.186474049956568e-06,
+      "loss": 0.1827,
+      "step": 10622
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0677326613521958,
+      "learning_rate": 9.184832551439777e-06,
+      "loss": 0.1999,
+      "step": 10623
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8733102154009217,
+      "learning_rate": 9.183191075034455e-06,
+      "loss": 0.1573,
+      "step": 10624
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4879165809895192,
+      "learning_rate": 9.181549620785115e-06,
+      "loss": 0.1958,
+      "step": 10625
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.989982301785905,
+      "learning_rate": 9.179908188736291e-06,
+      "loss": 0.1876,
+      "step": 10626
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9870998274556386,
+      "learning_rate": 9.178266778932506e-06,
+      "loss": 0.1604,
+      "step": 10627
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0754699021323235,
+      "learning_rate": 9.176625391418277e-06,
+      "loss": 0.1792,
+      "step": 10628
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8355366004814154,
+      "learning_rate": 9.174984026238134e-06,
+      "loss": 0.1895,
+      "step": 10629
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4989380803523193,
+      "learning_rate": 9.173342683436593e-06,
+      "loss": 0.2049,
+      "step": 10630
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7914223255584208,
+      "learning_rate": 9.171701363058179e-06,
+      "loss": 0.1691,
+      "step": 10631
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9393068625759057,
+      "learning_rate": 9.17006006514741e-06,
+      "loss": 0.1749,
+      "step": 10632
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1462009659205226,
+      "learning_rate": 9.168418789748815e-06,
+      "loss": 0.1781,
+      "step": 10633
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3736892729869854,
+      "learning_rate": 9.166777536906901e-06,
+      "loss": 0.1726,
+      "step": 10634
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1734054679774246,
+      "learning_rate": 9.165136306666196e-06,
+      "loss": 0.1598,
+      "step": 10635
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8581227962415139,
+      "learning_rate": 9.163495099071214e-06,
+      "loss": 0.1763,
+      "step": 10636
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9346363358837523,
+      "learning_rate": 9.161853914166479e-06,
+      "loss": 0.2053,
+      "step": 10637
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0907314457043735,
+      "learning_rate": 9.1602127519965e-06,
+      "loss": 0.1876,
+      "step": 10638
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9318701442009211,
+      "learning_rate": 9.1585716126058e-06,
+      "loss": 0.1759,
+      "step": 10639
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9744826376394506,
+      "learning_rate": 9.15693049603889e-06,
+      "loss": 0.2064,
+      "step": 10640
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0207950872373368,
+      "learning_rate": 9.155289402340291e-06,
+      "loss": 0.1799,
+      "step": 10641
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9647089007127575,
+      "learning_rate": 9.153648331554511e-06,
+      "loss": 0.1964,
+      "step": 10642
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1810311399011428,
+      "learning_rate": 9.152007283726073e-06,
+      "loss": 0.1655,
+      "step": 10643
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.67131220083229,
+      "learning_rate": 9.150366258899483e-06,
+      "loss": 0.2026,
+      "step": 10644
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0390141748363309,
+      "learning_rate": 9.148725257119253e-06,
+      "loss": 0.1978,
+      "step": 10645
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4274642279887897,
+      "learning_rate": 9.147084278429903e-06,
+      "loss": 0.1734,
+      "step": 10646
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1536905388237146,
+      "learning_rate": 9.145443322875937e-06,
+      "loss": 0.1664,
+      "step": 10647
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0672186751727026,
+      "learning_rate": 9.14380239050187e-06,
+      "loss": 0.1857,
+      "step": 10648
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1406716880450503,
+      "learning_rate": 9.14216148135221e-06,
+      "loss": 0.1912,
+      "step": 10649
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6124381866654078,
+      "learning_rate": 9.140520595471473e-06,
+      "loss": 0.1919,
+      "step": 10650
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9507612008810503,
+      "learning_rate": 9.138879732904157e-06,
+      "loss": 0.2101,
+      "step": 10651
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8238385961207768,
+      "learning_rate": 9.137238893694781e-06,
+      "loss": 0.2056,
+      "step": 10652
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9904799908904333,
+      "learning_rate": 9.135598077887846e-06,
+      "loss": 0.192,
+      "step": 10653
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0593204896708142,
+      "learning_rate": 9.133957285527868e-06,
+      "loss": 0.1807,
+      "step": 10654
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8334129482294383,
+      "learning_rate": 9.132316516659341e-06,
+      "loss": 0.1912,
+      "step": 10655
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.878014703805649,
+      "learning_rate": 9.130675771326783e-06,
+      "loss": 0.1927,
+      "step": 10656
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0126816393076499,
+      "learning_rate": 9.129035049574688e-06,
+      "loss": 0.2007,
+      "step": 10657
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.280017082713366,
+      "learning_rate": 9.12739435144757e-06,
+      "loss": 0.1823,
+      "step": 10658
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2509973026115213,
+      "learning_rate": 9.125753676989926e-06,
+      "loss": 0.1979,
+      "step": 10659
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.090298495463318,
+      "learning_rate": 9.124113026246268e-06,
+      "loss": 0.1854,
+      "step": 10660
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5921016493384281,
+      "learning_rate": 9.122472399261092e-06,
+      "loss": 0.1974,
+      "step": 10661
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9171169378239742,
+      "learning_rate": 9.1208317960789e-06,
+      "loss": 0.197,
+      "step": 10662
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8770555727852963,
+      "learning_rate": 9.119191216744197e-06,
+      "loss": 0.2032,
+      "step": 10663
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3314627171471414,
+      "learning_rate": 9.11755066130148e-06,
+      "loss": 0.1925,
+      "step": 10664
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8738649746145589,
+      "learning_rate": 9.115910129795252e-06,
+      "loss": 0.184,
+      "step": 10665
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9780393428276124,
+      "learning_rate": 9.11426962227001e-06,
+      "loss": 0.1919,
+      "step": 10666
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9004068467150715,
+      "learning_rate": 9.112629138770259e-06,
+      "loss": 0.1904,
+      "step": 10667
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0289144423469088,
+      "learning_rate": 9.110988679340488e-06,
+      "loss": 0.1721,
+      "step": 10668
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1838230125708527,
+      "learning_rate": 9.109348244025204e-06,
+      "loss": 0.1929,
+      "step": 10669
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9743941893483662,
+      "learning_rate": 9.107707832868896e-06,
+      "loss": 0.1949,
+      "step": 10670
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8473092924176169,
+      "learning_rate": 9.106067445916064e-06,
+      "loss": 0.1724,
+      "step": 10671
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.957361674100075,
+      "learning_rate": 9.104427083211201e-06,
+      "loss": 0.1669,
+      "step": 10672
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7641459772145157,
+      "learning_rate": 9.10278674479881e-06,
+      "loss": 0.166,
+      "step": 10673
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9649873289264107,
+      "learning_rate": 9.101146430723373e-06,
+      "loss": 0.157,
+      "step": 10674
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.695599698503498,
+      "learning_rate": 9.099506141029393e-06,
+      "loss": 0.1863,
+      "step": 10675
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.041754790008622,
+      "learning_rate": 9.097865875761356e-06,
+      "loss": 0.1821,
+      "step": 10676
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4290400240424552,
+      "learning_rate": 9.096225634963764e-06,
+      "loss": 0.1835,
+      "step": 10677
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2286969932742677,
+      "learning_rate": 9.094585418681098e-06,
+      "loss": 0.1774,
+      "step": 10678
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8841463674321048,
+      "learning_rate": 9.092945226957856e-06,
+      "loss": 0.1698,
+      "step": 10679
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.219517297991463,
+      "learning_rate": 9.091305059838528e-06,
+      "loss": 0.1917,
+      "step": 10680
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2140141811496525,
+      "learning_rate": 9.089664917367597e-06,
+      "loss": 0.1842,
+      "step": 10681
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2792334097251197,
+      "learning_rate": 9.088024799589557e-06,
+      "loss": 0.2023,
+      "step": 10682
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0618023545637925,
+      "learning_rate": 9.086384706548897e-06,
+      "loss": 0.1664,
+      "step": 10683
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.191493755751196,
+      "learning_rate": 9.084744638290105e-06,
+      "loss": 0.1846,
+      "step": 10684
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9933607568905328,
+      "learning_rate": 9.083104594857663e-06,
+      "loss": 0.1936,
+      "step": 10685
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8200920801777767,
+      "learning_rate": 9.081464576296066e-06,
+      "loss": 0.1865,
+      "step": 10686
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.971900461440888,
+      "learning_rate": 9.079824582649788e-06,
+      "loss": 0.1748,
+      "step": 10687
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8686920317236623,
+      "learning_rate": 9.078184613963324e-06,
+      "loss": 0.1749,
+      "step": 10688
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0556078077071391,
+      "learning_rate": 9.076544670281153e-06,
+      "loss": 0.1644,
+      "step": 10689
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8479512987307262,
+      "learning_rate": 9.074904751647762e-06,
+      "loss": 0.1655,
+      "step": 10690
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9580535758655278,
+      "learning_rate": 9.073264858107628e-06,
+      "loss": 0.2055,
+      "step": 10691
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0938096694712256,
+      "learning_rate": 9.071624989705241e-06,
+      "loss": 0.1681,
+      "step": 10692
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4343843624625687,
+      "learning_rate": 9.069985146485075e-06,
+      "loss": 0.1951,
+      "step": 10693
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9902205037582185,
+      "learning_rate": 9.068345328491619e-06,
+      "loss": 0.1865,
+      "step": 10694
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9372712378232515,
+      "learning_rate": 9.066705535769345e-06,
+      "loss": 0.1802,
+      "step": 10695
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4848712846231706,
+      "learning_rate": 9.065065768362736e-06,
+      "loss": 0.1754,
+      "step": 10696
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5807359672516952,
+      "learning_rate": 9.063426026316275e-06,
+      "loss": 0.1838,
+      "step": 10697
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.059155620869402,
+      "learning_rate": 9.061786309674431e-06,
+      "loss": 0.1754,
+      "step": 10698
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9541798586769945,
+      "learning_rate": 9.060146618481692e-06,
+      "loss": 0.1656,
+      "step": 10699
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.81118910909562,
+      "learning_rate": 9.058506952782522e-06,
+      "loss": 0.2161,
+      "step": 10700
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0132204443059734,
+      "learning_rate": 9.056867312621408e-06,
+      "loss": 0.1715,
+      "step": 10701
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0257858478307955,
+      "learning_rate": 9.05522769804282e-06,
+      "loss": 0.193,
+      "step": 10702
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8758821720127123,
+      "learning_rate": 9.053588109091236e-06,
+      "loss": 0.1795,
+      "step": 10703
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.459217006890036,
+      "learning_rate": 9.051948545811125e-06,
+      "loss": 0.1789,
+      "step": 10704
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0812146827278828,
+      "learning_rate": 9.050309008246965e-06,
+      "loss": 0.2045,
+      "step": 10705
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7067803049987202,
+      "learning_rate": 9.048669496443226e-06,
+      "loss": 0.197,
+      "step": 10706
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0480730399539124,
+      "learning_rate": 9.047030010444382e-06,
+      "loss": 0.1772,
+      "step": 10707
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9990838162286354,
+      "learning_rate": 9.045390550294901e-06,
+      "loss": 0.1904,
+      "step": 10708
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8913607947233958,
+      "learning_rate": 9.043751116039257e-06,
+      "loss": 0.182,
+      "step": 10709
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0511457664452766,
+      "learning_rate": 9.042111707721916e-06,
+      "loss": 0.1594,
+      "step": 10710
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0732583932499873,
+      "learning_rate": 9.040472325387352e-06,
+      "loss": 0.2012,
+      "step": 10711
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.903939596274699,
+      "learning_rate": 9.038832969080029e-06,
+      "loss": 0.1879,
+      "step": 10712
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.200570819981873,
+      "learning_rate": 9.037193638844417e-06,
+      "loss": 0.1818,
+      "step": 10713
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7451402376048245,
+      "learning_rate": 9.035554334724982e-06,
+      "loss": 0.1894,
+      "step": 10714
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0144393292155092,
+      "learning_rate": 9.033915056766187e-06,
+      "loss": 0.1624,
+      "step": 10715
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1598482461764674,
+      "learning_rate": 9.032275805012505e-06,
+      "loss": 0.187,
+      "step": 10716
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4088383790140577,
+      "learning_rate": 9.030636579508392e-06,
+      "loss": 0.1993,
+      "step": 10717
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9323936061752953,
+      "learning_rate": 9.02899738029832e-06,
+      "loss": 0.1877,
+      "step": 10718
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6577111888773488,
+      "learning_rate": 9.027358207426745e-06,
+      "loss": 0.2049,
+      "step": 10719
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9143010624235737,
+      "learning_rate": 9.025719060938138e-06,
+      "loss": 0.1708,
+      "step": 10720
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8855990663781108,
+      "learning_rate": 9.024079940876954e-06,
+      "loss": 0.199,
+      "step": 10721
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9750804804995725,
+      "learning_rate": 9.022440847287656e-06,
+      "loss": 0.2279,
+      "step": 10722
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8803501296190209,
+      "learning_rate": 9.020801780214705e-06,
+      "loss": 0.1813,
+      "step": 10723
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9742924036790803,
+      "learning_rate": 9.019162739702564e-06,
+      "loss": 0.1951,
+      "step": 10724
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0582094217135745,
+      "learning_rate": 9.017523725795688e-06,
+      "loss": 0.1769,
+      "step": 10725
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9670077934281341,
+      "learning_rate": 9.015884738538537e-06,
+      "loss": 0.1735,
+      "step": 10726
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9456609790651568,
+      "learning_rate": 9.014245777975565e-06,
+      "loss": 0.1664,
+      "step": 10727
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5450763527086255,
+      "learning_rate": 9.012606844151235e-06,
+      "loss": 0.2163,
+      "step": 10728
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8390513214365491,
+      "learning_rate": 9.010967937109997e-06,
+      "loss": 0.1823,
+      "step": 10729
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0167610566569403,
+      "learning_rate": 9.009329056896314e-06,
+      "loss": 0.2059,
+      "step": 10730
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8110446302914665,
+      "learning_rate": 9.007690203554636e-06,
+      "loss": 0.1828,
+      "step": 10731
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9537201974786012,
+      "learning_rate": 9.006051377129412e-06,
+      "loss": 0.1662,
+      "step": 10732
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8364630993085369,
+      "learning_rate": 9.004412577665107e-06,
+      "loss": 0.166,
+      "step": 10733
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9991791474471617,
+      "learning_rate": 9.002773805206161e-06,
+      "loss": 0.198,
+      "step": 10734
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.328538844649469,
+      "learning_rate": 9.001135059797036e-06,
+      "loss": 0.1678,
+      "step": 10735
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.333354113863571,
+      "learning_rate": 8.999496341482176e-06,
+      "loss": 0.1917,
+      "step": 10736
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1206005428077606,
+      "learning_rate": 8.997857650306038e-06,
+      "loss": 0.209,
+      "step": 10737
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1991656192292084,
+      "learning_rate": 8.996218986313063e-06,
+      "loss": 0.2033,
+      "step": 10738
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 11.08974914382019,
+      "learning_rate": 8.994580349547711e-06,
+      "loss": 0.1793,
+      "step": 10739
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8497449718022965,
+      "learning_rate": 8.992941740054418e-06,
+      "loss": 0.1951,
+      "step": 10740
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2854953153089266,
+      "learning_rate": 8.99130315787764e-06,
+      "loss": 0.1944,
+      "step": 10741
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.316795993709575,
+      "learning_rate": 8.989664603061818e-06,
+      "loss": 0.1785,
+      "step": 10742
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.858883046960564,
+      "learning_rate": 8.988026075651406e-06,
+      "loss": 0.1789,
+      "step": 10743
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7604592648249685,
+      "learning_rate": 8.986387575690839e-06,
+      "loss": 0.1655,
+      "step": 10744
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9078061178716357,
+      "learning_rate": 8.984749103224568e-06,
+      "loss": 0.2136,
+      "step": 10745
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7417197418190933,
+      "learning_rate": 8.983110658297033e-06,
+      "loss": 0.1641,
+      "step": 10746
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0208350376886899,
+      "learning_rate": 8.981472240952683e-06,
+      "loss": 0.1941,
+      "step": 10747
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0770493927619436,
+      "learning_rate": 8.979833851235952e-06,
+      "loss": 0.181,
+      "step": 10748
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2738105750757516,
+      "learning_rate": 8.978195489191288e-06,
+      "loss": 0.1891,
+      "step": 10749
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3508481124158724,
+      "learning_rate": 8.97655715486313e-06,
+      "loss": 0.1934,
+      "step": 10750
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9695221805853889,
+      "learning_rate": 8.974918848295914e-06,
+      "loss": 0.1796,
+      "step": 10751
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2229861774074742,
+      "learning_rate": 8.973280569534086e-06,
+      "loss": 0.1776,
+      "step": 10752
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9182531427775426,
+      "learning_rate": 8.971642318622076e-06,
+      "loss": 0.1921,
+      "step": 10753
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8634715877601445,
+      "learning_rate": 8.970004095604329e-06,
+      "loss": 0.177,
+      "step": 10754
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8793938254743472,
+      "learning_rate": 8.968365900525275e-06,
+      "loss": 0.1857,
+      "step": 10755
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9715431874467615,
+      "learning_rate": 8.966727733429362e-06,
+      "loss": 0.1889,
+      "step": 10756
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9027070073199415,
+      "learning_rate": 8.96508959436101e-06,
+      "loss": 0.1761,
+      "step": 10757
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1654803110367036,
+      "learning_rate": 8.963451483364665e-06,
+      "loss": 0.1877,
+      "step": 10758
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.002974639763267,
+      "learning_rate": 8.961813400484753e-06,
+      "loss": 0.186,
+      "step": 10759
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9182051316480477,
+      "learning_rate": 8.960175345765718e-06,
+      "loss": 0.1783,
+      "step": 10760
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1405988075873255,
+      "learning_rate": 8.95853731925198e-06,
+      "loss": 0.1648,
+      "step": 10761
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7103698934238166,
+      "learning_rate": 8.956899320987977e-06,
+      "loss": 0.1845,
+      "step": 10762
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9802152870441576,
+      "learning_rate": 8.955261351018138e-06,
+      "loss": 0.1906,
+      "step": 10763
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1017980194807244,
+      "learning_rate": 8.953623409386898e-06,
+      "loss": 0.1808,
+      "step": 10764
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8795915647215737,
+      "learning_rate": 8.951985496138679e-06,
+      "loss": 0.1646,
+      "step": 10765
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8785209703760559,
+      "learning_rate": 8.950347611317915e-06,
+      "loss": 0.2085,
+      "step": 10766
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8129306072891198,
+      "learning_rate": 8.94870975496903e-06,
+      "loss": 0.1803,
+      "step": 10767
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.5239210463731796,
+      "learning_rate": 8.94707192713645e-06,
+      "loss": 0.1723,
+      "step": 10768
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9476611032628168,
+      "learning_rate": 8.945434127864608e-06,
+      "loss": 0.1719,
+      "step": 10769
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7272317358149221,
+      "learning_rate": 8.94379635719792e-06,
+      "loss": 0.1572,
+      "step": 10770
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.484549886429762,
+      "learning_rate": 8.942158615180818e-06,
+      "loss": 0.1828,
+      "step": 10771
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1519687675896289,
+      "learning_rate": 8.940520901857722e-06,
+      "loss": 0.1835,
+      "step": 10772
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9810758635443829,
+      "learning_rate": 8.93888321727306e-06,
+      "loss": 0.1785,
+      "step": 10773
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7676471358143412,
+      "learning_rate": 8.937245561471247e-06,
+      "loss": 0.1815,
+      "step": 10774
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3673114707940612,
+      "learning_rate": 8.935607934496708e-06,
+      "loss": 0.1937,
+      "step": 10775
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.047631526141729,
+      "learning_rate": 8.933970336393864e-06,
+      "loss": 0.1663,
+      "step": 10776
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0940264402701274,
+      "learning_rate": 8.932332767207138e-06,
+      "loss": 0.1758,
+      "step": 10777
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7164913330577038,
+      "learning_rate": 8.930695226980942e-06,
+      "loss": 0.1674,
+      "step": 10778
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2698096854172805,
+      "learning_rate": 8.929057715759703e-06,
+      "loss": 0.1714,
+      "step": 10779
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.939943834510151,
+      "learning_rate": 8.92742023358783e-06,
+      "loss": 0.1808,
+      "step": 10780
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.2860861831418395,
+      "learning_rate": 8.925782780509745e-06,
+      "loss": 0.1863,
+      "step": 10781
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.6427009088574174,
+      "learning_rate": 8.92414535656986e-06,
+      "loss": 0.1753,
+      "step": 10782
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.307681135247615,
+      "learning_rate": 8.922507961812599e-06,
+      "loss": 0.1795,
+      "step": 10783
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0184505253136333,
+      "learning_rate": 8.920870596282368e-06,
+      "loss": 0.1885,
+      "step": 10784
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.040277807833205,
+      "learning_rate": 8.91923326002358e-06,
+      "loss": 0.1727,
+      "step": 10785
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0954583122249455,
+      "learning_rate": 8.917595953080656e-06,
+      "loss": 0.1815,
+      "step": 10786
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9721497354157779,
+      "learning_rate": 8.915958675497997e-06,
+      "loss": 0.1791,
+      "step": 10787
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9360953372361804,
+      "learning_rate": 8.914321427320024e-06,
+      "loss": 0.1773,
+      "step": 10788
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4024322278896943,
+      "learning_rate": 8.91268420859114e-06,
+      "loss": 0.1794,
+      "step": 10789
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9425720861101381,
+      "learning_rate": 8.911047019355763e-06,
+      "loss": 0.1712,
+      "step": 10790
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.639149419515541,
+      "learning_rate": 8.909409859658293e-06,
+      "loss": 0.1579,
+      "step": 10791
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0818029150587538,
+      "learning_rate": 8.907772729543145e-06,
+      "loss": 0.2021,
+      "step": 10792
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.409368473961371,
+      "learning_rate": 8.90613562905472e-06,
+      "loss": 0.17,
+      "step": 10793
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9131587173741051,
+      "learning_rate": 8.904498558237434e-06,
+      "loss": 0.1664,
+      "step": 10794
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9603316547360553,
+      "learning_rate": 8.902861517135682e-06,
+      "loss": 0.1652,
+      "step": 10795
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1837257062991657,
+      "learning_rate": 8.901224505793876e-06,
+      "loss": 0.1775,
+      "step": 10796
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8485619861393137,
+      "learning_rate": 8.899587524256414e-06,
+      "loss": 0.1721,
+      "step": 10797
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0226003384133822,
+      "learning_rate": 8.897950572567704e-06,
+      "loss": 0.1849,
+      "step": 10798
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9032907278532869,
+      "learning_rate": 8.896313650772144e-06,
+      "loss": 0.1971,
+      "step": 10799
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1700811482162825,
+      "learning_rate": 8.894676758914145e-06,
+      "loss": 0.1821,
+      "step": 10800
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3275915038076012,
+      "learning_rate": 8.893039897038097e-06,
+      "loss": 0.2052,
+      "step": 10801
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1586253791165513,
+      "learning_rate": 8.891403065188404e-06,
+      "loss": 0.1714,
+      "step": 10802
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0172789986237956,
+      "learning_rate": 8.88976626340947e-06,
+      "loss": 0.1748,
+      "step": 10803
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9277282769430485,
+      "learning_rate": 8.888129491745684e-06,
+      "loss": 0.1624,
+      "step": 10804
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7708823030251715,
+      "learning_rate": 8.88649275024145e-06,
+      "loss": 0.2086,
+      "step": 10805
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.458422843658293,
+      "learning_rate": 8.88485603894116e-06,
+      "loss": 0.2087,
+      "step": 10806
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.906221351091835,
+      "learning_rate": 8.883219357889218e-06,
+      "loss": 0.1811,
+      "step": 10807
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9681472452182767,
+      "learning_rate": 8.88158270713001e-06,
+      "loss": 0.1948,
+      "step": 10808
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0084461542273477,
+      "learning_rate": 8.87994608670794e-06,
+      "loss": 0.1857,
+      "step": 10809
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0272427474370724,
+      "learning_rate": 8.87830949666739e-06,
+      "loss": 0.1802,
+      "step": 10810
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0135686742121741,
+      "learning_rate": 8.87667293705276e-06,
+      "loss": 0.2147,
+      "step": 10811
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9696081518257766,
+      "learning_rate": 8.875036407908439e-06,
+      "loss": 0.1948,
+      "step": 10812
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0369553923642154,
+      "learning_rate": 8.873399909278821e-06,
+      "loss": 0.2007,
+      "step": 10813
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.083568254291667,
+      "learning_rate": 8.871763441208292e-06,
+      "loss": 0.1672,
+      "step": 10814
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9319992047044777,
+      "learning_rate": 8.870127003741245e-06,
+      "loss": 0.2072,
+      "step": 10815
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0894873647954664,
+      "learning_rate": 8.868490596922064e-06,
+      "loss": 0.1849,
+      "step": 10816
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8786214420077622,
+      "learning_rate": 8.866854220795144e-06,
+      "loss": 0.1663,
+      "step": 10817
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7965908011134126,
+      "learning_rate": 8.865217875404864e-06,
+      "loss": 0.1566,
+      "step": 10818
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.122771993981457,
+      "learning_rate": 8.863581560795614e-06,
+      "loss": 0.1663,
+      "step": 10819
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5355491727641055,
+      "learning_rate": 8.861945277011782e-06,
+      "loss": 0.1811,
+      "step": 10820
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0149352880199478,
+      "learning_rate": 8.860309024097744e-06,
+      "loss": 0.1896,
+      "step": 10821
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.15441128359665,
+      "learning_rate": 8.85867280209789e-06,
+      "loss": 0.19,
+      "step": 10822
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0242201683658139,
+      "learning_rate": 8.857036611056599e-06,
+      "loss": 0.1854,
+      "step": 10823
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0740247347675482,
+      "learning_rate": 8.855400451018256e-06,
+      "loss": 0.2088,
+      "step": 10824
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7181041405240665,
+      "learning_rate": 8.853764322027239e-06,
+      "loss": 0.1843,
+      "step": 10825
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.096691445469932,
+      "learning_rate": 8.852128224127931e-06,
+      "loss": 0.1879,
+      "step": 10826
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0328893655865132,
+      "learning_rate": 8.850492157364709e-06,
+      "loss": 0.1928,
+      "step": 10827
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1989050714425344,
+      "learning_rate": 8.848856121781953e-06,
+      "loss": 0.1722,
+      "step": 10828
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9972744783098172,
+      "learning_rate": 8.847220117424035e-06,
+      "loss": 0.1822,
+      "step": 10829
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.022753486245912,
+      "learning_rate": 8.845584144335344e-06,
+      "loss": 0.1798,
+      "step": 10830
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1575173533564185,
+      "learning_rate": 8.843948202560243e-06,
+      "loss": 0.1672,
+      "step": 10831
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1022591065507295,
+      "learning_rate": 8.842312292143115e-06,
+      "loss": 0.1953,
+      "step": 10832
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.074404331661858,
+      "learning_rate": 8.840676413128329e-06,
+      "loss": 0.2035,
+      "step": 10833
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9233624295011634,
+      "learning_rate": 8.839040565560264e-06,
+      "loss": 0.178,
+      "step": 10834
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1496136891813986,
+      "learning_rate": 8.837404749483285e-06,
+      "loss": 0.1718,
+      "step": 10835
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9570721283259703,
+      "learning_rate": 8.835768964941773e-06,
+      "loss": 0.194,
+      "step": 10836
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.000049498068047,
+      "learning_rate": 8.834133211980091e-06,
+      "loss": 0.1732,
+      "step": 10837
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9731055950131926,
+      "learning_rate": 8.83249749064261e-06,
+      "loss": 0.1644,
+      "step": 10838
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8763466169368689,
+      "learning_rate": 8.830861800973705e-06,
+      "loss": 0.1772,
+      "step": 10839
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9514424877296412,
+      "learning_rate": 8.829226143017735e-06,
+      "loss": 0.1578,
+      "step": 10840
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.072765787949337,
+      "learning_rate": 8.827590516819073e-06,
+      "loss": 0.1659,
+      "step": 10841
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.076068814527915,
+      "learning_rate": 8.825954922422084e-06,
+      "loss": 0.1831,
+      "step": 10842
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0256703193394343,
+      "learning_rate": 8.824319359871138e-06,
+      "loss": 0.1759,
+      "step": 10843
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9485220998311064,
+      "learning_rate": 8.822683829210591e-06,
+      "loss": 0.1959,
+      "step": 10844
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8247569049585812,
+      "learning_rate": 8.821048330484814e-06,
+      "loss": 0.1611,
+      "step": 10845
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9920482972212759,
+      "learning_rate": 8.819412863738165e-06,
+      "loss": 0.1807,
+      "step": 10846
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1803199349485702,
+      "learning_rate": 8.817777429015013e-06,
+      "loss": 0.1949,
+      "step": 10847
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8361111010536968,
+      "learning_rate": 8.816142026359711e-06,
+      "loss": 0.1904,
+      "step": 10848
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0204752253006832,
+      "learning_rate": 8.814506655816628e-06,
+      "loss": 0.2107,
+      "step": 10849
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8683096109613782,
+      "learning_rate": 8.812871317430114e-06,
+      "loss": 0.1856,
+      "step": 10850
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1192019964063356,
+      "learning_rate": 8.811236011244535e-06,
+      "loss": 0.1803,
+      "step": 10851
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8915818411611265,
+      "learning_rate": 8.809600737304245e-06,
+      "loss": 0.1713,
+      "step": 10852
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2922478281929404,
+      "learning_rate": 8.807965495653605e-06,
+      "loss": 0.1756,
+      "step": 10853
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1438370119106243,
+      "learning_rate": 8.806330286336966e-06,
+      "loss": 0.1737,
+      "step": 10854
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9674863155197395,
+      "learning_rate": 8.804695109398686e-06,
+      "loss": 0.1739,
+      "step": 10855
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8716670976435485,
+      "learning_rate": 8.803059964883121e-06,
+      "loss": 0.185,
+      "step": 10856
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8751877818085911,
+      "learning_rate": 8.801424852834617e-06,
+      "loss": 0.1864,
+      "step": 10857
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2810384916662814,
+      "learning_rate": 8.799789773297536e-06,
+      "loss": 0.2092,
+      "step": 10858
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.715194154870013,
+      "learning_rate": 8.79815472631622e-06,
+      "loss": 0.1817,
+      "step": 10859
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.957211967304257,
+      "learning_rate": 8.796519711935032e-06,
+      "loss": 0.1881,
+      "step": 10860
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.204919477838909,
+      "learning_rate": 8.794884730198309e-06,
+      "loss": 0.1823,
+      "step": 10861
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1796530871354738,
+      "learning_rate": 8.793249781150411e-06,
+      "loss": 0.216,
+      "step": 10862
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4713416941020465,
+      "learning_rate": 8.791614864835676e-06,
+      "loss": 0.1788,
+      "step": 10863
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9662566873754121,
+      "learning_rate": 8.789979981298457e-06,
+      "loss": 0.1884,
+      "step": 10864
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.356337540589256,
+      "learning_rate": 8.788345130583099e-06,
+      "loss": 0.1663,
+      "step": 10865
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7815317896632867,
+      "learning_rate": 8.786710312733952e-06,
+      "loss": 0.1656,
+      "step": 10866
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3871753218471037,
+      "learning_rate": 8.78507552779535e-06,
+      "loss": 0.1613,
+      "step": 10867
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5491548784204685,
+      "learning_rate": 8.783440775811646e-06,
+      "loss": 0.1772,
+      "step": 10868
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3689465680987378,
+      "learning_rate": 8.781806056827179e-06,
+      "loss": 0.1688,
+      "step": 10869
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2208364624605414,
+      "learning_rate": 8.780171370886295e-06,
+      "loss": 0.1632,
+      "step": 10870
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1374577018670873,
+      "learning_rate": 8.778536718033329e-06,
+      "loss": 0.2031,
+      "step": 10871
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.064865021837396,
+      "learning_rate": 8.776902098312622e-06,
+      "loss": 0.1848,
+      "step": 10872
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1104820656639132,
+      "learning_rate": 8.775267511768518e-06,
+      "loss": 0.1866,
+      "step": 10873
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9836540178033396,
+      "learning_rate": 8.773632958445348e-06,
+      "loss": 0.182,
+      "step": 10874
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7715029190728462,
+      "learning_rate": 8.771998438387458e-06,
+      "loss": 0.1629,
+      "step": 10875
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9027018301738213,
+      "learning_rate": 8.770363951639175e-06,
+      "loss": 0.1903,
+      "step": 10876
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0277633203112935,
+      "learning_rate": 8.768729498244841e-06,
+      "loss": 0.1688,
+      "step": 10877
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2082709822081161,
+      "learning_rate": 8.767095078248788e-06,
+      "loss": 0.1884,
+      "step": 10878
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4968536355503077,
+      "learning_rate": 8.765460691695353e-06,
+      "loss": 0.1846,
+      "step": 10879
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8723340832563421,
+      "learning_rate": 8.763826338628865e-06,
+      "loss": 0.1659,
+      "step": 10880
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0983506797438705,
+      "learning_rate": 8.762192019093658e-06,
+      "loss": 0.1557,
+      "step": 10881
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.665668563682617,
+      "learning_rate": 8.76055773313406e-06,
+      "loss": 0.1948,
+      "step": 10882
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1360990318938495,
+      "learning_rate": 8.758923480794407e-06,
+      "loss": 0.1858,
+      "step": 10883
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0668625114928212,
+      "learning_rate": 8.757289262119021e-06,
+      "loss": 0.1843,
+      "step": 10884
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5359141009129476,
+      "learning_rate": 8.755655077152237e-06,
+      "loss": 0.1967,
+      "step": 10885
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0503386454467547,
+      "learning_rate": 8.754020925938375e-06,
+      "loss": 0.1732,
+      "step": 10886
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0101469154598328,
+      "learning_rate": 8.752386808521774e-06,
+      "loss": 0.1775,
+      "step": 10887
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1936552339879511,
+      "learning_rate": 8.750752724946744e-06,
+      "loss": 0.1859,
+      "step": 10888
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.083771152185385,
+      "learning_rate": 8.749118675257622e-06,
+      "loss": 0.1906,
+      "step": 10889
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1874420465891897,
+      "learning_rate": 8.747484659498725e-06,
+      "loss": 0.1761,
+      "step": 10890
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9178432426616299,
+      "learning_rate": 8.745850677714373e-06,
+      "loss": 0.1913,
+      "step": 10891
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9437458995593613,
+      "learning_rate": 8.744216729948898e-06,
+      "loss": 0.1849,
+      "step": 10892
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0195680018807618,
+      "learning_rate": 8.74258281624661e-06,
+      "loss": 0.1661,
+      "step": 10893
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9040885233318499,
+      "learning_rate": 8.740948936651838e-06,
+      "loss": 0.1879,
+      "step": 10894
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2327915094808635,
+      "learning_rate": 8.739315091208893e-06,
+      "loss": 0.1977,
+      "step": 10895
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2229661061013206,
+      "learning_rate": 8.737681279962103e-06,
+      "loss": 0.1607,
+      "step": 10896
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8743363590350692,
+      "learning_rate": 8.736047502955776e-06,
+      "loss": 0.1775,
+      "step": 10897
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1094227460806871,
+      "learning_rate": 8.734413760234233e-06,
+      "loss": 0.1764,
+      "step": 10898
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3084970668744595,
+      "learning_rate": 8.732780051841785e-06,
+      "loss": 0.1815,
+      "step": 10899
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6092401166985617,
+      "learning_rate": 8.731146377822755e-06,
+      "loss": 0.1936,
+      "step": 10900
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0475620352949029,
+      "learning_rate": 8.729512738221448e-06,
+      "loss": 0.2011,
+      "step": 10901
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9645061349451889,
+      "learning_rate": 8.72787913308218e-06,
+      "loss": 0.1767,
+      "step": 10902
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0792427949313164,
+      "learning_rate": 8.726245562449261e-06,
+      "loss": 0.1862,
+      "step": 10903
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1064164071055667,
+      "learning_rate": 8.724612026367008e-06,
+      "loss": 0.188,
+      "step": 10904
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2077951842578907,
+      "learning_rate": 8.72297852487972e-06,
+      "loss": 0.2033,
+      "step": 10905
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9631462825847468,
+      "learning_rate": 8.721345058031718e-06,
+      "loss": 0.1773,
+      "step": 10906
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0126229462231389,
+      "learning_rate": 8.7197116258673e-06,
+      "loss": 0.1764,
+      "step": 10907
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.111625593594014,
+      "learning_rate": 8.718078228430775e-06,
+      "loss": 0.1739,
+      "step": 10908
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.136754878674941,
+      "learning_rate": 8.716444865766454e-06,
+      "loss": 0.1699,
+      "step": 10909
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0117881332904743,
+      "learning_rate": 8.714811537918634e-06,
+      "loss": 0.2098,
+      "step": 10910
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9200787097045166,
+      "learning_rate": 8.713178244931626e-06,
+      "loss": 0.1763,
+      "step": 10911
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1680809124305545,
+      "learning_rate": 8.71154498684973e-06,
+      "loss": 0.1767,
+      "step": 10912
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.994348721844077,
+      "learning_rate": 8.709911763717251e-06,
+      "loss": 0.1888,
+      "step": 10913
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.848085721687801,
+      "learning_rate": 8.708278575578485e-06,
+      "loss": 0.158,
+      "step": 10914
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2588476989472734,
+      "learning_rate": 8.706645422477739e-06,
+      "loss": 0.1978,
+      "step": 10915
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8154000665521712,
+      "learning_rate": 8.705012304459306e-06,
+      "loss": 0.1759,
+      "step": 10916
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0771918217834016,
+      "learning_rate": 8.703379221567491e-06,
+      "loss": 0.1771,
+      "step": 10917
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1485676136008025,
+      "learning_rate": 8.701746173846583e-06,
+      "loss": 0.1665,
+      "step": 10918
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4594060433552434,
+      "learning_rate": 8.70011316134089e-06,
+      "loss": 0.2013,
+      "step": 10919
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1035150344192732,
+      "learning_rate": 8.698480184094695e-06,
+      "loss": 0.1896,
+      "step": 10920
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.79687252374784,
+      "learning_rate": 8.696847242152301e-06,
+      "loss": 0.1588,
+      "step": 10921
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8751652874434868,
+      "learning_rate": 8.695214335557997e-06,
+      "loss": 0.1703,
+      "step": 10922
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.283108327990949,
+      "learning_rate": 8.693581464356083e-06,
+      "loss": 0.1891,
+      "step": 10923
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.30940533827719,
+      "learning_rate": 8.691948628590841e-06,
+      "loss": 0.1774,
+      "step": 10924
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2612497366281696,
+      "learning_rate": 8.690315828306566e-06,
+      "loss": 0.1899,
+      "step": 10925
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1632980454401722,
+      "learning_rate": 8.688683063547551e-06,
+      "loss": 0.1785,
+      "step": 10926
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2054669236060747,
+      "learning_rate": 8.68705033435808e-06,
+      "loss": 0.18,
+      "step": 10927
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8880601277908344,
+      "learning_rate": 8.685417640782444e-06,
+      "loss": 0.1753,
+      "step": 10928
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2596056413723435,
+      "learning_rate": 8.683784982864925e-06,
+      "loss": 0.1873,
+      "step": 10929
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0722207155737384,
+      "learning_rate": 8.682152360649819e-06,
+      "loss": 0.174,
+      "step": 10930
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0243942383285416,
+      "learning_rate": 8.6805197741814e-06,
+      "loss": 0.1703,
+      "step": 10931
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.04686861248303,
+      "learning_rate": 8.678887223503962e-06,
+      "loss": 0.1897,
+      "step": 10932
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.99105527675413,
+      "learning_rate": 8.677254708661775e-06,
+      "loss": 0.1561,
+      "step": 10933
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2436201373640272,
+      "learning_rate": 8.675622229699134e-06,
+      "loss": 0.1735,
+      "step": 10934
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0608032324133017,
+      "learning_rate": 8.67398978666031e-06,
+      "loss": 0.1764,
+      "step": 10935
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0845256537351942,
+      "learning_rate": 8.672357379589595e-06,
+      "loss": 0.1746,
+      "step": 10936
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1726740536790237,
+      "learning_rate": 8.670725008531255e-06,
+      "loss": 0.203,
+      "step": 10937
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8982383138239131,
+      "learning_rate": 8.669092673529577e-06,
+      "loss": 0.1936,
+      "step": 10938
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1467761252995048,
+      "learning_rate": 8.667460374628834e-06,
+      "loss": 0.2081,
+      "step": 10939
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.279402231030666,
+      "learning_rate": 8.665828111873307e-06,
+      "loss": 0.1743,
+      "step": 10940
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3727098270038447,
+      "learning_rate": 8.664195885307265e-06,
+      "loss": 0.1823,
+      "step": 10941
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6401924818960485,
+      "learning_rate": 8.662563694974983e-06,
+      "loss": 0.1907,
+      "step": 10942
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.123455051565522,
+      "learning_rate": 8.660931540920743e-06,
+      "loss": 0.1908,
+      "step": 10943
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.838274020948395,
+      "learning_rate": 8.659299423188803e-06,
+      "loss": 0.1799,
+      "step": 10944
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0452364543364976,
+      "learning_rate": 8.657667341823449e-06,
+      "loss": 0.1792,
+      "step": 10945
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8190999827722713,
+      "learning_rate": 8.656035296868938e-06,
+      "loss": 0.1693,
+      "step": 10946
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4522244007392222,
+      "learning_rate": 8.654403288369548e-06,
+      "loss": 0.166,
+      "step": 10947
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3287989010272494,
+      "learning_rate": 8.652771316369544e-06,
+      "loss": 0.1847,
+      "step": 10948
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0890163794301415,
+      "learning_rate": 8.651139380913197e-06,
+      "loss": 0.1977,
+      "step": 10949
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0666571494962853,
+      "learning_rate": 8.649507482044766e-06,
+      "loss": 0.1751,
+      "step": 10950
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8103740471153004,
+      "learning_rate": 8.647875619808523e-06,
+      "loss": 0.1765,
+      "step": 10951
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0313607099358908,
+      "learning_rate": 8.64624379424873e-06,
+      "loss": 0.1665,
+      "step": 10952
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0468097800952842,
+      "learning_rate": 8.644612005409654e-06,
+      "loss": 0.1795,
+      "step": 10953
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1057397636613135,
+      "learning_rate": 8.642980253335547e-06,
+      "loss": 0.2016,
+      "step": 10954
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8451430566992679,
+      "learning_rate": 8.641348538070683e-06,
+      "loss": 0.175,
+      "step": 10955
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7806195927151729,
+      "learning_rate": 8.639716859659312e-06,
+      "loss": 0.1901,
+      "step": 10956
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8798299824358545,
+      "learning_rate": 8.638085218145704e-06,
+      "loss": 0.1769,
+      "step": 10957
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5650750606191544,
+      "learning_rate": 8.63645361357411e-06,
+      "loss": 0.1877,
+      "step": 10958
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1279481372195475,
+      "learning_rate": 8.634822045988784e-06,
+      "loss": 0.1831,
+      "step": 10959
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2830148866132574,
+      "learning_rate": 8.633190515433992e-06,
+      "loss": 0.1901,
+      "step": 10960
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2855953567844727,
+      "learning_rate": 8.63155902195398e-06,
+      "loss": 0.1865,
+      "step": 10961
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8777387129501347,
+      "learning_rate": 8.62992756559301e-06,
+      "loss": 0.1688,
+      "step": 10962
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8700630299582908,
+      "learning_rate": 8.628296146395331e-06,
+      "loss": 0.1704,
+      "step": 10963
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.103315427349816,
+      "learning_rate": 8.626664764405196e-06,
+      "loss": 0.1724,
+      "step": 10964
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0764734465086485,
+      "learning_rate": 8.625033419666856e-06,
+      "loss": 0.1882,
+      "step": 10965
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0634970577767235,
+      "learning_rate": 8.623402112224567e-06,
+      "loss": 0.1972,
+      "step": 10966
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0115594755172888,
+      "learning_rate": 8.621770842122569e-06,
+      "loss": 0.2047,
+      "step": 10967
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0603757146365727,
+      "learning_rate": 8.620139609405115e-06,
+      "loss": 0.1647,
+      "step": 10968
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0003635283977106,
+      "learning_rate": 8.618508414116452e-06,
+      "loss": 0.1721,
+      "step": 10969
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1566118633153082,
+      "learning_rate": 8.616877256300829e-06,
+      "loss": 0.1855,
+      "step": 10970
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8767666422857754,
+      "learning_rate": 8.615246136002486e-06,
+      "loss": 0.1422,
+      "step": 10971
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9756034260032039,
+      "learning_rate": 8.613615053265673e-06,
+      "loss": 0.1666,
+      "step": 10972
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.054608091695435,
+      "learning_rate": 8.611984008134626e-06,
+      "loss": 0.2045,
+      "step": 10973
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0026363374092324,
+      "learning_rate": 8.610353000653592e-06,
+      "loss": 0.1857,
+      "step": 10974
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0371375238714557,
+      "learning_rate": 8.608722030866812e-06,
+      "loss": 0.1849,
+      "step": 10975
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9701005160752207,
+      "learning_rate": 8.607091098818528e-06,
+      "loss": 0.1735,
+      "step": 10976
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0497538225267664,
+      "learning_rate": 8.605460204552975e-06,
+      "loss": 0.1926,
+      "step": 10977
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9234896721451861,
+      "learning_rate": 8.60382934811439e-06,
+      "loss": 0.1928,
+      "step": 10978
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9119449965995741,
+      "learning_rate": 8.602198529547016e-06,
+      "loss": 0.1795,
+      "step": 10979
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.98527998422671,
+      "learning_rate": 8.600567748895083e-06,
+      "loss": 0.1869,
+      "step": 10980
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1053717143578772,
+      "learning_rate": 8.598937006202832e-06,
+      "loss": 0.1964,
+      "step": 10981
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.468621052173339,
+      "learning_rate": 8.59730630151449e-06,
+      "loss": 0.1734,
+      "step": 10982
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.952452351528307,
+      "learning_rate": 8.595675634874299e-06,
+      "loss": 0.1841,
+      "step": 10983
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1959355986594147,
+      "learning_rate": 8.594045006326481e-06,
+      "loss": 0.194,
+      "step": 10984
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9707998389931917,
+      "learning_rate": 8.592414415915275e-06,
+      "loss": 0.1864,
+      "step": 10985
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.382033789392686,
+      "learning_rate": 8.590783863684904e-06,
+      "loss": 0.2151,
+      "step": 10986
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5081807414030657,
+      "learning_rate": 8.589153349679602e-06,
+      "loss": 0.1716,
+      "step": 10987
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.261638494448365,
+      "learning_rate": 8.587522873943595e-06,
+      "loss": 0.1997,
+      "step": 10988
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.462716210009533,
+      "learning_rate": 8.585892436521113e-06,
+      "loss": 0.188,
+      "step": 10989
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4971491727761224,
+      "learning_rate": 8.584262037456374e-06,
+      "loss": 0.1831,
+      "step": 10990
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9349659575444129,
+      "learning_rate": 8.582631676793609e-06,
+      "loss": 0.1834,
+      "step": 10991
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9905856958238987,
+      "learning_rate": 8.581001354577037e-06,
+      "loss": 0.1975,
+      "step": 10992
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9028068531467937,
+      "learning_rate": 8.57937107085089e-06,
+      "loss": 0.1784,
+      "step": 10993
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9150736772460405,
+      "learning_rate": 8.577740825659379e-06,
+      "loss": 0.1729,
+      "step": 10994
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9936047580820794,
+      "learning_rate": 8.576110619046726e-06,
+      "loss": 0.2087,
+      "step": 10995
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2040788857862377,
+      "learning_rate": 8.574480451057158e-06,
+      "loss": 0.1729,
+      "step": 10996
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8647923487542004,
+      "learning_rate": 8.572850321734884e-06,
+      "loss": 0.1611,
+      "step": 10997
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0258661585027546,
+      "learning_rate": 8.571220231124129e-06,
+      "loss": 0.1889,
+      "step": 10998
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9404572152675427,
+      "learning_rate": 8.569590179269103e-06,
+      "loss": 0.1859,
+      "step": 10999
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3241112379396596,
+      "learning_rate": 8.567960166214026e-06,
+      "loss": 0.1805,
+      "step": 11000
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2502395125362873,
+      "learning_rate": 8.566330192003108e-06,
+      "loss": 0.1694,
+      "step": 11001
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3002683332018317,
+      "learning_rate": 8.564700256680568e-06,
+      "loss": 0.168,
+      "step": 11002
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.159540772474298,
+      "learning_rate": 8.563070360290611e-06,
+      "loss": 0.1874,
+      "step": 11003
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9030555139116054,
+      "learning_rate": 8.561440502877454e-06,
+      "loss": 0.1761,
+      "step": 11004
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.89022901095595,
+      "learning_rate": 8.559810684485302e-06,
+      "loss": 0.1863,
+      "step": 11005
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2859747477095844,
+      "learning_rate": 8.558180905158372e-06,
+      "loss": 0.1866,
+      "step": 11006
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1070866647981865,
+      "learning_rate": 8.55655116494086e-06,
+      "loss": 0.2022,
+      "step": 11007
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0914719993337787,
+      "learning_rate": 8.554921463876983e-06,
+      "loss": 0.1655,
+      "step": 11008
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2357096687659934,
+      "learning_rate": 8.55329180201094e-06,
+      "loss": 0.203,
+      "step": 11009
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1989549010584921,
+      "learning_rate": 8.551662179386944e-06,
+      "loss": 0.1722,
+      "step": 11010
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.088420664118368,
+      "learning_rate": 8.55003259604919e-06,
+      "loss": 0.1696,
+      "step": 11011
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0460955577335986,
+      "learning_rate": 8.54840305204188e-06,
+      "loss": 0.1639,
+      "step": 11012
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8399521974271892,
+      "learning_rate": 8.546773547409227e-06,
+      "loss": 0.168,
+      "step": 11013
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1979374294097083,
+      "learning_rate": 8.545144082195417e-06,
+      "loss": 0.1817,
+      "step": 11014
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3236855265368384,
+      "learning_rate": 8.54351465644466e-06,
+      "loss": 0.1872,
+      "step": 11015
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.563503817505762,
+      "learning_rate": 8.541885270201146e-06,
+      "loss": 0.1858,
+      "step": 11016
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9908741381061342,
+      "learning_rate": 8.540255923509081e-06,
+      "loss": 0.2116,
+      "step": 11017
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.184467292305344,
+      "learning_rate": 8.538626616412651e-06,
+      "loss": 0.1809,
+      "step": 11018
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2502518651558443,
+      "learning_rate": 8.536997348956065e-06,
+      "loss": 0.1994,
+      "step": 11019
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1074915593033396,
+      "learning_rate": 8.5353681211835e-06,
+      "loss": 0.1774,
+      "step": 11020
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3377825391510998,
+      "learning_rate": 8.533738933139162e-06,
+      "loss": 0.2249,
+      "step": 11021
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8265730358711466,
+      "learning_rate": 8.532109784867235e-06,
+      "loss": 0.1889,
+      "step": 11022
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.147024717062173,
+      "learning_rate": 8.530480676411919e-06,
+      "loss": 0.1922,
+      "step": 11023
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9714762838124623,
+      "learning_rate": 8.52885160781739e-06,
+      "loss": 0.1623,
+      "step": 11024
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4244966383242854,
+      "learning_rate": 8.52722257912785e-06,
+      "loss": 0.1986,
+      "step": 11025
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.406073801878377,
+      "learning_rate": 8.525593590387476e-06,
+      "loss": 0.1853,
+      "step": 11026
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9625221752761264,
+      "learning_rate": 8.523964641640466e-06,
+      "loss": 0.1741,
+      "step": 11027
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6320938086216492,
+      "learning_rate": 8.522335732930996e-06,
+      "loss": 0.1879,
+      "step": 11028
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8104481171462272,
+      "learning_rate": 8.52070686430325e-06,
+      "loss": 0.167,
+      "step": 11029
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0081096113409695,
+      "learning_rate": 8.519078035801415e-06,
+      "loss": 0.1695,
+      "step": 11030
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8971930904927773,
+      "learning_rate": 8.51744924746967e-06,
+      "loss": 0.1832,
+      "step": 11031
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9465995537342011,
+      "learning_rate": 8.515820499352203e-06,
+      "loss": 0.1698,
+      "step": 11032
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2279426320906355,
+      "learning_rate": 8.514191791493183e-06,
+      "loss": 0.1648,
+      "step": 11033
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.096622128478326,
+      "learning_rate": 8.512563123936796e-06,
+      "loss": 0.1885,
+      "step": 11034
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9072936887908979,
+      "learning_rate": 8.510934496727218e-06,
+      "loss": 0.186,
+      "step": 11035
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2281823546434798,
+      "learning_rate": 8.50930590990863e-06,
+      "loss": 0.1806,
+      "step": 11036
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7826850654708606,
+      "learning_rate": 8.5076773635252e-06,
+      "loss": 0.1985,
+      "step": 11037
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7724179175255349,
+      "learning_rate": 8.506048857621106e-06,
+      "loss": 0.1779,
+      "step": 11038
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1463623013180657,
+      "learning_rate": 8.50442039224052e-06,
+      "loss": 0.1776,
+      "step": 11039
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8442033414257292,
+      "learning_rate": 8.50279196742762e-06,
+      "loss": 0.1574,
+      "step": 11040
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7464494523210246,
+      "learning_rate": 8.501163583226567e-06,
+      "loss": 0.1837,
+      "step": 11041
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.03783109296676,
+      "learning_rate": 8.499535239681541e-06,
+      "loss": 0.1846,
+      "step": 11042
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.186059856182854,
+      "learning_rate": 8.497906936836704e-06,
+      "loss": 0.1718,
+      "step": 11043
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6469341544249463,
+      "learning_rate": 8.496278674736226e-06,
+      "loss": 0.1484,
+      "step": 11044
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.039933904642872,
+      "learning_rate": 8.494650453424272e-06,
+      "loss": 0.2039,
+      "step": 11045
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0120314170011517,
+      "learning_rate": 8.493022272945014e-06,
+      "loss": 0.1794,
+      "step": 11046
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0318104615833152,
+      "learning_rate": 8.49139413334261e-06,
+      "loss": 0.1917,
+      "step": 11047
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.042770661802242,
+      "learning_rate": 8.489766034661225e-06,
+      "loss": 0.2006,
+      "step": 11048
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0643753590812406,
+      "learning_rate": 8.488137976945023e-06,
+      "loss": 0.2061,
+      "step": 11049
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9171635729586125,
+      "learning_rate": 8.48650996023816e-06,
+      "loss": 0.1804,
+      "step": 11050
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8872809410093716,
+      "learning_rate": 8.484881984584803e-06,
+      "loss": 0.1678,
+      "step": 11051
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8891633300766486,
+      "learning_rate": 8.483254050029105e-06,
+      "loss": 0.1966,
+      "step": 11052
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0598293338182732,
+      "learning_rate": 8.481626156615231e-06,
+      "loss": 0.1981,
+      "step": 11053
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0567734225104917,
+      "learning_rate": 8.479998304387329e-06,
+      "loss": 0.1826,
+      "step": 11054
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7260869610454411,
+      "learning_rate": 8.478370493389563e-06,
+      "loss": 0.184,
+      "step": 11055
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1388976808393028,
+      "learning_rate": 8.47674272366608e-06,
+      "loss": 0.1711,
+      "step": 11056
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9298321738607229,
+      "learning_rate": 8.475114995261038e-06,
+      "loss": 0.1799,
+      "step": 11057
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6325973917833867,
+      "learning_rate": 8.473487308218585e-06,
+      "loss": 0.2003,
+      "step": 11058
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8499000723973356,
+      "learning_rate": 8.47185966258288e-06,
+      "loss": 0.1757,
+      "step": 11059
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8913225991349657,
+      "learning_rate": 8.470232058398063e-06,
+      "loss": 0.1657,
+      "step": 11060
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1076538002812075,
+      "learning_rate": 8.468604495708292e-06,
+      "loss": 0.1799,
+      "step": 11061
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9903752364490356,
+      "learning_rate": 8.466976974557706e-06,
+      "loss": 0.1719,
+      "step": 11062
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8559784269414673,
+      "learning_rate": 8.465349494990461e-06,
+      "loss": 0.1761,
+      "step": 11063
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3156245006771254,
+      "learning_rate": 8.463722057050696e-06,
+      "loss": 0.1677,
+      "step": 11064
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.943210871044513,
+      "learning_rate": 8.462094660782555e-06,
+      "loss": 0.1644,
+      "step": 11065
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9269504655641544,
+      "learning_rate": 8.460467306230187e-06,
+      "loss": 0.2061,
+      "step": 11066
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8738119366074828,
+      "learning_rate": 8.458839993437726e-06,
+      "loss": 0.2071,
+      "step": 11067
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8343929887803908,
+      "learning_rate": 8.457212722449322e-06,
+      "loss": 0.1685,
+      "step": 11068
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7622864170582623,
+      "learning_rate": 8.455585493309107e-06,
+      "loss": 0.1819,
+      "step": 11069
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8684169737198935,
+      "learning_rate": 8.453958306061223e-06,
+      "loss": 0.1858,
+      "step": 11070
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.294786211809637,
+      "learning_rate": 8.452331160749804e-06,
+      "loss": 0.1853,
+      "step": 11071
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8926249284110674,
+      "learning_rate": 8.450704057418996e-06,
+      "loss": 0.206,
+      "step": 11072
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3117790446383297,
+      "learning_rate": 8.449076996112924e-06,
+      "loss": 0.1691,
+      "step": 11073
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.869685388091908,
+      "learning_rate": 8.447449976875726e-06,
+      "loss": 0.1864,
+      "step": 11074
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7568117156894011,
+      "learning_rate": 8.445822999751536e-06,
+      "loss": 0.1803,
+      "step": 11075
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7174381127014311,
+      "learning_rate": 8.444196064784487e-06,
+      "loss": 0.1794,
+      "step": 11076
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9146953380622976,
+      "learning_rate": 8.442569172018703e-06,
+      "loss": 0.2146,
+      "step": 11077
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3439061263436667,
+      "learning_rate": 8.440942321498322e-06,
+      "loss": 0.1911,
+      "step": 11078
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.794186945195168,
+      "learning_rate": 8.439315513267465e-06,
+      "loss": 0.1677,
+      "step": 11079
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1085470147484742,
+      "learning_rate": 8.437688747370267e-06,
+      "loss": 0.1719,
+      "step": 11080
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7518099221003346,
+      "learning_rate": 8.43606202385085e-06,
+      "loss": 0.1733,
+      "step": 11081
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9856665916024756,
+      "learning_rate": 8.434435342753335e-06,
+      "loss": 0.1834,
+      "step": 11082
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.82839232903437,
+      "learning_rate": 8.432808704121854e-06,
+      "loss": 0.207,
+      "step": 11083
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3434290434642995,
+      "learning_rate": 8.431182108000522e-06,
+      "loss": 0.1664,
+      "step": 11084
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0161057228711095,
+      "learning_rate": 8.429555554433466e-06,
+      "loss": 0.1824,
+      "step": 11085
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8666489308505075,
+      "learning_rate": 8.427929043464802e-06,
+      "loss": 0.1797,
+      "step": 11086
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0791313722788873,
+      "learning_rate": 8.426302575138652e-06,
+      "loss": 0.1925,
+      "step": 11087
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9458821418898986,
+      "learning_rate": 8.424676149499133e-06,
+      "loss": 0.1704,
+      "step": 11088
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.266642190112638,
+      "learning_rate": 8.423049766590368e-06,
+      "loss": 0.1862,
+      "step": 11089
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0161348196522118,
+      "learning_rate": 8.42142342645646e-06,
+      "loss": 0.1985,
+      "step": 11090
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8030952803424762,
+      "learning_rate": 8.419797129141535e-06,
+      "loss": 0.174,
+      "step": 11091
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9562934368345417,
+      "learning_rate": 8.418170874689698e-06,
+      "loss": 0.1998,
+      "step": 11092
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7236144403554686,
+      "learning_rate": 8.416544663145073e-06,
+      "loss": 0.1681,
+      "step": 11093
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7208411969746172,
+      "learning_rate": 8.414918494551756e-06,
+      "loss": 0.1795,
+      "step": 11094
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0597353018878433,
+      "learning_rate": 8.41329236895387e-06,
+      "loss": 0.1828,
+      "step": 11095
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9124523581849326,
+      "learning_rate": 8.411666286395512e-06,
+      "loss": 0.1724,
+      "step": 11096
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9149754997881748,
+      "learning_rate": 8.410040246920799e-06,
+      "loss": 0.1765,
+      "step": 11097
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8623594055291297,
+      "learning_rate": 8.408414250573836e-06,
+      "loss": 0.1808,
+      "step": 11098
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6064872872688194,
+      "learning_rate": 8.406788297398722e-06,
+      "loss": 0.2015,
+      "step": 11099
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8449701891957053,
+      "learning_rate": 8.40516238743957e-06,
+      "loss": 0.1716,
+      "step": 11100
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8100056755848879,
+      "learning_rate": 8.403536520740474e-06,
+      "loss": 0.1733,
+      "step": 11101
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.958813365387284,
+      "learning_rate": 8.401910697345545e-06,
+      "loss": 0.2077,
+      "step": 11102
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7220376893487891,
+      "learning_rate": 8.400284917298873e-06,
+      "loss": 0.1976,
+      "step": 11103
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2678557390705776,
+      "learning_rate": 8.398659180644566e-06,
+      "loss": 0.2011,
+      "step": 11104
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1575418812541884,
+      "learning_rate": 8.397033487426717e-06,
+      "loss": 0.1832,
+      "step": 11105
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8109308744591895,
+      "learning_rate": 8.395407837689429e-06,
+      "loss": 0.186,
+      "step": 11106
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8894855499170428,
+      "learning_rate": 8.393782231476791e-06,
+      "loss": 0.1618,
+      "step": 11107
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9600703395707851,
+      "learning_rate": 8.392156668832904e-06,
+      "loss": 0.1905,
+      "step": 11108
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7495019141844758,
+      "learning_rate": 8.390531149801855e-06,
+      "loss": 0.1779,
+      "step": 11109
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8162713651475454,
+      "learning_rate": 8.38890567442774e-06,
+      "loss": 0.1886,
+      "step": 11110
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8336747419583592,
+      "learning_rate": 8.387280242754647e-06,
+      "loss": 0.1572,
+      "step": 11111
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2507564505391258,
+      "learning_rate": 8.385654854826674e-06,
+      "loss": 0.1887,
+      "step": 11112
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3709963705184494,
+      "learning_rate": 8.384029510687901e-06,
+      "loss": 0.1749,
+      "step": 11113
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2214279227277296,
+      "learning_rate": 8.38240421038242e-06,
+      "loss": 0.1783,
+      "step": 11114
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1108161355604176,
+      "learning_rate": 8.380778953954314e-06,
+      "loss": 0.1769,
+      "step": 11115
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8516403521844824,
+      "learning_rate": 8.379153741447674e-06,
+      "loss": 0.1869,
+      "step": 11116
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5159161379968633,
+      "learning_rate": 8.377528572906577e-06,
+      "loss": 0.1771,
+      "step": 11117
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1689390073254047,
+      "learning_rate": 8.375903448375109e-06,
+      "loss": 0.1902,
+      "step": 11118
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9519725613019046,
+      "learning_rate": 8.374278367897356e-06,
+      "loss": 0.1678,
+      "step": 11119
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1234418790420935,
+      "learning_rate": 8.37265333151739e-06,
+      "loss": 0.19,
+      "step": 11120
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8878798498587716,
+      "learning_rate": 8.371028339279294e-06,
+      "loss": 0.1852,
+      "step": 11121
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8025824103558893,
+      "learning_rate": 8.369403391227147e-06,
+      "loss": 0.1799,
+      "step": 11122
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9236287372839385,
+      "learning_rate": 8.367778487405028e-06,
+      "loss": 0.1707,
+      "step": 11123
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.91283308205594,
+      "learning_rate": 8.366153627857007e-06,
+      "loss": 0.1595,
+      "step": 11124
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9898467211961541,
+      "learning_rate": 8.364528812627164e-06,
+      "loss": 0.1914,
+      "step": 11125
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.496393384975324,
+      "learning_rate": 8.362904041759565e-06,
+      "loss": 0.1833,
+      "step": 11126
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1537680813326736,
+      "learning_rate": 8.361279315298289e-06,
+      "loss": 0.1994,
+      "step": 11127
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.041084597870637,
+      "learning_rate": 8.359654633287402e-06,
+      "loss": 0.1973,
+      "step": 11128
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0126117950716194,
+      "learning_rate": 8.358029995770979e-06,
+      "loss": 0.1643,
+      "step": 11129
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0569021104004925,
+      "learning_rate": 8.356405402793081e-06,
+      "loss": 0.1698,
+      "step": 11130
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.026689494861531,
+      "learning_rate": 8.354780854397782e-06,
+      "loss": 0.1609,
+      "step": 11131
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0032381083929853,
+      "learning_rate": 8.353156350629144e-06,
+      "loss": 0.1742,
+      "step": 11132
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9080894368846892,
+      "learning_rate": 8.351531891531235e-06,
+      "loss": 0.1905,
+      "step": 11133
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8909433565269967,
+      "learning_rate": 8.349907477148117e-06,
+      "loss": 0.1918,
+      "step": 11134
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9240676936852567,
+      "learning_rate": 8.348283107523848e-06,
+      "loss": 0.1654,
+      "step": 11135
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0737051026902602,
+      "learning_rate": 8.346658782702497e-06,
+      "loss": 0.1875,
+      "step": 11136
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2452176936417192,
+      "learning_rate": 8.345034502728117e-06,
+      "loss": 0.1878,
+      "step": 11137
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0940724955406431,
+      "learning_rate": 8.343410267644772e-06,
+      "loss": 0.1712,
+      "step": 11138
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2688468526066765,
+      "learning_rate": 8.341786077496513e-06,
+      "loss": 0.1796,
+      "step": 11139
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1625892788590617,
+      "learning_rate": 8.340161932327405e-06,
+      "loss": 0.1892,
+      "step": 11140
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8787948953079177,
+      "learning_rate": 8.338537832181494e-06,
+      "loss": 0.1679,
+      "step": 11141
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.433149178929281,
+      "learning_rate": 8.336913777102844e-06,
+      "loss": 0.1955,
+      "step": 11142
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0963719115563995,
+      "learning_rate": 8.335289767135497e-06,
+      "loss": 0.1678,
+      "step": 11143
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8733510208261461,
+      "learning_rate": 8.333665802323511e-06,
+      "loss": 0.1749,
+      "step": 11144
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0572328492621077,
+      "learning_rate": 8.33204188271093e-06,
+      "loss": 0.1763,
+      "step": 11145
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9655645976323488,
+      "learning_rate": 8.330418008341814e-06,
+      "loss": 0.1817,
+      "step": 11146
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0939906086573947,
+      "learning_rate": 8.328794179260199e-06,
+      "loss": 0.181,
+      "step": 11147
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.243143463741493,
+      "learning_rate": 8.327170395510137e-06,
+      "loss": 0.182,
+      "step": 11148
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0128360439150017,
+      "learning_rate": 8.325546657135673e-06,
+      "loss": 0.1882,
+      "step": 11149
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2114392175763768,
+      "learning_rate": 8.323922964180853e-06,
+      "loss": 0.1819,
+      "step": 11150
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.111825839568996,
+      "learning_rate": 8.322299316689717e-06,
+      "loss": 0.1881,
+      "step": 11151
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9514854567930829,
+      "learning_rate": 8.320675714706304e-06,
+      "loss": 0.1885,
+      "step": 11152
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.729532871086949,
+      "learning_rate": 8.319052158274659e-06,
+      "loss": 0.1857,
+      "step": 11153
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8476108959359467,
+      "learning_rate": 8.317428647438816e-06,
+      "loss": 0.1958,
+      "step": 11154
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4892354448303768,
+      "learning_rate": 8.31580518224282e-06,
+      "loss": 0.1747,
+      "step": 11155
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9187311605693036,
+      "learning_rate": 8.3141817627307e-06,
+      "loss": 0.1769,
+      "step": 11156
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.072172900218075,
+      "learning_rate": 8.312558388946497e-06,
+      "loss": 0.1733,
+      "step": 11157
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1420718748445113,
+      "learning_rate": 8.310935060934242e-06,
+      "loss": 0.1699,
+      "step": 11158
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.913240919562509,
+      "learning_rate": 8.30931177873797e-06,
+      "loss": 0.1633,
+      "step": 11159
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.051073321740484,
+      "learning_rate": 8.307688542401709e-06,
+      "loss": 0.1867,
+      "step": 11160
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8171469423370246,
+      "learning_rate": 8.306065351969494e-06,
+      "loss": 0.1793,
+      "step": 11161
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8922136251925066,
+      "learning_rate": 8.30444220748535e-06,
+      "loss": 0.2011,
+      "step": 11162
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9938900223149604,
+      "learning_rate": 8.302819108993311e-06,
+      "loss": 0.1852,
+      "step": 11163
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4287792786208626,
+      "learning_rate": 8.301196056537398e-06,
+      "loss": 0.1674,
+      "step": 11164
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7941065747802549,
+      "learning_rate": 8.299573050161637e-06,
+      "loss": 0.1573,
+      "step": 11165
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8134429355374344,
+      "learning_rate": 8.297950089910053e-06,
+      "loss": 0.1637,
+      "step": 11166
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0627650354860654,
+      "learning_rate": 8.29632717582667e-06,
+      "loss": 0.1735,
+      "step": 11167
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1882568419001993,
+      "learning_rate": 8.29470430795551e-06,
+      "loss": 0.2061,
+      "step": 11168
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9156524069604196,
+      "learning_rate": 8.293081486340587e-06,
+      "loss": 0.1754,
+      "step": 11169
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9397394528457855,
+      "learning_rate": 8.291458711025928e-06,
+      "loss": 0.1683,
+      "step": 11170
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9827543442383088,
+      "learning_rate": 8.289835982055546e-06,
+      "loss": 0.1969,
+      "step": 11171
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.001792440456909,
+      "learning_rate": 8.288213299473464e-06,
+      "loss": 0.1875,
+      "step": 11172
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.083625888509629,
+      "learning_rate": 8.286590663323689e-06,
+      "loss": 0.1876,
+      "step": 11173
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.220390786988613,
+      "learning_rate": 8.28496807365024e-06,
+      "loss": 0.1822,
+      "step": 11174
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9251997557124158,
+      "learning_rate": 8.283345530497128e-06,
+      "loss": 0.1911,
+      "step": 11175
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9765254071523586,
+      "learning_rate": 8.281723033908369e-06,
+      "loss": 0.1863,
+      "step": 11176
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4798687261341534,
+      "learning_rate": 8.280100583927965e-06,
+      "loss": 0.1853,
+      "step": 11177
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.813600417515956,
+      "learning_rate": 8.278478180599935e-06,
+      "loss": 0.1628,
+      "step": 11178
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0435886610526612,
+      "learning_rate": 8.276855823968278e-06,
+      "loss": 0.1875,
+      "step": 11179
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8112724641517984,
+      "learning_rate": 8.275233514077003e-06,
+      "loss": 0.1714,
+      "step": 11180
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3049418762007479,
+      "learning_rate": 8.273611250970118e-06,
+      "loss": 0.1742,
+      "step": 11181
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2920212964599433,
+      "learning_rate": 8.271989034691628e-06,
+      "loss": 0.1938,
+      "step": 11182
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9233125397139785,
+      "learning_rate": 8.270366865285528e-06,
+      "loss": 0.1923,
+      "step": 11183
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1376499852517181,
+      "learning_rate": 8.268744742795827e-06,
+      "loss": 0.1899,
+      "step": 11184
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.094168903904755,
+      "learning_rate": 8.26712266726652e-06,
+      "loss": 0.1805,
+      "step": 11185
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9333248112316166,
+      "learning_rate": 8.265500638741615e-06,
+      "loss": 0.1801,
+      "step": 11186
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.823286028623516,
+      "learning_rate": 8.263878657265099e-06,
+      "loss": 0.1779,
+      "step": 11187
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7560114879876358,
+      "learning_rate": 8.262256722880972e-06,
+      "loss": 0.1864,
+      "step": 11188
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.336291099600056,
+      "learning_rate": 8.260634835633232e-06,
+      "loss": 0.1729,
+      "step": 11189
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7182598354778278,
+      "learning_rate": 8.259012995565868e-06,
+      "loss": 0.1624,
+      "step": 11190
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7813488338610487,
+      "learning_rate": 8.257391202722877e-06,
+      "loss": 0.1972,
+      "step": 11191
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3591205075647002,
+      "learning_rate": 8.255769457148245e-06,
+      "loss": 0.197,
+      "step": 11192
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7968283715538477,
+      "learning_rate": 8.254147758885967e-06,
+      "loss": 0.1754,
+      "step": 11193
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7972409968258353,
+      "learning_rate": 8.252526107980027e-06,
+      "loss": 0.177,
+      "step": 11194
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.092037423266726,
+      "learning_rate": 8.25090450447442e-06,
+      "loss": 0.198,
+      "step": 11195
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.993110992833045,
+      "learning_rate": 8.249282948413123e-06,
+      "loss": 0.2043,
+      "step": 11196
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0829880769116116,
+      "learning_rate": 8.247661439840126e-06,
+      "loss": 0.1982,
+      "step": 11197
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.15991828521091,
+      "learning_rate": 8.246039978799409e-06,
+      "loss": 0.1755,
+      "step": 11198
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8749820174230736,
+      "learning_rate": 8.244418565334962e-06,
+      "loss": 0.1856,
+      "step": 11199
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0155370756575854,
+      "learning_rate": 8.242797199490757e-06,
+      "loss": 0.1975,
+      "step": 11200
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.954844012432073,
+      "learning_rate": 8.241175881310776e-06,
+      "loss": 0.15,
+      "step": 11201
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.393593559664671,
+      "learning_rate": 8.239554610839e-06,
+      "loss": 0.1904,
+      "step": 11202
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0309112780357765,
+      "learning_rate": 8.237933388119409e-06,
+      "loss": 0.1784,
+      "step": 11203
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2242804853714373,
+      "learning_rate": 8.236312213195972e-06,
+      "loss": 0.1842,
+      "step": 11204
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.092592283605871,
+      "learning_rate": 8.234691086112662e-06,
+      "loss": 0.1838,
+      "step": 11205
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7873812308441436,
+      "learning_rate": 8.23307000691346e-06,
+      "loss": 0.1828,
+      "step": 11206
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1405892274453442,
+      "learning_rate": 8.231448975642329e-06,
+      "loss": 0.2055,
+      "step": 11207
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9035227624336404,
+      "learning_rate": 8.229827992343251e-06,
+      "loss": 0.1863,
+      "step": 11208
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9515723670404533,
+      "learning_rate": 8.228207057060184e-06,
+      "loss": 0.1696,
+      "step": 11209
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2242351733938421,
+      "learning_rate": 8.226586169837103e-06,
+      "loss": 0.213,
+      "step": 11210
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0159569573403826,
+      "learning_rate": 8.224965330717967e-06,
+      "loss": 0.1848,
+      "step": 11211
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8746281789128181,
+      "learning_rate": 8.223344539746755e-06,
+      "loss": 0.1847,
+      "step": 11212
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9441233355578249,
+      "learning_rate": 8.221723796967416e-06,
+      "loss": 0.1997,
+      "step": 11213
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9518524036191406,
+      "learning_rate": 8.220103102423923e-06,
+      "loss": 0.1789,
+      "step": 11214
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.779774266436401,
+      "learning_rate": 8.21848245616023e-06,
+      "loss": 0.1759,
+      "step": 11215
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3450116809336112,
+      "learning_rate": 8.216861858220307e-06,
+      "loss": 0.1888,
+      "step": 11216
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.229616077127556,
+      "learning_rate": 8.215241308648102e-06,
+      "loss": 0.194,
+      "step": 11217
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9018917733087243,
+      "learning_rate": 8.213620807487583e-06,
+      "loss": 0.168,
+      "step": 11218
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.079239171539652,
+      "learning_rate": 8.212000354782695e-06,
+      "loss": 0.19,
+      "step": 11219
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8430855813690791,
+      "learning_rate": 8.210379950577398e-06,
+      "loss": 0.1706,
+      "step": 11220
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8230309276074811,
+      "learning_rate": 8.208759594915652e-06,
+      "loss": 0.1828,
+      "step": 11221
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.050508723669258,
+      "learning_rate": 8.207139287841397e-06,
+      "loss": 0.1818,
+      "step": 11222
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8161318295355628,
+      "learning_rate": 8.205519029398592e-06,
+      "loss": 0.1982,
+      "step": 11223
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8558635578387,
+      "learning_rate": 8.203898819631183e-06,
+      "loss": 0.1853,
+      "step": 11224
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8714033130034283,
+      "learning_rate": 8.202278658583125e-06,
+      "loss": 0.1865,
+      "step": 11225
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8827239327377759,
+      "learning_rate": 8.200658546298354e-06,
+      "loss": 0.1877,
+      "step": 11226
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2222996881129682,
+      "learning_rate": 8.199038482820824e-06,
+      "loss": 0.1913,
+      "step": 11227
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8298858935965866,
+      "learning_rate": 8.197418468194476e-06,
+      "loss": 0.179,
+      "step": 11228
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8088596354338242,
+      "learning_rate": 8.195798502463256e-06,
+      "loss": 0.1849,
+      "step": 11229
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3270774863738242,
+      "learning_rate": 8.194178585671102e-06,
+      "loss": 0.1787,
+      "step": 11230
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1797488127169842,
+      "learning_rate": 8.192558717861956e-06,
+      "loss": 0.1669,
+      "step": 11231
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.790298626038553,
+      "learning_rate": 8.190938899079756e-06,
+      "loss": 0.1583,
+      "step": 11232
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7010232969764986,
+      "learning_rate": 8.189319129368445e-06,
+      "loss": 0.1859,
+      "step": 11233
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9080372020704749,
+      "learning_rate": 8.18769940877195e-06,
+      "loss": 0.1892,
+      "step": 11234
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0889779620400961,
+      "learning_rate": 8.186079737334214e-06,
+      "loss": 0.1793,
+      "step": 11235
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.91466018716544,
+      "learning_rate": 8.184460115099165e-06,
+      "loss": 0.1735,
+      "step": 11236
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8871193225037902,
+      "learning_rate": 8.182840542110739e-06,
+      "loss": 0.1565,
+      "step": 11237
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0285746494638235,
+      "learning_rate": 8.181221018412868e-06,
+      "loss": 0.1718,
+      "step": 11238
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.029761854841097,
+      "learning_rate": 8.179601544049475e-06,
+      "loss": 0.187,
+      "step": 11239
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1239798590547327,
+      "learning_rate": 8.177982119064497e-06,
+      "loss": 0.1579,
+      "step": 11240
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.275798523162937,
+      "learning_rate": 8.176362743501853e-06,
+      "loss": 0.1851,
+      "step": 11241
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1377935943260051,
+      "learning_rate": 8.174743417405479e-06,
+      "loss": 0.1923,
+      "step": 11242
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7527062986335642,
+      "learning_rate": 8.173124140819285e-06,
+      "loss": 0.1969,
+      "step": 11243
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0965192679867608,
+      "learning_rate": 8.171504913787208e-06,
+      "loss": 0.1695,
+      "step": 11244
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8068912443926077,
+      "learning_rate": 8.169885736353158e-06,
+      "loss": 0.1896,
+      "step": 11245
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8361653287998111,
+      "learning_rate": 8.168266608561068e-06,
+      "loss": 0.1828,
+      "step": 11246
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2573793588577753,
+      "learning_rate": 8.166647530454843e-06,
+      "loss": 0.1468,
+      "step": 11247
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9078758393689891,
+      "learning_rate": 8.165028502078412e-06,
+      "loss": 0.1715,
+      "step": 11248
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7759762634961216,
+      "learning_rate": 8.163409523475682e-06,
+      "loss": 0.1895,
+      "step": 11249
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8809090715288981,
+      "learning_rate": 8.161790594690577e-06,
+      "loss": 0.1923,
+      "step": 11250
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.07713204744277,
+      "learning_rate": 8.160171715767002e-06,
+      "loss": 0.1645,
+      "step": 11251
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8341872450950087,
+      "learning_rate": 8.158552886748878e-06,
+      "loss": 0.1677,
+      "step": 11252
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.642327976245687,
+      "learning_rate": 8.156934107680106e-06,
+      "loss": 0.1772,
+      "step": 11253
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.822319012757017,
+      "learning_rate": 8.155315378604602e-06,
+      "loss": 0.1802,
+      "step": 11254
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0442418900022932,
+      "learning_rate": 8.153696699566272e-06,
+      "loss": 0.1798,
+      "step": 11255
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8988307675062374,
+      "learning_rate": 8.152078070609027e-06,
+      "loss": 0.1967,
+      "step": 11256
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1779903856225626,
+      "learning_rate": 8.150459491776765e-06,
+      "loss": 0.1847,
+      "step": 11257
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8690379175414268,
+      "learning_rate": 8.148840963113392e-06,
+      "loss": 0.1668,
+      "step": 11258
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.08253893833268,
+      "learning_rate": 8.147222484662817e-06,
+      "loss": 0.1828,
+      "step": 11259
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.802304156833683,
+      "learning_rate": 8.145604056468935e-06,
+      "loss": 0.177,
+      "step": 11260
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8294670880912641,
+      "learning_rate": 8.14398567857565e-06,
+      "loss": 0.1621,
+      "step": 11261
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7540643717119795,
+      "learning_rate": 8.142367351026853e-06,
+      "loss": 0.1481,
+      "step": 11262
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.26321843148518,
+      "learning_rate": 8.140749073866449e-06,
+      "loss": 0.1903,
+      "step": 11263
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3476642102613676,
+      "learning_rate": 8.139130847138332e-06,
+      "loss": 0.1826,
+      "step": 11264
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2458429990147137,
+      "learning_rate": 8.137512670886397e-06,
+      "loss": 0.1746,
+      "step": 11265
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.834594923921182,
+      "learning_rate": 8.135894545154533e-06,
+      "loss": 0.2108,
+      "step": 11266
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.129814446121592,
+      "learning_rate": 8.134276469986638e-06,
+      "loss": 0.1707,
+      "step": 11267
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.136819679803226,
+      "learning_rate": 8.132658445426595e-06,
+      "loss": 0.1569,
+      "step": 11268
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9788466534491648,
+      "learning_rate": 8.131040471518302e-06,
+      "loss": 0.173,
+      "step": 11269
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0440416991646695,
+      "learning_rate": 8.129422548305637e-06,
+      "loss": 0.1741,
+      "step": 11270
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1616215131566432,
+      "learning_rate": 8.127804675832494e-06,
+      "loss": 0.1861,
+      "step": 11271
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1600212126434577,
+      "learning_rate": 8.126186854142752e-06,
+      "loss": 0.2059,
+      "step": 11272
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6623548564513533,
+      "learning_rate": 8.124569083280303e-06,
+      "loss": 0.1855,
+      "step": 11273
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.185514362249397,
+      "learning_rate": 8.122951363289022e-06,
+      "loss": 0.1988,
+      "step": 11274
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.536505725783757,
+      "learning_rate": 8.12133369421279e-06,
+      "loss": 0.1921,
+      "step": 11275
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.047302910115353,
+      "learning_rate": 8.119716076095485e-06,
+      "loss": 0.2041,
+      "step": 11276
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1227914996846082,
+      "learning_rate": 8.118098508980989e-06,
+      "loss": 0.1765,
+      "step": 11277
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8278666307866703,
+      "learning_rate": 8.116480992913181e-06,
+      "loss": 0.1863,
+      "step": 11278
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2470534589076179,
+      "learning_rate": 8.114863527935929e-06,
+      "loss": 0.1943,
+      "step": 11279
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7763928476800777,
+      "learning_rate": 8.11324611409311e-06,
+      "loss": 0.1893,
+      "step": 11280
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7925818183893052,
+      "learning_rate": 8.111628751428595e-06,
+      "loss": 0.1634,
+      "step": 11281
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8936196251979709,
+      "learning_rate": 8.110011439986262e-06,
+      "loss": 0.163,
+      "step": 11282
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1152052160492107,
+      "learning_rate": 8.10839417980997e-06,
+      "loss": 0.1821,
+      "step": 11283
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8475383954634401,
+      "learning_rate": 8.106776970943597e-06,
+      "loss": 0.1888,
+      "step": 11284
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9449434158508707,
+      "learning_rate": 8.105159813431002e-06,
+      "loss": 0.1883,
+      "step": 11285
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0582156036341694,
+      "learning_rate": 8.103542707316058e-06,
+      "loss": 0.1877,
+      "step": 11286
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6783740262877755,
+      "learning_rate": 8.10192565264262e-06,
+      "loss": 0.1699,
+      "step": 11287
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2497198731430568,
+      "learning_rate": 8.100308649454563e-06,
+      "loss": 0.1756,
+      "step": 11288
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7127747891379077,
+      "learning_rate": 8.098691697795737e-06,
+      "loss": 0.1665,
+      "step": 11289
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8696899217535935,
+      "learning_rate": 8.097074797710007e-06,
+      "loss": 0.1796,
+      "step": 11290
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8810934002107135,
+      "learning_rate": 8.095457949241233e-06,
+      "loss": 0.1798,
+      "step": 11291
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.086874398334323,
+      "learning_rate": 8.093841152433265e-06,
+      "loss": 0.1778,
+      "step": 11292
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2700382193260717,
+      "learning_rate": 8.092224407329965e-06,
+      "loss": 0.1766,
+      "step": 11293
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0137503311295675,
+      "learning_rate": 8.090607713975182e-06,
+      "loss": 0.1728,
+      "step": 11294
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.280165089528958,
+      "learning_rate": 8.08899107241278e-06,
+      "loss": 0.1919,
+      "step": 11295
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2938704499783362,
+      "learning_rate": 8.087374482686598e-06,
+      "loss": 0.2076,
+      "step": 11296
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1512383258265375,
+      "learning_rate": 8.085757944840493e-06,
+      "loss": 0.2066,
+      "step": 11297
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7807085489114645,
+      "learning_rate": 8.08414145891831e-06,
+      "loss": 0.2115,
+      "step": 11298
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.05569501529721,
+      "learning_rate": 8.082525024963902e-06,
+      "loss": 0.1771,
+      "step": 11299
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9034811271517589,
+      "learning_rate": 8.080908643021107e-06,
+      "loss": 0.2346,
+      "step": 11300
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0660529540221337,
+      "learning_rate": 8.079292313133778e-06,
+      "loss": 0.1686,
+      "step": 11301
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2568586287865866,
+      "learning_rate": 8.077676035345748e-06,
+      "loss": 0.1725,
+      "step": 11302
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8573594071467343,
+      "learning_rate": 8.076059809700866e-06,
+      "loss": 0.1938,
+      "step": 11303
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0561172525104656,
+      "learning_rate": 8.07444363624297e-06,
+      "loss": 0.1567,
+      "step": 11304
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9525099461969996,
+      "learning_rate": 8.072827515015902e-06,
+      "loss": 0.1808,
+      "step": 11305
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1728442563294836,
+      "learning_rate": 8.071211446063495e-06,
+      "loss": 0.1774,
+      "step": 11306
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2680203679857898,
+      "learning_rate": 8.069595429429586e-06,
+      "loss": 0.1642,
+      "step": 11307
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9485894628866848,
+      "learning_rate": 8.067979465158013e-06,
+      "loss": 0.1728,
+      "step": 11308
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9960977651544729,
+      "learning_rate": 8.0663635532926e-06,
+      "loss": 0.1747,
+      "step": 11309
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3899001437254705,
+      "learning_rate": 8.06474769387719e-06,
+      "loss": 0.1401,
+      "step": 11310
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8600698725128217,
+      "learning_rate": 8.063131886955605e-06,
+      "loss": 0.1904,
+      "step": 11311
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9555677688751293,
+      "learning_rate": 8.061516132571679e-06,
+      "loss": 0.1964,
+      "step": 11312
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9597259602120783,
+      "learning_rate": 8.059900430769234e-06,
+      "loss": 0.185,
+      "step": 11313
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.111605309098347,
+      "learning_rate": 8.058284781592107e-06,
+      "loss": 0.1626,
+      "step": 11314
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2225140069439981,
+      "learning_rate": 8.056669185084108e-06,
+      "loss": 0.1598,
+      "step": 11315
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2846536877935064,
+      "learning_rate": 8.05505364128907e-06,
+      "loss": 0.1752,
+      "step": 11316
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8394495233121696,
+      "learning_rate": 8.053438150250808e-06,
+      "loss": 0.172,
+      "step": 11317
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9968223731163737,
+      "learning_rate": 8.051822712013151e-06,
+      "loss": 0.1919,
+      "step": 11318
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.123603368114008,
+      "learning_rate": 8.05020732661991e-06,
+      "loss": 0.1887,
+      "step": 11319
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7884087579364176,
+      "learning_rate": 8.048591994114906e-06,
+      "loss": 0.1702,
+      "step": 11320
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4011551740812698,
+      "learning_rate": 8.046976714541953e-06,
+      "loss": 0.164,
+      "step": 11321
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8102925771158213,
+      "learning_rate": 8.045361487944872e-06,
+      "loss": 0.2049,
+      "step": 11322
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7791722073468371,
+      "learning_rate": 8.043746314367466e-06,
+      "loss": 0.1966,
+      "step": 11323
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2138756039240277,
+      "learning_rate": 8.042131193853553e-06,
+      "loss": 0.1648,
+      "step": 11324
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1604943883973953,
+      "learning_rate": 8.04051612644694e-06,
+      "loss": 0.1767,
+      "step": 11325
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7364800594337658,
+      "learning_rate": 8.038901112191443e-06,
+      "loss": 0.157,
+      "step": 11326
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9303579297888586,
+      "learning_rate": 8.037286151130864e-06,
+      "loss": 0.1788,
+      "step": 11327
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9261735222055252,
+      "learning_rate": 8.035671243309005e-06,
+      "loss": 0.2088,
+      "step": 11328
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.146395985557706,
+      "learning_rate": 8.034056388769676e-06,
+      "loss": 0.1908,
+      "step": 11329
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8834531220910166,
+      "learning_rate": 8.032441587556676e-06,
+      "loss": 0.182,
+      "step": 11330
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7483513003591301,
+      "learning_rate": 8.030826839713814e-06,
+      "loss": 0.1804,
+      "step": 11331
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9241777250512747,
+      "learning_rate": 8.02921214528488e-06,
+      "loss": 0.1819,
+      "step": 11332
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8444863698741362,
+      "learning_rate": 8.02759750431368e-06,
+      "loss": 0.1852,
+      "step": 11333
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3926107355585513,
+      "learning_rate": 8.025982916844008e-06,
+      "loss": 0.1818,
+      "step": 11334
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8435975489819109,
+      "learning_rate": 8.024368382919665e-06,
+      "loss": 0.1855,
+      "step": 11335
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3527644387003557,
+      "learning_rate": 8.022753902584436e-06,
+      "loss": 0.206,
+      "step": 11336
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9316082958199031,
+      "learning_rate": 8.021139475882122e-06,
+      "loss": 0.1661,
+      "step": 11337
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7054655059974458,
+      "learning_rate": 8.01952510285651e-06,
+      "loss": 0.1804,
+      "step": 11338
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.908661234462497,
+      "learning_rate": 8.017910783551394e-06,
+      "loss": 0.176,
+      "step": 11339
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7743444896095891,
+      "learning_rate": 8.016296518010558e-06,
+      "loss": 0.1833,
+      "step": 11340
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9433594583525929,
+      "learning_rate": 8.014682306277792e-06,
+      "loss": 0.1766,
+      "step": 11341
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.122567249465266,
+      "learning_rate": 8.013068148396878e-06,
+      "loss": 0.1981,
+      "step": 11342
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8714340710402263,
+      "learning_rate": 8.011454044411606e-06,
+      "loss": 0.1819,
+      "step": 11343
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.572982517943364,
+      "learning_rate": 8.009839994365757e-06,
+      "loss": 0.1771,
+      "step": 11344
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.033687137192266,
+      "learning_rate": 8.008225998303107e-06,
+      "loss": 0.1911,
+      "step": 11345
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0309772673882498,
+      "learning_rate": 8.00661205626744e-06,
+      "loss": 0.2027,
+      "step": 11346
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9153445694050915,
+      "learning_rate": 8.004998168302531e-06,
+      "loss": 0.1847,
+      "step": 11347
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8893563536780098,
+      "learning_rate": 8.003384334452165e-06,
+      "loss": 0.1967,
+      "step": 11348
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9472506688885662,
+      "learning_rate": 8.001770554760107e-06,
+      "loss": 0.1759,
+      "step": 11349
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8856873542075918,
+      "learning_rate": 8.000156829270136e-06,
+      "loss": 0.1713,
+      "step": 11350
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8354670246876854,
+      "learning_rate": 7.998543158026025e-06,
+      "loss": 0.1784,
+      "step": 11351
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8998962124671674,
+      "learning_rate": 7.996929541071545e-06,
+      "loss": 0.189,
+      "step": 11352
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6963210550002428,
+      "learning_rate": 7.995315978450462e-06,
+      "loss": 0.1902,
+      "step": 11353
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7976868012803862,
+      "learning_rate": 7.993702470206547e-06,
+      "loss": 0.1982,
+      "step": 11354
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8559801835244099,
+      "learning_rate": 7.992089016383565e-06,
+      "loss": 0.1885,
+      "step": 11355
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9255002415299407,
+      "learning_rate": 7.990475617025286e-06,
+      "loss": 0.1789,
+      "step": 11356
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0647958227251417,
+      "learning_rate": 7.988862272175464e-06,
+      "loss": 0.1602,
+      "step": 11357
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0094849945715152,
+      "learning_rate": 7.987248981877872e-06,
+      "loss": 0.1838,
+      "step": 11358
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7584797503213979,
+      "learning_rate": 7.985635746176261e-06,
+      "loss": 0.1643,
+      "step": 11359
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9401437229638838,
+      "learning_rate": 7.984022565114396e-06,
+      "loss": 0.2103,
+      "step": 11360
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9411254100791999,
+      "learning_rate": 7.982409438736034e-06,
+      "loss": 0.1814,
+      "step": 11361
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.467728278191582,
+      "learning_rate": 7.980796367084925e-06,
+      "loss": 0.1642,
+      "step": 11362
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7544167417792813,
+      "learning_rate": 7.979183350204833e-06,
+      "loss": 0.1721,
+      "step": 11363
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7551276841736471,
+      "learning_rate": 7.977570388139503e-06,
+      "loss": 0.1781,
+      "step": 11364
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9763581032326041,
+      "learning_rate": 7.975957480932695e-06,
+      "loss": 0.1714,
+      "step": 11365
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7418208277674712,
+      "learning_rate": 7.974344628628151e-06,
+      "loss": 0.1668,
+      "step": 11366
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8384863204541823,
+      "learning_rate": 7.972731831269624e-06,
+      "loss": 0.1927,
+      "step": 11367
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8651724563330945,
+      "learning_rate": 7.97111908890086e-06,
+      "loss": 0.1944,
+      "step": 11368
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0535952841849294,
+      "learning_rate": 7.96950640156561e-06,
+      "loss": 0.2104,
+      "step": 11369
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.437367487342895,
+      "learning_rate": 7.967893769307608e-06,
+      "loss": 0.1829,
+      "step": 11370
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0217015755872205,
+      "learning_rate": 7.966281192170607e-06,
+      "loss": 0.1721,
+      "step": 11371
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9469785199313246,
+      "learning_rate": 7.964668670198339e-06,
+      "loss": 0.1768,
+      "step": 11372
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8351899718021233,
+      "learning_rate": 7.963056203434552e-06,
+      "loss": 0.1773,
+      "step": 11373
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8950493743963949,
+      "learning_rate": 7.961443791922975e-06,
+      "loss": 0.1957,
+      "step": 11374
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5883478193156267,
+      "learning_rate": 7.959831435707357e-06,
+      "loss": 0.1735,
+      "step": 11375
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0007651701611708,
+      "learning_rate": 7.958219134831423e-06,
+      "loss": 0.1675,
+      "step": 11376
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.183435973584841,
+      "learning_rate": 7.95660688933891e-06,
+      "loss": 0.161,
+      "step": 11377
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8598891102521062,
+      "learning_rate": 7.954994699273555e-06,
+      "loss": 0.1785,
+      "step": 11378
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0475770668796385,
+      "learning_rate": 7.953382564679078e-06,
+      "loss": 0.1632,
+      "step": 11379
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7656142930701677,
+      "learning_rate": 7.951770485599218e-06,
+      "loss": 0.1835,
+      "step": 11380
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7597755941350266,
+      "learning_rate": 7.950158462077697e-06,
+      "loss": 0.175,
+      "step": 11381
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1075597327894078,
+      "learning_rate": 7.948546494158247e-06,
+      "loss": 0.1563,
+      "step": 11382
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.919981907436329,
+      "learning_rate": 7.946934581884585e-06,
+      "loss": 0.1861,
+      "step": 11383
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8663663674300841,
+      "learning_rate": 7.945322725300444e-06,
+      "loss": 0.1851,
+      "step": 11384
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0200009715386287,
+      "learning_rate": 7.943710924449535e-06,
+      "loss": 0.1677,
+      "step": 11385
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7627219820305977,
+      "learning_rate": 7.942099179375585e-06,
+      "loss": 0.1654,
+      "step": 11386
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0960879832055765,
+      "learning_rate": 7.940487490122309e-06,
+      "loss": 0.1573,
+      "step": 11387
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9655957129166881,
+      "learning_rate": 7.93887585673343e-06,
+      "loss": 0.1622,
+      "step": 11388
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0674479446431646,
+      "learning_rate": 7.937264279252657e-06,
+      "loss": 0.1907,
+      "step": 11389
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7380337556095736,
+      "learning_rate": 7.935652757723709e-06,
+      "loss": 0.1808,
+      "step": 11390
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.027981513065775,
+      "learning_rate": 7.934041292190293e-06,
+      "loss": 0.2013,
+      "step": 11391
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7875171759545012,
+      "learning_rate": 7.93242988269613e-06,
+      "loss": 0.1519,
+      "step": 11392
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.210993758514946,
+      "learning_rate": 7.930818529284917e-06,
+      "loss": 0.1916,
+      "step": 11393
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7811719689055883,
+      "learning_rate": 7.92920723200037e-06,
+      "loss": 0.1937,
+      "step": 11394
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8521424976568306,
+      "learning_rate": 7.927595990886194e-06,
+      "loss": 0.1843,
+      "step": 11395
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.190810739178779,
+      "learning_rate": 7.925984805986096e-06,
+      "loss": 0.1903,
+      "step": 11396
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8073776952125132,
+      "learning_rate": 7.924373677343778e-06,
+      "loss": 0.1905,
+      "step": 11397
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9438162305535253,
+      "learning_rate": 7.922762605002938e-06,
+      "loss": 0.1794,
+      "step": 11398
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0413964995074405,
+      "learning_rate": 7.92115158900728e-06,
+      "loss": 0.1932,
+      "step": 11399
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8659254048488056,
+      "learning_rate": 7.9195406294005e-06,
+      "loss": 0.1756,
+      "step": 11400
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8148011974376181,
+      "learning_rate": 7.917929726226305e-06,
+      "loss": 0.1778,
+      "step": 11401
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8350748355467905,
+      "learning_rate": 7.916318879528377e-06,
+      "loss": 0.1828,
+      "step": 11402
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0380626632884626,
+      "learning_rate": 7.91470808935042e-06,
+      "loss": 0.1881,
+      "step": 11403
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.924415556504601,
+      "learning_rate": 7.913097355736122e-06,
+      "loss": 0.1809,
+      "step": 11404
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7720068220086019,
+      "learning_rate": 7.91148667872918e-06,
+      "loss": 0.1791,
+      "step": 11405
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8069146836895251,
+      "learning_rate": 7.909876058373275e-06,
+      "loss": 0.1721,
+      "step": 11406
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7883066391198847,
+      "learning_rate": 7.908265494712105e-06,
+      "loss": 0.2014,
+      "step": 11407
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8683896446322161,
+      "learning_rate": 7.906654987789346e-06,
+      "loss": 0.1462,
+      "step": 11408
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7508193646444697,
+      "learning_rate": 7.905044537648693e-06,
+      "loss": 0.1703,
+      "step": 11409
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5055920321230043,
+      "learning_rate": 7.903434144333824e-06,
+      "loss": 0.1907,
+      "step": 11410
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.888206103381393,
+      "learning_rate": 7.901823807888423e-06,
+      "loss": 0.1638,
+      "step": 11411
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7553488596562018,
+      "learning_rate": 7.900213528356167e-06,
+      "loss": 0.1638,
+      "step": 11412
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0805557840987123,
+      "learning_rate": 7.898603305780741e-06,
+      "loss": 0.1854,
+      "step": 11413
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9065778378349452,
+      "learning_rate": 7.89699314020582e-06,
+      "loss": 0.1778,
+      "step": 11414
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9282067156258447,
+      "learning_rate": 7.895383031675074e-06,
+      "loss": 0.1867,
+      "step": 11415
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9695888508795255,
+      "learning_rate": 7.893772980232186e-06,
+      "loss": 0.1935,
+      "step": 11416
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8176310094625524,
+      "learning_rate": 7.89216298592082e-06,
+      "loss": 0.1776,
+      "step": 11417
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8476605704866567,
+      "learning_rate": 7.89055304878466e-06,
+      "loss": 0.1779,
+      "step": 11418
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7864485731998299,
+      "learning_rate": 7.88894316886736e-06,
+      "loss": 0.1586,
+      "step": 11419
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8945343470837589,
+      "learning_rate": 7.8873333462126e-06,
+      "loss": 0.1903,
+      "step": 11420
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0553523338334387,
+      "learning_rate": 7.885723580864039e-06,
+      "loss": 0.1942,
+      "step": 11421
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4852096499603038,
+      "learning_rate": 7.884113872865352e-06,
+      "loss": 0.1909,
+      "step": 11422
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.918085514806113,
+      "learning_rate": 7.882504222260187e-06,
+      "loss": 0.2025,
+      "step": 11423
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7621380179939279,
+      "learning_rate": 7.880894629092222e-06,
+      "loss": 0.1904,
+      "step": 11424
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8141253588089128,
+      "learning_rate": 7.879285093405105e-06,
+      "loss": 0.198,
+      "step": 11425
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9091609362931211,
+      "learning_rate": 7.877675615242502e-06,
+      "loss": 0.1942,
+      "step": 11426
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.779397878754563,
+      "learning_rate": 7.876066194648066e-06,
+      "loss": 0.1855,
+      "step": 11427
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.894861345972049,
+      "learning_rate": 7.874456831665457e-06,
+      "loss": 0.1745,
+      "step": 11428
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8408203658866416,
+      "learning_rate": 7.872847526338324e-06,
+      "loss": 0.1742,
+      "step": 11429
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.811850660399226,
+      "learning_rate": 7.871238278710322e-06,
+      "loss": 0.1826,
+      "step": 11430
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8983352155401033,
+      "learning_rate": 7.869629088825105e-06,
+      "loss": 0.187,
+      "step": 11431
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.727772267058997,
+      "learning_rate": 7.868019956726318e-06,
+      "loss": 0.1926,
+      "step": 11432
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9780031004620156,
+      "learning_rate": 7.866410882457609e-06,
+      "loss": 0.1567,
+      "step": 11433
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9214707777852454,
+      "learning_rate": 7.864801866062624e-06,
+      "loss": 0.1843,
+      "step": 11434
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1192018977233658,
+      "learning_rate": 7.863192907585013e-06,
+      "loss": 0.1722,
+      "step": 11435
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7960680607754248,
+      "learning_rate": 7.861584007068411e-06,
+      "loss": 0.1832,
+      "step": 11436
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9839711925450026,
+      "learning_rate": 7.859975164556468e-06,
+      "loss": 0.17,
+      "step": 11437
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0211951625381592,
+      "learning_rate": 7.858366380092814e-06,
+      "loss": 0.1893,
+      "step": 11438
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.425161321405098,
+      "learning_rate": 7.856757653721097e-06,
+      "loss": 0.1898,
+      "step": 11439
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8515906228888953,
+      "learning_rate": 7.855148985484946e-06,
+      "loss": 0.1855,
+      "step": 11440
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8177327295630341,
+      "learning_rate": 7.853540375428006e-06,
+      "loss": 0.1662,
+      "step": 11441
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8708247841613003,
+      "learning_rate": 7.851931823593897e-06,
+      "loss": 0.1933,
+      "step": 11442
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7436403301338063,
+      "learning_rate": 7.850323330026264e-06,
+      "loss": 0.1819,
+      "step": 11443
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8529954048960217,
+      "learning_rate": 7.848714894768729e-06,
+      "loss": 0.1752,
+      "step": 11444
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.86165304634647,
+      "learning_rate": 7.847106517864927e-06,
+      "loss": 0.1675,
+      "step": 11445
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.795950397385831,
+      "learning_rate": 7.84549819935848e-06,
+      "loss": 0.1692,
+      "step": 11446
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8796257453564453,
+      "learning_rate": 7.843889939293017e-06,
+      "loss": 0.1604,
+      "step": 11447
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8178018265488591,
+      "learning_rate": 7.842281737712164e-06,
+      "loss": 0.1763,
+      "step": 11448
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2412486625603991,
+      "learning_rate": 7.840673594659535e-06,
+      "loss": 0.1635,
+      "step": 11449
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.138558639655599,
+      "learning_rate": 7.839065510178763e-06,
+      "loss": 0.151,
+      "step": 11450
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8575207245711235,
+      "learning_rate": 7.837457484313452e-06,
+      "loss": 0.1722,
+      "step": 11451
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1377203906398663,
+      "learning_rate": 7.835849517107237e-06,
+      "loss": 0.1826,
+      "step": 11452
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9724400883078111,
+      "learning_rate": 7.834241608603722e-06,
+      "loss": 0.1759,
+      "step": 11453
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.823551349347765,
+      "learning_rate": 7.83263375884653e-06,
+      "loss": 0.1969,
+      "step": 11454
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8901574635583533,
+      "learning_rate": 7.831025967879265e-06,
+      "loss": 0.1827,
+      "step": 11455
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9064639578916673,
+      "learning_rate": 7.829418235745547e-06,
+      "loss": 0.1743,
+      "step": 11456
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1197538740125437,
+      "learning_rate": 7.827810562488978e-06,
+      "loss": 0.1735,
+      "step": 11457
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7804009424828282,
+      "learning_rate": 7.826202948153174e-06,
+      "loss": 0.1559,
+      "step": 11458
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.784302834791359,
+      "learning_rate": 7.824595392781735e-06,
+      "loss": 0.1955,
+      "step": 11459
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.955136017244413,
+      "learning_rate": 7.822987896418269e-06,
+      "loss": 0.1802,
+      "step": 11460
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.11692735362752,
+      "learning_rate": 7.821380459106379e-06,
+      "loss": 0.1784,
+      "step": 11461
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.686381091202176,
+      "learning_rate": 7.81977308088967e-06,
+      "loss": 0.1852,
+      "step": 11462
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9285742766920333,
+      "learning_rate": 7.818165761811736e-06,
+      "loss": 0.1724,
+      "step": 11463
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5463402825839065,
+      "learning_rate": 7.81655850191618e-06,
+      "loss": 0.1889,
+      "step": 11464
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8834470298152093,
+      "learning_rate": 7.814951301246597e-06,
+      "loss": 0.2023,
+      "step": 11465
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9091701075412179,
+      "learning_rate": 7.813344159846588e-06,
+      "loss": 0.1946,
+      "step": 11466
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8394475643175278,
+      "learning_rate": 7.811737077759742e-06,
+      "loss": 0.1687,
+      "step": 11467
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6724614209465645,
+      "learning_rate": 7.810130055029646e-06,
+      "loss": 0.1806,
+      "step": 11468
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9378148453117767,
+      "learning_rate": 7.808523091699898e-06,
+      "loss": 0.1773,
+      "step": 11469
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5997825594312163,
+      "learning_rate": 7.806916187814084e-06,
+      "loss": 0.1728,
+      "step": 11470
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7187489821039615,
+      "learning_rate": 7.805309343415796e-06,
+      "loss": 0.1724,
+      "step": 11471
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7144109041965934,
+      "learning_rate": 7.803702558548611e-06,
+      "loss": 0.1796,
+      "step": 11472
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9252592793590034,
+      "learning_rate": 7.802095833256121e-06,
+      "loss": 0.2001,
+      "step": 11473
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.786306724546287,
+      "learning_rate": 7.800489167581903e-06,
+      "loss": 0.1819,
+      "step": 11474
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9248066516176838,
+      "learning_rate": 7.798882561569546e-06,
+      "loss": 0.1855,
+      "step": 11475
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1746878367894527,
+      "learning_rate": 7.797276015262619e-06,
+      "loss": 0.1822,
+      "step": 11476
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9795575556369396,
+      "learning_rate": 7.795669528704707e-06,
+      "loss": 0.1721,
+      "step": 11477
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9468774083981828,
+      "learning_rate": 7.794063101939381e-06,
+      "loss": 0.1758,
+      "step": 11478
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8166886570894248,
+      "learning_rate": 7.792456735010223e-06,
+      "loss": 0.1815,
+      "step": 11479
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9671715919770977,
+      "learning_rate": 7.790850427960795e-06,
+      "loss": 0.201,
+      "step": 11480
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.357756688548844,
+      "learning_rate": 7.789244180834679e-06,
+      "loss": 0.1946,
+      "step": 11481
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8662586324346514,
+      "learning_rate": 7.787637993675434e-06,
+      "loss": 0.1871,
+      "step": 11482
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.066321292841425,
+      "learning_rate": 7.786031866526636e-06,
+      "loss": 0.193,
+      "step": 11483
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8948737952595948,
+      "learning_rate": 7.784425799431852e-06,
+      "loss": 0.1686,
+      "step": 11484
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1269647961952738,
+      "learning_rate": 7.782819792434638e-06,
+      "loss": 0.174,
+      "step": 11485
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2486855535449515,
+      "learning_rate": 7.781213845578564e-06,
+      "loss": 0.181,
+      "step": 11486
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2255911988677148,
+      "learning_rate": 7.779607958907189e-06,
+      "loss": 0.2061,
+      "step": 11487
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6697046944408509,
+      "learning_rate": 7.778002132464077e-06,
+      "loss": 0.1637,
+      "step": 11488
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1222438318590247,
+      "learning_rate": 7.77639636629278e-06,
+      "loss": 0.2002,
+      "step": 11489
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8560548327618978,
+      "learning_rate": 7.774790660436857e-06,
+      "loss": 0.1879,
+      "step": 11490
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0547481846929991,
+      "learning_rate": 7.773185014939863e-06,
+      "loss": 0.2034,
+      "step": 11491
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9656593366959516,
+      "learning_rate": 7.771579429845353e-06,
+      "loss": 0.1883,
+      "step": 11492
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.933937310893644,
+      "learning_rate": 7.769973905196875e-06,
+      "loss": 0.1685,
+      "step": 11493
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0471187930082608,
+      "learning_rate": 7.768368441037983e-06,
+      "loss": 0.1507,
+      "step": 11494
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8376089379719395,
+      "learning_rate": 7.766763037412219e-06,
+      "loss": 0.1825,
+      "step": 11495
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9080888688549174,
+      "learning_rate": 7.765157694363138e-06,
+      "loss": 0.2009,
+      "step": 11496
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8024492902468562,
+      "learning_rate": 7.763552411934277e-06,
+      "loss": 0.1785,
+      "step": 11497
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7860320019469207,
+      "learning_rate": 7.761947190169188e-06,
+      "loss": 0.1718,
+      "step": 11498
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0387294443548292,
+      "learning_rate": 7.760342029111403e-06,
+      "loss": 0.1829,
+      "step": 11499
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9896381467940631,
+      "learning_rate": 7.758736928804469e-06,
+      "loss": 0.1924,
+      "step": 11500
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0018665717164907,
+      "learning_rate": 7.757131889291925e-06,
+      "loss": 0.1735,
+      "step": 11501
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9894239106183639,
+      "learning_rate": 7.7555269106173e-06,
+      "loss": 0.1751,
+      "step": 11502
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8985708137642188,
+      "learning_rate": 7.753921992824139e-06,
+      "loss": 0.1964,
+      "step": 11503
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1661604244942374,
+      "learning_rate": 7.752317135955966e-06,
+      "loss": 0.1713,
+      "step": 11504
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8612006619090127,
+      "learning_rate": 7.750712340056323e-06,
+      "loss": 0.1987,
+      "step": 11505
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7680521604795602,
+      "learning_rate": 7.74910760516873e-06,
+      "loss": 0.1889,
+      "step": 11506
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9533786515303825,
+      "learning_rate": 7.747502931336726e-06,
+      "loss": 0.1727,
+      "step": 11507
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7931110934670622,
+      "learning_rate": 7.745898318603826e-06,
+      "loss": 0.1815,
+      "step": 11508
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9576609653751136,
+      "learning_rate": 7.744293767013564e-06,
+      "loss": 0.1833,
+      "step": 11509
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9853274013225376,
+      "learning_rate": 7.742689276609459e-06,
+      "loss": 0.1607,
+      "step": 11510
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8886805308841748,
+      "learning_rate": 7.74108484743504e-06,
+      "loss": 0.1788,
+      "step": 11511
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9910320281706113,
+      "learning_rate": 7.739480479533818e-06,
+      "loss": 0.1914,
+      "step": 11512
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9545227317871207,
+      "learning_rate": 7.737876172949317e-06,
+      "loss": 0.1613,
+      "step": 11513
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7746215432805033,
+      "learning_rate": 7.73627192772505e-06,
+      "loss": 0.1784,
+      "step": 11514
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1258108748177553,
+      "learning_rate": 7.73466774390454e-06,
+      "loss": 0.1493,
+      "step": 11515
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1555646260024954,
+      "learning_rate": 7.73306362153129e-06,
+      "loss": 0.1875,
+      "step": 11516
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0687324728634058,
+      "learning_rate": 7.73145956064882e-06,
+      "loss": 0.1546,
+      "step": 11517
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.106528469273153,
+      "learning_rate": 7.72985556130064e-06,
+      "loss": 0.1887,
+      "step": 11518
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.837873803762906,
+      "learning_rate": 7.728251623530253e-06,
+      "loss": 0.1703,
+      "step": 11519
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9792314296285167,
+      "learning_rate": 7.726647747381171e-06,
+      "loss": 0.1769,
+      "step": 11520
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.036453018636925,
+      "learning_rate": 7.725043932896895e-06,
+      "loss": 0.1763,
+      "step": 11521
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6120548814324298,
+      "learning_rate": 7.723440180120932e-06,
+      "loss": 0.1884,
+      "step": 11522
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.952343846906782,
+      "learning_rate": 7.72183648909678e-06,
+      "loss": 0.1594,
+      "step": 11523
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8464787482722366,
+      "learning_rate": 7.720232859867946e-06,
+      "loss": 0.1709,
+      "step": 11524
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.84795764242471,
+      "learning_rate": 7.71862929247792e-06,
+      "loss": 0.1867,
+      "step": 11525
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0405235420112218,
+      "learning_rate": 7.717025786970208e-06,
+      "loss": 0.176,
+      "step": 11526
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.781721864117291,
+      "learning_rate": 7.715422343388296e-06,
+      "loss": 0.1801,
+      "step": 11527
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9009036342142364,
+      "learning_rate": 7.713818961775686e-06,
+      "loss": 0.1775,
+      "step": 11528
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.889676087190506,
+      "learning_rate": 7.712215642175862e-06,
+      "loss": 0.1806,
+      "step": 11529
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.537611419687043,
+      "learning_rate": 7.710612384632321e-06,
+      "loss": 0.1986,
+      "step": 11530
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9885073435263368,
+      "learning_rate": 7.709009189188546e-06,
+      "loss": 0.2099,
+      "step": 11531
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.157845104013614,
+      "learning_rate": 7.70740605588803e-06,
+      "loss": 0.1774,
+      "step": 11532
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5695577270079757,
+      "learning_rate": 7.70580298477425e-06,
+      "loss": 0.1729,
+      "step": 11533
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7935269584611251,
+      "learning_rate": 7.704199975890698e-06,
+      "loss": 0.1811,
+      "step": 11534
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.302641620541749,
+      "learning_rate": 7.702597029280848e-06,
+      "loss": 0.1637,
+      "step": 11535
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0267924968478008,
+      "learning_rate": 7.700994144988183e-06,
+      "loss": 0.1992,
+      "step": 11536
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0705791649139917,
+      "learning_rate": 7.699391323056184e-06,
+      "loss": 0.1867,
+      "step": 11537
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0443080061831171,
+      "learning_rate": 7.697788563528323e-06,
+      "loss": 0.1743,
+      "step": 11538
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.91455270222627,
+      "learning_rate": 7.696185866448079e-06,
+      "loss": 0.1781,
+      "step": 11539
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9676560323614188,
+      "learning_rate": 7.694583231858921e-06,
+      "loss": 0.1898,
+      "step": 11540
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8959806001854843,
+      "learning_rate": 7.692980659804327e-06,
+      "loss": 0.1763,
+      "step": 11541
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0203472241064597,
+      "learning_rate": 7.691378150327759e-06,
+      "loss": 0.1764,
+      "step": 11542
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1974668086090456,
+      "learning_rate": 7.689775703472691e-06,
+      "loss": 0.1866,
+      "step": 11543
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7475706747831443,
+      "learning_rate": 7.688173319282586e-06,
+      "loss": 0.1976,
+      "step": 11544
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.935656524326181,
+      "learning_rate": 7.686570997800914e-06,
+      "loss": 0.1605,
+      "step": 11545
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8084136464393837,
+      "learning_rate": 7.68496873907113e-06,
+      "loss": 0.2016,
+      "step": 11546
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9290286115022512,
+      "learning_rate": 7.683366543136703e-06,
+      "loss": 0.1894,
+      "step": 11547
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4974991685353187,
+      "learning_rate": 7.681764410041087e-06,
+      "loss": 0.1957,
+      "step": 11548
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8232639356455898,
+      "learning_rate": 7.680162339827744e-06,
+      "loss": 0.1841,
+      "step": 11549
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9511264667655934,
+      "learning_rate": 7.678560332540126e-06,
+      "loss": 0.1775,
+      "step": 11550
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9769977386312791,
+      "learning_rate": 7.676958388221693e-06,
+      "loss": 0.1797,
+      "step": 11551
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.167911556068729,
+      "learning_rate": 7.675356506915892e-06,
+      "loss": 0.1962,
+      "step": 11552
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8463201079565178,
+      "learning_rate": 7.67375468866618e-06,
+      "loss": 0.1771,
+      "step": 11553
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2272626867654062,
+      "learning_rate": 7.672152933516005e-06,
+      "loss": 0.1973,
+      "step": 11554
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.094380301002933,
+      "learning_rate": 7.670551241508809e-06,
+      "loss": 0.1678,
+      "step": 11555
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.915630672766507,
+      "learning_rate": 7.668949612688044e-06,
+      "loss": 0.1837,
+      "step": 11556
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7441259642618644,
+      "learning_rate": 7.667348047097151e-06,
+      "loss": 0.1744,
+      "step": 11557
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.761767486300642,
+      "learning_rate": 7.665746544779577e-06,
+      "loss": 0.172,
+      "step": 11558
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6978167876222314,
+      "learning_rate": 7.664145105778755e-06,
+      "loss": 0.1646,
+      "step": 11559
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8980322487996942,
+      "learning_rate": 7.662543730138136e-06,
+      "loss": 0.1795,
+      "step": 11560
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1376196282122772,
+      "learning_rate": 7.660942417901145e-06,
+      "loss": 0.1729,
+      "step": 11561
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.372232682551164,
+      "learning_rate": 7.659341169111222e-06,
+      "loss": 0.1562,
+      "step": 11562
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9663702113761,
+      "learning_rate": 7.657739983811803e-06,
+      "loss": 0.2076,
+      "step": 11563
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0129999221068482,
+      "learning_rate": 7.656138862046323e-06,
+      "loss": 0.1767,
+      "step": 11564
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9004518055692233,
+      "learning_rate": 7.654537803858205e-06,
+      "loss": 0.186,
+      "step": 11565
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8210445742427821,
+      "learning_rate": 7.652936809290883e-06,
+      "loss": 0.2024,
+      "step": 11566
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9394699808210847,
+      "learning_rate": 7.65133587838778e-06,
+      "loss": 0.1847,
+      "step": 11567
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9609354948526633,
+      "learning_rate": 7.649735011192329e-06,
+      "loss": 0.1912,
+      "step": 11568
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.654273035810828,
+      "learning_rate": 7.648134207747944e-06,
+      "loss": 0.1791,
+      "step": 11569
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8138671952721602,
+      "learning_rate": 7.646533468098054e-06,
+      "loss": 0.1994,
+      "step": 11570
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9362668301075935,
+      "learning_rate": 7.644932792286078e-06,
+      "loss": 0.2046,
+      "step": 11571
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8749552721286621,
+      "learning_rate": 7.64333218035543e-06,
+      "loss": 0.1914,
+      "step": 11572
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8649832225609275,
+      "learning_rate": 7.64173163234953e-06,
+      "loss": 0.1579,
+      "step": 11573
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.171430802862261,
+      "learning_rate": 7.640131148311791e-06,
+      "loss": 0.19,
+      "step": 11574
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2259477037304825,
+      "learning_rate": 7.638530728285633e-06,
+      "loss": 0.1808,
+      "step": 11575
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4841824061106357,
+      "learning_rate": 7.636930372314457e-06,
+      "loss": 0.19,
+      "step": 11576
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9639021119081354,
+      "learning_rate": 7.635330080441684e-06,
+      "loss": 0.1811,
+      "step": 11577
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9701950679898839,
+      "learning_rate": 7.633729852710711e-06,
+      "loss": 0.1714,
+      "step": 11578
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0768229910505707,
+      "learning_rate": 7.632129689164951e-06,
+      "loss": 0.2228,
+      "step": 11579
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3353356856633012,
+      "learning_rate": 7.630529589847807e-06,
+      "loss": 0.1921,
+      "step": 11580
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3622609087993482,
+      "learning_rate": 7.628929554802683e-06,
+      "loss": 0.1911,
+      "step": 11581
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8975720323589307,
+      "learning_rate": 7.627329584072976e-06,
+      "loss": 0.1796,
+      "step": 11582
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9448305446099222,
+      "learning_rate": 7.625729677702089e-06,
+      "loss": 0.197,
+      "step": 11583
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.414447297300965,
+      "learning_rate": 7.624129835733418e-06,
+      "loss": 0.1847,
+      "step": 11584
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1403047850573143,
+      "learning_rate": 7.622530058210363e-06,
+      "loss": 0.1705,
+      "step": 11585
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7769183214743727,
+      "learning_rate": 7.62093034517631e-06,
+      "loss": 0.1736,
+      "step": 11586
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8832212847030111,
+      "learning_rate": 7.619330696674658e-06,
+      "loss": 0.1749,
+      "step": 11587
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0324852829353546,
+      "learning_rate": 7.6177311127487984e-06,
+      "loss": 0.17,
+      "step": 11588
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0803947078290193,
+      "learning_rate": 7.616131593442111e-06,
+      "loss": 0.1657,
+      "step": 11589
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3593432430994428,
+      "learning_rate": 7.614532138797994e-06,
+      "loss": 0.2016,
+      "step": 11590
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1694049472944021,
+      "learning_rate": 7.6129327488598225e-06,
+      "loss": 0.1627,
+      "step": 11591
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9088413784789771,
+      "learning_rate": 7.611333423670988e-06,
+      "loss": 0.184,
+      "step": 11592
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2524174806149122,
+      "learning_rate": 7.609734163274867e-06,
+      "loss": 0.1803,
+      "step": 11593
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8086586822961561,
+      "learning_rate": 7.608134967714846e-06,
+      "loss": 0.1657,
+      "step": 11594
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0821545601165399,
+      "learning_rate": 7.606535837034295e-06,
+      "loss": 0.1814,
+      "step": 11595
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1909757173673818,
+      "learning_rate": 7.604936771276596e-06,
+      "loss": 0.1925,
+      "step": 11596
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1089161855517706,
+      "learning_rate": 7.603337770485122e-06,
+      "loss": 0.2052,
+      "step": 11597
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9012722774072845,
+      "learning_rate": 7.601738834703249e-06,
+      "loss": 0.1869,
+      "step": 11598
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7300428580712354,
+      "learning_rate": 7.600139963974341e-06,
+      "loss": 0.1819,
+      "step": 11599
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9574724234455277,
+      "learning_rate": 7.598541158341774e-06,
+      "loss": 0.1799,
+      "step": 11600
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2387333692896372,
+      "learning_rate": 7.5969424178489134e-06,
+      "loss": 0.17,
+      "step": 11601
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.907897336158818,
+      "learning_rate": 7.5953437425391284e-06,
+      "loss": 0.1887,
+      "step": 11602
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4548819371977963,
+      "learning_rate": 7.593745132455776e-06,
+      "loss": 0.1931,
+      "step": 11603
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0247766606758641,
+      "learning_rate": 7.592146587642227e-06,
+      "loss": 0.1961,
+      "step": 11604
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.237579976507833,
+      "learning_rate": 7.5905481081418365e-06,
+      "loss": 0.1764,
+      "step": 11605
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9641963466945946,
+      "learning_rate": 7.588949693997962e-06,
+      "loss": 0.1806,
+      "step": 11606
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0322631441623458,
+      "learning_rate": 7.587351345253968e-06,
+      "loss": 0.1658,
+      "step": 11607
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8648225298624148,
+      "learning_rate": 7.585753061953199e-06,
+      "loss": 0.1559,
+      "step": 11608
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8683653108191068,
+      "learning_rate": 7.584154844139019e-06,
+      "loss": 0.1815,
+      "step": 11609
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.291889450017652,
+      "learning_rate": 7.582556691854772e-06,
+      "loss": 0.1659,
+      "step": 11610
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.074740180118149,
+      "learning_rate": 7.580958605143816e-06,
+      "loss": 0.1688,
+      "step": 11611
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8152071906384597,
+      "learning_rate": 7.579360584049489e-06,
+      "loss": 0.1947,
+      "step": 11612
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1801412874720607,
+      "learning_rate": 7.577762628615146e-06,
+      "loss": 0.1859,
+      "step": 11613
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.034604716980372,
+      "learning_rate": 7.576164738884126e-06,
+      "loss": 0.1831,
+      "step": 11614
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9786064729118367,
+      "learning_rate": 7.574566914899779e-06,
+      "loss": 0.1892,
+      "step": 11615
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1607546685118113,
+      "learning_rate": 7.572969156705437e-06,
+      "loss": 0.1854,
+      "step": 11616
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8754173581044404,
+      "learning_rate": 7.571371464344448e-06,
+      "loss": 0.1663,
+      "step": 11617
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9708501178409704,
+      "learning_rate": 7.5697738378601406e-06,
+      "loss": 0.2012,
+      "step": 11618
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0361952174520443,
+      "learning_rate": 7.568176277295858e-06,
+      "loss": 0.1848,
+      "step": 11619
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.31195992549576,
+      "learning_rate": 7.566578782694928e-06,
+      "loss": 0.1803,
+      "step": 11620
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0422363977834639,
+      "learning_rate": 7.564981354100691e-06,
+      "loss": 0.1711,
+      "step": 11621
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7208745670437828,
+      "learning_rate": 7.563383991556468e-06,
+      "loss": 0.1681,
+      "step": 11622
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4976767554787114,
+      "learning_rate": 7.561786695105593e-06,
+      "loss": 0.1692,
+      "step": 11623
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1242864380247015,
+      "learning_rate": 7.5601894647913955e-06,
+      "loss": 0.1749,
+      "step": 11624
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1508597414092376,
+      "learning_rate": 7.55859230065719e-06,
+      "loss": 0.165,
+      "step": 11625
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9724639392749773,
+      "learning_rate": 7.556995202746311e-06,
+      "loss": 0.1606,
+      "step": 11626
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7715827756096528,
+      "learning_rate": 7.555398171102072e-06,
+      "loss": 0.1771,
+      "step": 11627
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3122835713518486,
+      "learning_rate": 7.5538012057677985e-06,
+      "loss": 0.1746,
+      "step": 11628
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5803497435595275,
+      "learning_rate": 7.5522043067868034e-06,
+      "loss": 0.1704,
+      "step": 11629
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1103465528131011,
+      "learning_rate": 7.550607474202407e-06,
+      "loss": 0.1813,
+      "step": 11630
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.998855733673789,
+      "learning_rate": 7.549010708057919e-06,
+      "loss": 0.1905,
+      "step": 11631
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5212565968715075,
+      "learning_rate": 7.5474140083966544e-06,
+      "loss": 0.1947,
+      "step": 11632
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.973467654139691,
+      "learning_rate": 7.545817375261921e-06,
+      "loss": 0.1931,
+      "step": 11633
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.941498013978441,
+      "learning_rate": 7.544220808697036e-06,
+      "loss": 0.1799,
+      "step": 11634
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7971826901881703,
+      "learning_rate": 7.542624308745294e-06,
+      "loss": 0.1502,
+      "step": 11635
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8618617274901741,
+      "learning_rate": 7.541027875450011e-06,
+      "loss": 0.1794,
+      "step": 11636
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8212372318307198,
+      "learning_rate": 7.539431508854482e-06,
+      "loss": 0.1771,
+      "step": 11637
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8450855924204931,
+      "learning_rate": 7.537835209002015e-06,
+      "loss": 0.1869,
+      "step": 11638
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1731059901473806,
+      "learning_rate": 7.536238975935906e-06,
+      "loss": 0.1832,
+      "step": 11639
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8971279390004603,
+      "learning_rate": 7.534642809699455e-06,
+      "loss": 0.1556,
+      "step": 11640
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.279437219873534,
+      "learning_rate": 7.533046710335959e-06,
+      "loss": 0.2015,
+      "step": 11641
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3048828184889434,
+      "learning_rate": 7.531450677888706e-06,
+      "loss": 0.192,
+      "step": 11642
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0784834754808177,
+      "learning_rate": 7.529854712400996e-06,
+      "loss": 0.1835,
+      "step": 11643
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8167241931927995,
+      "learning_rate": 7.528258813916113e-06,
+      "loss": 0.175,
+      "step": 11644
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9433414492128455,
+      "learning_rate": 7.5266629824773506e-06,
+      "loss": 0.1663,
+      "step": 11645
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9127783252817138,
+      "learning_rate": 7.525067218127994e-06,
+      "loss": 0.1837,
+      "step": 11646
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8962957098263203,
+      "learning_rate": 7.523471520911332e-06,
+      "loss": 0.1634,
+      "step": 11647
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9029306025815423,
+      "learning_rate": 7.521875890870641e-06,
+      "loss": 0.1654,
+      "step": 11648
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.472572365862214,
+      "learning_rate": 7.520280328049209e-06,
+      "loss": 0.1952,
+      "step": 11649
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8529342448124863,
+      "learning_rate": 7.518684832490311e-06,
+      "loss": 0.1789,
+      "step": 11650
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0128886629264706,
+      "learning_rate": 7.51708940423723e-06,
+      "loss": 0.1848,
+      "step": 11651
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8367934406403794,
+      "learning_rate": 7.5154940433332354e-06,
+      "loss": 0.1677,
+      "step": 11652
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7232024685942742,
+      "learning_rate": 7.513898749821607e-06,
+      "loss": 0.189,
+      "step": 11653
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9930819245270107,
+      "learning_rate": 7.5123035237456145e-06,
+      "loss": 0.19,
+      "step": 11654
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.813383634612531,
+      "learning_rate": 7.510708365148534e-06,
+      "loss": 0.1614,
+      "step": 11655
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0880834403732438,
+      "learning_rate": 7.509113274073624e-06,
+      "loss": 0.1766,
+      "step": 11656
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0095368000735225,
+      "learning_rate": 7.507518250564162e-06,
+      "loss": 0.1605,
+      "step": 11657
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8563653106348561,
+      "learning_rate": 7.505923294663407e-06,
+      "loss": 0.1922,
+      "step": 11658
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9991615525590314,
+      "learning_rate": 7.5043284064146195e-06,
+      "loss": 0.1761,
+      "step": 11659
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8596434111528071,
+      "learning_rate": 7.50273358586107e-06,
+      "loss": 0.1658,
+      "step": 11660
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8708077223179779,
+      "learning_rate": 7.501138833046009e-06,
+      "loss": 0.17,
+      "step": 11661
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9135953852571289,
+      "learning_rate": 7.499544148012701e-06,
+      "loss": 0.1691,
+      "step": 11662
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1459883181004817,
+      "learning_rate": 7.4979495308043956e-06,
+      "loss": 0.1632,
+      "step": 11663
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0935777455155868,
+      "learning_rate": 7.496354981464355e-06,
+      "loss": 0.1817,
+      "step": 11664
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9206013159913617,
+      "learning_rate": 7.494760500035824e-06,
+      "loss": 0.1793,
+      "step": 11665
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9309447002221981,
+      "learning_rate": 7.493166086562057e-06,
+      "loss": 0.1669,
+      "step": 11666
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.847841347921362,
+      "learning_rate": 7.4915717410862985e-06,
+      "loss": 0.1515,
+      "step": 11667
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8058277967072249,
+      "learning_rate": 7.489977463651805e-06,
+      "loss": 0.1861,
+      "step": 11668
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0099173797801007,
+      "learning_rate": 7.488383254301809e-06,
+      "loss": 0.1436,
+      "step": 11669
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.146285183247257,
+      "learning_rate": 7.4867891130795625e-06,
+      "loss": 0.1771,
+      "step": 11670
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9109887240430162,
+      "learning_rate": 7.4851950400283e-06,
+      "loss": 0.1819,
+      "step": 11671
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0952192354772308,
+      "learning_rate": 7.483601035191265e-06,
+      "loss": 0.1899,
+      "step": 11672
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7876174516574952,
+      "learning_rate": 7.482007098611694e-06,
+      "loss": 0.1723,
+      "step": 11673
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9015007081376896,
+      "learning_rate": 7.480413230332826e-06,
+      "loss": 0.189,
+      "step": 11674
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1568448543002259,
+      "learning_rate": 7.478819430397888e-06,
+      "loss": 0.1847,
+      "step": 11675
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.333752114037828,
+      "learning_rate": 7.4772256988501145e-06,
+      "loss": 0.1753,
+      "step": 11676
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8048652598854352,
+      "learning_rate": 7.4756320357327406e-06,
+      "loss": 0.1843,
+      "step": 11677
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8734669456812948,
+      "learning_rate": 7.474038441088987e-06,
+      "loss": 0.1565,
+      "step": 11678
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6532642783039464,
+      "learning_rate": 7.472444914962084e-06,
+      "loss": 0.1814,
+      "step": 11679
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0401268716166456,
+      "learning_rate": 7.470851457395254e-06,
+      "loss": 0.182,
+      "step": 11680
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4617086415693403,
+      "learning_rate": 7.469258068431724e-06,
+      "loss": 0.2092,
+      "step": 11681
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8619882782489962,
+      "learning_rate": 7.467664748114709e-06,
+      "loss": 0.1808,
+      "step": 11682
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9972063938232998,
+      "learning_rate": 7.466071496487432e-06,
+      "loss": 0.1763,
+      "step": 11683
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8998667580089821,
+      "learning_rate": 7.4644783135931076e-06,
+      "loss": 0.2041,
+      "step": 11684
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8611236553868576,
+      "learning_rate": 7.462885199474956e-06,
+      "loss": 0.1838,
+      "step": 11685
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5463735071999285,
+      "learning_rate": 7.461292154176183e-06,
+      "loss": 0.164,
+      "step": 11686
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9550680424020984,
+      "learning_rate": 7.459699177740006e-06,
+      "loss": 0.1831,
+      "step": 11687
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0858276043205346,
+      "learning_rate": 7.4581062702096295e-06,
+      "loss": 0.1742,
+      "step": 11688
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1960381092670782,
+      "learning_rate": 7.456513431628266e-06,
+      "loss": 0.2053,
+      "step": 11689
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.127318310682017,
+      "learning_rate": 7.454920662039118e-06,
+      "loss": 0.1765,
+      "step": 11690
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8470517961431754,
+      "learning_rate": 7.4533279614853935e-06,
+      "loss": 0.1777,
+      "step": 11691
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9298438108234138,
+      "learning_rate": 7.451735330010288e-06,
+      "loss": 0.1779,
+      "step": 11692
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8895116893439401,
+      "learning_rate": 7.450142767657009e-06,
+      "loss": 0.1914,
+      "step": 11693
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2369497075642366,
+      "learning_rate": 7.448550274468752e-06,
+      "loss": 0.1621,
+      "step": 11694
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7270373270555748,
+      "learning_rate": 7.4469578504887094e-06,
+      "loss": 0.1751,
+      "step": 11695
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0363877318541614,
+      "learning_rate": 7.445365495760082e-06,
+      "loss": 0.1832,
+      "step": 11696
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.473420292278499,
+      "learning_rate": 7.443773210326057e-06,
+      "loss": 0.1829,
+      "step": 11697
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0751573517871695,
+      "learning_rate": 7.4421809942298305e-06,
+      "loss": 0.1953,
+      "step": 11698
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9729007669559803,
+      "learning_rate": 7.440588847514587e-06,
+      "loss": 0.1871,
+      "step": 11699
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9134498944853251,
+      "learning_rate": 7.43899677022352e-06,
+      "loss": 0.1819,
+      "step": 11700
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9035168835614964,
+      "learning_rate": 7.437404762399805e-06,
+      "loss": 0.1806,
+      "step": 11701
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1679300875833856,
+      "learning_rate": 7.435812824086632e-06,
+      "loss": 0.1791,
+      "step": 11702
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8328750754816019,
+      "learning_rate": 7.434220955327181e-06,
+      "loss": 0.1596,
+      "step": 11703
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.354004614535836,
+      "learning_rate": 7.4326291561646345e-06,
+      "loss": 0.1971,
+      "step": 11704
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2102840579710277,
+      "learning_rate": 7.4310374266421625e-06,
+      "loss": 0.1502,
+      "step": 11705
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0356127604610017,
+      "learning_rate": 7.429445766802949e-06,
+      "loss": 0.1783,
+      "step": 11706
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8936017296139833,
+      "learning_rate": 7.427854176690161e-06,
+      "loss": 0.1605,
+      "step": 11707
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4480314333799575,
+      "learning_rate": 7.426262656346979e-06,
+      "loss": 0.1862,
+      "step": 11708
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7651554481125494,
+      "learning_rate": 7.424671205816562e-06,
+      "loss": 0.1647,
+      "step": 11709
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.125730464461837,
+      "learning_rate": 7.4230798251420865e-06,
+      "loss": 0.2053,
+      "step": 11710
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0174053546119461,
+      "learning_rate": 7.421488514366719e-06,
+      "loss": 0.1772,
+      "step": 11711
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0451093555191666,
+      "learning_rate": 7.419897273533616e-06,
+      "loss": 0.1855,
+      "step": 11712
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.770060246482943,
+      "learning_rate": 7.41830610268595e-06,
+      "loss": 0.1818,
+      "step": 11713
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2007239777086642,
+      "learning_rate": 7.416715001866873e-06,
+      "loss": 0.184,
+      "step": 11714
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1565457242610033,
+      "learning_rate": 7.415123971119549e-06,
+      "loss": 0.1896,
+      "step": 11715
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9928207158486638,
+      "learning_rate": 7.4135330104871315e-06,
+      "loss": 0.1813,
+      "step": 11716
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1910867213681666,
+      "learning_rate": 7.411942120012781e-06,
+      "loss": 0.1823,
+      "step": 11717
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1848318570575225,
+      "learning_rate": 7.4103512997396434e-06,
+      "loss": 0.1878,
+      "step": 11718
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2032769546753332,
+      "learning_rate": 7.408760549710874e-06,
+      "loss": 0.1879,
+      "step": 11719
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7872865465004285,
+      "learning_rate": 7.40716986996962e-06,
+      "loss": 0.1565,
+      "step": 11720
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1778944733024477,
+      "learning_rate": 7.405579260559033e-06,
+      "loss": 0.1751,
+      "step": 11721
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6036289736231404,
+      "learning_rate": 7.4039887215222515e-06,
+      "loss": 0.1724,
+      "step": 11722
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7619375542330458,
+      "learning_rate": 7.402398252902425e-06,
+      "loss": 0.169,
+      "step": 11723
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0087057949594154,
+      "learning_rate": 7.400807854742689e-06,
+      "loss": 0.1545,
+      "step": 11724
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.008949785204754,
+      "learning_rate": 7.399217527086192e-06,
+      "loss": 0.1866,
+      "step": 11725
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3812962714038624,
+      "learning_rate": 7.397627269976062e-06,
+      "loss": 0.1737,
+      "step": 11726
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0063596374669785,
+      "learning_rate": 7.396037083455444e-06,
+      "loss": 0.1723,
+      "step": 11727
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9817553872537692,
+      "learning_rate": 7.394446967567464e-06,
+      "loss": 0.1541,
+      "step": 11728
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9842802734124819,
+      "learning_rate": 7.392856922355256e-06,
+      "loss": 0.1935,
+      "step": 11729
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.932910795695672,
+      "learning_rate": 7.3912669478619555e-06,
+      "loss": 0.1928,
+      "step": 11730
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9222028155289357,
+      "learning_rate": 7.389677044130682e-06,
+      "loss": 0.1671,
+      "step": 11731
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0197399095709354,
+      "learning_rate": 7.3880872112045685e-06,
+      "loss": 0.1622,
+      "step": 11732
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3319641346176132,
+      "learning_rate": 7.386497449126735e-06,
+      "loss": 0.1827,
+      "step": 11733
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.352246925301863,
+      "learning_rate": 7.384907757940309e-06,
+      "loss": 0.1698,
+      "step": 11734
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2899917207740446,
+      "learning_rate": 7.383318137688403e-06,
+      "loss": 0.1954,
+      "step": 11735
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8328991342166945,
+      "learning_rate": 7.381728588414143e-06,
+      "loss": 0.1813,
+      "step": 11736
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0932700968247486,
+      "learning_rate": 7.380139110160642e-06,
+      "loss": 0.1883,
+      "step": 11737
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8564260352071182,
+      "learning_rate": 7.378549702971018e-06,
+      "loss": 0.1905,
+      "step": 11738
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2383818779988518,
+      "learning_rate": 7.3769603668883794e-06,
+      "loss": 0.1784,
+      "step": 11739
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9750897662967867,
+      "learning_rate": 7.375371101955842e-06,
+      "loss": 0.1606,
+      "step": 11740
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0507731167395518,
+      "learning_rate": 7.373781908216507e-06,
+      "loss": 0.174,
+      "step": 11741
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.124368626038594,
+      "learning_rate": 7.372192785713489e-06,
+      "loss": 0.1719,
+      "step": 11742
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7837162544033184,
+      "learning_rate": 7.370603734489887e-06,
+      "loss": 0.159,
+      "step": 11743
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8844333450966607,
+      "learning_rate": 7.3690147545888124e-06,
+      "loss": 0.2085,
+      "step": 11744
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2703046460006664,
+      "learning_rate": 7.36742584605336e-06,
+      "loss": 0.1899,
+      "step": 11745
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0418363545100195,
+      "learning_rate": 7.365837008926626e-06,
+      "loss": 0.1729,
+      "step": 11746
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9182324287203589,
+      "learning_rate": 7.364248243251717e-06,
+      "loss": 0.1685,
+      "step": 11747
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4520427465418861,
+      "learning_rate": 7.362659549071719e-06,
+      "loss": 0.203,
+      "step": 11748
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8872792416800015,
+      "learning_rate": 7.361070926429732e-06,
+      "loss": 0.1743,
+      "step": 11749
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9247274282638314,
+      "learning_rate": 7.359482375368843e-06,
+      "loss": 0.1945,
+      "step": 11750
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0557085759243954,
+      "learning_rate": 7.357893895932148e-06,
+      "loss": 0.1891,
+      "step": 11751
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8863500229256194,
+      "learning_rate": 7.356305488162725e-06,
+      "loss": 0.1874,
+      "step": 11752
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6322796010243121,
+      "learning_rate": 7.3547171521036705e-06,
+      "loss": 0.1572,
+      "step": 11753
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1802599977822752,
+      "learning_rate": 7.353128887798058e-06,
+      "loss": 0.1852,
+      "step": 11754
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9031348770048434,
+      "learning_rate": 7.351540695288977e-06,
+      "loss": 0.1597,
+      "step": 11755
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3039431244166906,
+      "learning_rate": 7.349952574619501e-06,
+      "loss": 0.1914,
+      "step": 11756
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8685734518172754,
+      "learning_rate": 7.3483645258327145e-06,
+      "loss": 0.1813,
+      "step": 11757
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0207187065630585,
+      "learning_rate": 7.346776548971687e-06,
+      "loss": 0.198,
+      "step": 11758
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0524126214529708,
+      "learning_rate": 7.345188644079497e-06,
+      "loss": 0.1833,
+      "step": 11759
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8143507974511598,
+      "learning_rate": 7.3436008111992145e-06,
+      "loss": 0.1725,
+      "step": 11760
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8899011291190664,
+      "learning_rate": 7.342013050373913e-06,
+      "loss": 0.1763,
+      "step": 11761
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9923921534932216,
+      "learning_rate": 7.340425361646653e-06,
+      "loss": 0.1843,
+      "step": 11762
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7734285672321267,
+      "learning_rate": 7.338837745060508e-06,
+      "loss": 0.1713,
+      "step": 11763
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9661992269342103,
+      "learning_rate": 7.337250200658541e-06,
+      "loss": 0.1829,
+      "step": 11764
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9162980784153151,
+      "learning_rate": 7.335662728483808e-06,
+      "loss": 0.1751,
+      "step": 11765
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6337898949008385,
+      "learning_rate": 7.33407532857938e-06,
+      "loss": 0.1695,
+      "step": 11766
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1236715789822076,
+      "learning_rate": 7.332488000988303e-06,
+      "loss": 0.1855,
+      "step": 11767
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.157821899167246,
+      "learning_rate": 7.330900745753643e-06,
+      "loss": 0.1811,
+      "step": 11768
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9014010739784609,
+      "learning_rate": 7.329313562918449e-06,
+      "loss": 0.1748,
+      "step": 11769
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.13217901184811,
+      "learning_rate": 7.327726452525779e-06,
+      "loss": 0.1707,
+      "step": 11770
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9580859428605121,
+      "learning_rate": 7.326139414618676e-06,
+      "loss": 0.1931,
+      "step": 11771
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9840830862799712,
+      "learning_rate": 7.324552449240194e-06,
+      "loss": 0.1843,
+      "step": 11772
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9225628334085767,
+      "learning_rate": 7.322965556433377e-06,
+      "loss": 0.1967,
+      "step": 11773
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8297351049262989,
+      "learning_rate": 7.321378736241274e-06,
+      "loss": 0.1662,
+      "step": 11774
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9391575526978314,
+      "learning_rate": 7.31979198870692e-06,
+      "loss": 0.1803,
+      "step": 11775
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8288981321997082,
+      "learning_rate": 7.318205313873361e-06,
+      "loss": 0.1572,
+      "step": 11776
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0773086929648976,
+      "learning_rate": 7.316618711783634e-06,
+      "loss": 0.2063,
+      "step": 11777
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1221116623307208,
+      "learning_rate": 7.315032182480779e-06,
+      "loss": 0.181,
+      "step": 11778
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0846276680180476,
+      "learning_rate": 7.313445726007824e-06,
+      "loss": 0.1942,
+      "step": 11779
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9901471340187721,
+      "learning_rate": 7.311859342407809e-06,
+      "loss": 0.1945,
+      "step": 11780
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8943804861069643,
+      "learning_rate": 7.310273031723759e-06,
+      "loss": 0.1569,
+      "step": 11781
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8785188472162674,
+      "learning_rate": 7.3086867939987025e-06,
+      "loss": 0.1808,
+      "step": 11782
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2553735501788523,
+      "learning_rate": 7.307100629275674e-06,
+      "loss": 0.1856,
+      "step": 11783
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7430195608217605,
+      "learning_rate": 7.305514537597689e-06,
+      "loss": 0.1745,
+      "step": 11784
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8762282857033288,
+      "learning_rate": 7.303928519007776e-06,
+      "loss": 0.1854,
+      "step": 11785
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.973383706689411,
+      "learning_rate": 7.302342573548952e-06,
+      "loss": 0.2029,
+      "step": 11786
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2048069280922016,
+      "learning_rate": 7.300756701264242e-06,
+      "loss": 0.1754,
+      "step": 11787
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.330264833832681,
+      "learning_rate": 7.299170902196655e-06,
+      "loss": 0.18,
+      "step": 11788
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8528049614191048,
+      "learning_rate": 7.297585176389212e-06,
+      "loss": 0.1892,
+      "step": 11789
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9450155315439861,
+      "learning_rate": 7.295999523884921e-06,
+      "loss": 0.168,
+      "step": 11790
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9239769701781485,
+      "learning_rate": 7.294413944726801e-06,
+      "loss": 0.1762,
+      "step": 11791
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1140158903085968,
+      "learning_rate": 7.292828438957851e-06,
+      "loss": 0.1669,
+      "step": 11792
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7712235884304885,
+      "learning_rate": 7.291243006621084e-06,
+      "loss": 0.1739,
+      "step": 11793
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0059461187022227,
+      "learning_rate": 7.289657647759501e-06,
+      "loss": 0.163,
+      "step": 11794
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9109098709361532,
+      "learning_rate": 7.288072362416112e-06,
+      "loss": 0.1922,
+      "step": 11795
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9489883402080502,
+      "learning_rate": 7.28648715063391e-06,
+      "loss": 0.16,
+      "step": 11796
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0745977552488903,
+      "learning_rate": 7.2849020124559015e-06,
+      "loss": 0.1783,
+      "step": 11797
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9793700759659624,
+      "learning_rate": 7.283316947925075e-06,
+      "loss": 0.2013,
+      "step": 11798
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9171954822057149,
+      "learning_rate": 7.28173195708443e-06,
+      "loss": 0.1914,
+      "step": 11799
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9517807304858662,
+      "learning_rate": 7.280147039976963e-06,
+      "loss": 0.1757,
+      "step": 11800
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9207643617133459,
+      "learning_rate": 7.278562196645656e-06,
+      "loss": 0.1978,
+      "step": 11801
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2926951372931075,
+      "learning_rate": 7.276977427133505e-06,
+      "loss": 0.2026,
+      "step": 11802
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1629358351213168,
+      "learning_rate": 7.275392731483495e-06,
+      "loss": 0.1937,
+      "step": 11803
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9957458561200179,
+      "learning_rate": 7.273808109738614e-06,
+      "loss": 0.1602,
+      "step": 11804
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4993122787846052,
+      "learning_rate": 7.272223561941837e-06,
+      "loss": 0.1736,
+      "step": 11805
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3441130422411014,
+      "learning_rate": 7.270639088136154e-06,
+      "loss": 0.1756,
+      "step": 11806
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9196103667338753,
+      "learning_rate": 7.269054688364535e-06,
+      "loss": 0.1838,
+      "step": 11807
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0711928579048098,
+      "learning_rate": 7.2674703626699685e-06,
+      "loss": 0.1731,
+      "step": 11808
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1727203349150228,
+      "learning_rate": 7.265886111095417e-06,
+      "loss": 0.2018,
+      "step": 11809
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.19263133780406,
+      "learning_rate": 7.264301933683864e-06,
+      "loss": 0.1866,
+      "step": 11810
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9406769046267162,
+      "learning_rate": 7.262717830478272e-06,
+      "loss": 0.1979,
+      "step": 11811
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9433538747251087,
+      "learning_rate": 7.261133801521614e-06,
+      "loss": 0.1738,
+      "step": 11812
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9999985459271858,
+      "learning_rate": 7.259549846856855e-06,
+      "loss": 0.1768,
+      "step": 11813
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.089501907253873,
+      "learning_rate": 7.257965966526966e-06,
+      "loss": 0.1793,
+      "step": 11814
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1513213292828732,
+      "learning_rate": 7.256382160574902e-06,
+      "loss": 0.1967,
+      "step": 11815
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9955636307423217,
+      "learning_rate": 7.254798429043626e-06,
+      "loss": 0.1881,
+      "step": 11816
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9804159920200639,
+      "learning_rate": 7.2532147719761e-06,
+      "loss": 0.1697,
+      "step": 11817
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1111216413748202,
+      "learning_rate": 7.251631189415275e-06,
+      "loss": 0.1902,
+      "step": 11818
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0571395710658018,
+      "learning_rate": 7.250047681404113e-06,
+      "loss": 0.2079,
+      "step": 11819
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1265894734074346,
+      "learning_rate": 7.248464247985558e-06,
+      "loss": 0.192,
+      "step": 11820
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8806117017761279,
+      "learning_rate": 7.246880889202572e-06,
+      "loss": 0.155,
+      "step": 11821
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9507389103704897,
+      "learning_rate": 7.245297605098093e-06,
+      "loss": 0.1478,
+      "step": 11822
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2828201375748278,
+      "learning_rate": 7.243714395715076e-06,
+      "loss": 0.1679,
+      "step": 11823
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.137428365187284,
+      "learning_rate": 7.242131261096457e-06,
+      "loss": 0.1591,
+      "step": 11824
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9450990752849141,
+      "learning_rate": 7.240548201285186e-06,
+      "loss": 0.1715,
+      "step": 11825
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9348710156084757,
+      "learning_rate": 7.238965216324199e-06,
+      "loss": 0.1897,
+      "step": 11826
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2180133790398777,
+      "learning_rate": 7.2373823062564396e-06,
+      "loss": 0.1676,
+      "step": 11827
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4145821638061544,
+      "learning_rate": 7.235799471124838e-06,
+      "loss": 0.1753,
+      "step": 11828
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.479901347156181,
+      "learning_rate": 7.234216710972333e-06,
+      "loss": 0.1736,
+      "step": 11829
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0622603323670472,
+      "learning_rate": 7.232634025841853e-06,
+      "loss": 0.1667,
+      "step": 11830
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8733703542173635,
+      "learning_rate": 7.231051415776338e-06,
+      "loss": 0.1768,
+      "step": 11831
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7926624980613768,
+      "learning_rate": 7.2294688808187045e-06,
+      "loss": 0.1545,
+      "step": 11832
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8767977208712247,
+      "learning_rate": 7.227886421011886e-06,
+      "loss": 0.1706,
+      "step": 11833
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9092722362527742,
+      "learning_rate": 7.226304036398808e-06,
+      "loss": 0.179,
+      "step": 11834
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9596541254080619,
+      "learning_rate": 7.224721727022384e-06,
+      "loss": 0.1998,
+      "step": 11835
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2707631192094582,
+      "learning_rate": 7.223139492925544e-06,
+      "loss": 0.1838,
+      "step": 11836
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4641264206313687,
+      "learning_rate": 7.221557334151199e-06,
+      "loss": 0.1954,
+      "step": 11837
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0628876393856026,
+      "learning_rate": 7.21997525074227e-06,
+      "loss": 0.1831,
+      "step": 11838
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8944603679621254,
+      "learning_rate": 7.218393242741667e-06,
+      "loss": 0.1661,
+      "step": 11839
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8294708528212483,
+      "learning_rate": 7.2168113101923085e-06,
+      "loss": 0.1595,
+      "step": 11840
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9545116326224699,
+      "learning_rate": 7.215229453137097e-06,
+      "loss": 0.1907,
+      "step": 11841
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5609981833288196,
+      "learning_rate": 7.213647671618945e-06,
+      "loss": 0.2064,
+      "step": 11842
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2486096471058774,
+      "learning_rate": 7.212065965680755e-06,
+      "loss": 0.1733,
+      "step": 11843
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9830128638983774,
+      "learning_rate": 7.210484335365438e-06,
+      "loss": 0.1777,
+      "step": 11844
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1222461529512562,
+      "learning_rate": 7.208902780715888e-06,
+      "loss": 0.1849,
+      "step": 11845
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9584478418686001,
+      "learning_rate": 7.207321301775008e-06,
+      "loss": 0.1651,
+      "step": 11846
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0400600781793397,
+      "learning_rate": 7.205739898585693e-06,
+      "loss": 0.2062,
+      "step": 11847
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.101326132606511,
+      "learning_rate": 7.204158571190847e-06,
+      "loss": 0.1983,
+      "step": 11848
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9634348795913377,
+      "learning_rate": 7.202577319633353e-06,
+      "loss": 0.1736,
+      "step": 11849
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8806079665225696,
+      "learning_rate": 7.200996143956111e-06,
+      "loss": 0.1777,
+      "step": 11850
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1346657628354975,
+      "learning_rate": 7.199415044202004e-06,
+      "loss": 0.1747,
+      "step": 11851
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7454248060467574,
+      "learning_rate": 7.1978340204139205e-06,
+      "loss": 0.229,
+      "step": 11852
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8133745687679493,
+      "learning_rate": 7.196253072634751e-06,
+      "loss": 0.1764,
+      "step": 11853
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9192064226925258,
+      "learning_rate": 7.194672200907373e-06,
+      "loss": 0.1787,
+      "step": 11854
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9325417312120747,
+      "learning_rate": 7.193091405274671e-06,
+      "loss": 0.1826,
+      "step": 11855
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8704646661650539,
+      "learning_rate": 7.19151068577952e-06,
+      "loss": 0.1806,
+      "step": 11856
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9427144139646895,
+      "learning_rate": 7.189930042464806e-06,
+      "loss": 0.1696,
+      "step": 11857
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7537092950958617,
+      "learning_rate": 7.188349475373393e-06,
+      "loss": 0.16,
+      "step": 11858
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.35349247055404,
+      "learning_rate": 7.186768984548162e-06,
+      "loss": 0.1975,
+      "step": 11859
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1981338252651514,
+      "learning_rate": 7.185188570031979e-06,
+      "loss": 0.19,
+      "step": 11860
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.001442490909649,
+      "learning_rate": 7.18360823186772e-06,
+      "loss": 0.1824,
+      "step": 11861
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8589437176538642,
+      "learning_rate": 7.182027970098242e-06,
+      "loss": 0.1514,
+      "step": 11862
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3932418492827867,
+      "learning_rate": 7.180447784766418e-06,
+      "loss": 0.1654,
+      "step": 11863
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.680823262999588,
+      "learning_rate": 7.178867675915104e-06,
+      "loss": 0.1665,
+      "step": 11864
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8770714709369166,
+      "learning_rate": 7.177287643587164e-06,
+      "loss": 0.1608,
+      "step": 11865
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4125927927249728,
+      "learning_rate": 7.175707687825455e-06,
+      "loss": 0.1637,
+      "step": 11866
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9644948475142574,
+      "learning_rate": 7.1741278086728395e-06,
+      "loss": 0.1644,
+      "step": 11867
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0163483995642684,
+      "learning_rate": 7.172548006172164e-06,
+      "loss": 0.202,
+      "step": 11868
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0437266979742186,
+      "learning_rate": 7.170968280366281e-06,
+      "loss": 0.1642,
+      "step": 11869
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8267503919071744,
+      "learning_rate": 7.1693886312980475e-06,
+      "loss": 0.1618,
+      "step": 11870
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1865281005301913,
+      "learning_rate": 7.1678090590103035e-06,
+      "loss": 0.1923,
+      "step": 11871
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9358919918321995,
+      "learning_rate": 7.166229563545901e-06,
+      "loss": 0.1808,
+      "step": 11872
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.050465910789463,
+      "learning_rate": 7.164650144947679e-06,
+      "loss": 0.1753,
+      "step": 11873
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0196472656738296,
+      "learning_rate": 7.163070803258486e-06,
+      "loss": 0.2121,
+      "step": 11874
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.553112700593638,
+      "learning_rate": 7.161491538521154e-06,
+      "loss": 0.1763,
+      "step": 11875
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.012673181882685,
+      "learning_rate": 7.159912350778528e-06,
+      "loss": 0.1572,
+      "step": 11876
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.216975617994192,
+      "learning_rate": 7.158333240073436e-06,
+      "loss": 0.1983,
+      "step": 11877
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9765534866537904,
+      "learning_rate": 7.156754206448718e-06,
+      "loss": 0.1688,
+      "step": 11878
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7845748811610675,
+      "learning_rate": 7.1551752499472005e-06,
+      "loss": 0.19,
+      "step": 11879
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.424597105206769,
+      "learning_rate": 7.153596370611719e-06,
+      "loss": 0.1788,
+      "step": 11880
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3585442671939263,
+      "learning_rate": 7.152017568485092e-06,
+      "loss": 0.1835,
+      "step": 11881
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7360441143193666,
+      "learning_rate": 7.150438843610152e-06,
+      "loss": 0.189,
+      "step": 11882
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9452059160457844,
+      "learning_rate": 7.148860196029717e-06,
+      "loss": 0.2015,
+      "step": 11883
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2493024368049896,
+      "learning_rate": 7.147281625786615e-06,
+      "loss": 0.1752,
+      "step": 11884
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.883847817100227,
+      "learning_rate": 7.145703132923657e-06,
+      "loss": 0.1823,
+      "step": 11885
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.906830011163742,
+      "learning_rate": 7.144124717483661e-06,
+      "loss": 0.175,
+      "step": 11886
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9714690367328279,
+      "learning_rate": 7.1425463795094476e-06,
+      "loss": 0.163,
+      "step": 11887
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1441987658528388,
+      "learning_rate": 7.14096811904382e-06,
+      "loss": 0.1728,
+      "step": 11888
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9804796842368632,
+      "learning_rate": 7.139389936129599e-06,
+      "loss": 0.1925,
+      "step": 11889
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2599611555433587,
+      "learning_rate": 7.1378118308095835e-06,
+      "loss": 0.1869,
+      "step": 11890
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0853613521821157,
+      "learning_rate": 7.136233803126584e-06,
+      "loss": 0.1751,
+      "step": 11891
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6086541355655184,
+      "learning_rate": 7.1346558531234046e-06,
+      "loss": 0.1936,
+      "step": 11892
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1234407378780253,
+      "learning_rate": 7.133077980842851e-06,
+      "loss": 0.1855,
+      "step": 11893
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1006285198696308,
+      "learning_rate": 7.1315001863277135e-06,
+      "loss": 0.1831,
+      "step": 11894
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.128971427493001,
+      "learning_rate": 7.129922469620798e-06,
+      "loss": 0.1685,
+      "step": 11895
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.878266359471762,
+      "learning_rate": 7.128344830764895e-06,
+      "loss": 0.1826,
+      "step": 11896
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4104581272003494,
+      "learning_rate": 7.126767269802806e-06,
+      "loss": 0.1746,
+      "step": 11897
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1426165349149782,
+      "learning_rate": 7.125189786777312e-06,
+      "loss": 0.1707,
+      "step": 11898
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3657276142785477,
+      "learning_rate": 7.123612381731211e-06,
+      "loss": 0.1663,
+      "step": 11899
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.014592637511237,
+      "learning_rate": 7.122035054707283e-06,
+      "loss": 0.1808,
+      "step": 11900
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.059580390073793,
+      "learning_rate": 7.1204578057483206e-06,
+      "loss": 0.1851,
+      "step": 11901
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1218977907688872,
+      "learning_rate": 7.1188806348971e-06,
+      "loss": 0.192,
+      "step": 11902
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8440398569371811,
+      "learning_rate": 7.117303542196406e-06,
+      "loss": 0.1728,
+      "step": 11903
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0808151942873958,
+      "learning_rate": 7.1157265276890195e-06,
+      "loss": 0.1944,
+      "step": 11904
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.194100575050271,
+      "learning_rate": 7.114149591417709e-06,
+      "loss": 0.1842,
+      "step": 11905
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9826936837280463,
+      "learning_rate": 7.112572733425257e-06,
+      "loss": 0.1748,
+      "step": 11906
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.234040095182392,
+      "learning_rate": 7.11099595375443e-06,
+      "loss": 0.1875,
+      "step": 11907
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9664874950364423,
+      "learning_rate": 7.1094192524480025e-06,
+      "loss": 0.1562,
+      "step": 11908
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7664885752023356,
+      "learning_rate": 7.107842629548738e-06,
+      "loss": 0.1667,
+      "step": 11909
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8510699508661388,
+      "learning_rate": 7.106266085099412e-06,
+      "loss": 0.1942,
+      "step": 11910
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.783986982922015,
+      "learning_rate": 7.104689619142775e-06,
+      "loss": 0.1705,
+      "step": 11911
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7713147450056765,
+      "learning_rate": 7.103113231721599e-06,
+      "loss": 0.1825,
+      "step": 11912
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8175255022774114,
+      "learning_rate": 7.101536922878638e-06,
+      "loss": 0.1721,
+      "step": 11913
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5835071361004465,
+      "learning_rate": 7.0999606926566554e-06,
+      "loss": 0.1853,
+      "step": 11914
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4009658683843862,
+      "learning_rate": 7.0983845410984e-06,
+      "loss": 0.187,
+      "step": 11915
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0056035099506464,
+      "learning_rate": 7.096808468246629e-06,
+      "loss": 0.1735,
+      "step": 11916
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0911482613971564,
+      "learning_rate": 7.095232474144089e-06,
+      "loss": 0.1802,
+      "step": 11917
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9992152632875951,
+      "learning_rate": 7.0936565588335386e-06,
+      "loss": 0.1792,
+      "step": 11918
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3219752435746845,
+      "learning_rate": 7.092080722357713e-06,
+      "loss": 0.1683,
+      "step": 11919
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1595662144721943,
+      "learning_rate": 7.090504964759366e-06,
+      "loss": 0.1739,
+      "step": 11920
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0274332516187044,
+      "learning_rate": 7.0889292860812344e-06,
+      "loss": 0.1694,
+      "step": 11921
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7438566646687913,
+      "learning_rate": 7.087353686366059e-06,
+      "loss": 0.1743,
+      "step": 11922
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7802668225695544,
+      "learning_rate": 7.085778165656581e-06,
+      "loss": 0.1644,
+      "step": 11923
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.044161670020728,
+      "learning_rate": 7.084202723995533e-06,
+      "loss": 0.1855,
+      "step": 11924
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8374455063926431,
+      "learning_rate": 7.082627361425652e-06,
+      "loss": 0.1736,
+      "step": 11925
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0659359053226045,
+      "learning_rate": 7.081052077989668e-06,
+      "loss": 0.1615,
+      "step": 11926
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0140527059862043,
+      "learning_rate": 7.0794768737303135e-06,
+      "loss": 0.1732,
+      "step": 11927
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9475497891387962,
+      "learning_rate": 7.07790174869031e-06,
+      "loss": 0.1554,
+      "step": 11928
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8568709849395552,
+      "learning_rate": 7.076326702912388e-06,
+      "loss": 0.1957,
+      "step": 11929
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0945804443320943,
+      "learning_rate": 7.0747517364392694e-06,
+      "loss": 0.1911,
+      "step": 11930
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.042478240027438,
+      "learning_rate": 7.073176849313678e-06,
+      "loss": 0.193,
+      "step": 11931
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9780085983785852,
+      "learning_rate": 7.071602041578325e-06,
+      "loss": 0.2077,
+      "step": 11932
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8456134232105421,
+      "learning_rate": 7.0700273132759374e-06,
+      "loss": 0.1859,
+      "step": 11933
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0665183972480694,
+      "learning_rate": 7.068452664449219e-06,
+      "loss": 0.189,
+      "step": 11934
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.7600552381034,
+      "learning_rate": 7.066878095140892e-06,
+      "loss": 0.1883,
+      "step": 11935
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7283221691703692,
+      "learning_rate": 7.065303605393659e-06,
+      "loss": 0.1785,
+      "step": 11936
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.944965829965286,
+      "learning_rate": 7.0637291952502355e-06,
+      "loss": 0.191,
+      "step": 11937
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2512234638902588,
+      "learning_rate": 7.062154864753321e-06,
+      "loss": 0.1475,
+      "step": 11938
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8827605647103475,
+      "learning_rate": 7.0605806139456205e-06,
+      "loss": 0.1713,
+      "step": 11939
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9519568727584321,
+      "learning_rate": 7.05900644286984e-06,
+      "loss": 0.1674,
+      "step": 11940
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0050025496707462,
+      "learning_rate": 7.057432351568671e-06,
+      "loss": 0.1691,
+      "step": 11941
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1691513647351677,
+      "learning_rate": 7.05585834008482e-06,
+      "loss": 0.1958,
+      "step": 11942
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.082537787781271,
+      "learning_rate": 7.054284408460974e-06,
+      "loss": 0.1719,
+      "step": 11943
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0093555069161197,
+      "learning_rate": 7.052710556739835e-06,
+      "loss": 0.1721,
+      "step": 11944
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.367301110642045,
+      "learning_rate": 7.051136784964083e-06,
+      "loss": 0.1933,
+      "step": 11945
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8675198631297651,
+      "learning_rate": 7.049563093176418e-06,
+      "loss": 0.1958,
+      "step": 11946
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8474305343227716,
+      "learning_rate": 7.047989481419516e-06,
+      "loss": 0.1806,
+      "step": 11947
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2477633360015177,
+      "learning_rate": 7.0464159497360675e-06,
+      "loss": 0.2078,
+      "step": 11948
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.37049127199201,
+      "learning_rate": 7.044842498168752e-06,
+      "loss": 0.1967,
+      "step": 11949
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.081084129031886,
+      "learning_rate": 7.043269126760255e-06,
+      "loss": 0.1803,
+      "step": 11950
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0286902605759753,
+      "learning_rate": 7.041695835553245e-06,
+      "loss": 0.1862,
+      "step": 11951
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9401373508312021,
+      "learning_rate": 7.040122624590405e-06,
+      "loss": 0.1669,
+      "step": 11952
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.520605516020862,
+      "learning_rate": 7.038549493914404e-06,
+      "loss": 0.1609,
+      "step": 11953
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1363660276509426,
+      "learning_rate": 7.03697644356792e-06,
+      "loss": 0.1696,
+      "step": 11954
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.269973237982814,
+      "learning_rate": 7.035403473593614e-06,
+      "loss": 0.1637,
+      "step": 11955
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0499017769975487,
+      "learning_rate": 7.0338305840341535e-06,
+      "loss": 0.1809,
+      "step": 11956
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9581650638544686,
+      "learning_rate": 7.032257774932212e-06,
+      "loss": 0.202,
+      "step": 11957
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1165249053239028,
+      "learning_rate": 7.030685046330441e-06,
+      "loss": 0.1592,
+      "step": 11958
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.303437766903637,
+      "learning_rate": 7.02911239827151e-06,
+      "loss": 0.1623,
+      "step": 11959
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8286811962313174,
+      "learning_rate": 7.027539830798069e-06,
+      "loss": 0.1495,
+      "step": 11960
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4931590119544556,
+      "learning_rate": 7.02596734395278e-06,
+      "loss": 0.1829,
+      "step": 11961
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9899743456092847,
+      "learning_rate": 7.024394937778293e-06,
+      "loss": 0.178,
+      "step": 11962
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1519840371806078,
+      "learning_rate": 7.022822612317265e-06,
+      "loss": 0.2048,
+      "step": 11963
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9024290437544673,
+      "learning_rate": 7.021250367612338e-06,
+      "loss": 0.1552,
+      "step": 11964
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9192970344585016,
+      "learning_rate": 7.019678203706164e-06,
+      "loss": 0.1918,
+      "step": 11965
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9435428055972597,
+      "learning_rate": 7.018106120641386e-06,
+      "loss": 0.1798,
+      "step": 11966
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9971047001488691,
+      "learning_rate": 7.016534118460652e-06,
+      "loss": 0.1695,
+      "step": 11967
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1748885021929922,
+      "learning_rate": 7.014962197206594e-06,
+      "loss": 0.175,
+      "step": 11968
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9969796770369592,
+      "learning_rate": 7.013390356921858e-06,
+      "loss": 0.1897,
+      "step": 11969
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9595108683100719,
+      "learning_rate": 7.011818597649074e-06,
+      "loss": 0.1694,
+      "step": 11970
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9267782372098969,
+      "learning_rate": 7.010246919430884e-06,
+      "loss": 0.1956,
+      "step": 11971
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8777993144052069,
+      "learning_rate": 7.008675322309911e-06,
+      "loss": 0.1847,
+      "step": 11972
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7337740983158976,
+      "learning_rate": 7.0071038063287935e-06,
+      "loss": 0.1991,
+      "step": 11973
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0782367779242605,
+      "learning_rate": 7.005532371530152e-06,
+      "loss": 0.201,
+      "step": 11974
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.501456289799915,
+      "learning_rate": 7.003961017956611e-06,
+      "loss": 0.1869,
+      "step": 11975
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1395341072149203,
+      "learning_rate": 7.002389745650801e-06,
+      "loss": 0.2017,
+      "step": 11976
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1050083080002921,
+      "learning_rate": 7.000818554655335e-06,
+      "loss": 0.1697,
+      "step": 11977
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7721561404794662,
+      "learning_rate": 6.9992474450128375e-06,
+      "loss": 0.1542,
+      "step": 11978
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.043825789826784,
+      "learning_rate": 6.997676416765919e-06,
+      "loss": 0.1771,
+      "step": 11979
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.473082441806016,
+      "learning_rate": 6.996105469957204e-06,
+      "loss": 0.1672,
+      "step": 11980
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1440472364717835,
+      "learning_rate": 6.994534604629291e-06,
+      "loss": 0.1782,
+      "step": 11981
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9716464255364722,
+      "learning_rate": 6.9929638208247994e-06,
+      "loss": 0.189,
+      "step": 11982
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.069848684646463,
+      "learning_rate": 6.991393118586333e-06,
+      "loss": 0.199,
+      "step": 11983
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.909313878941909,
+      "learning_rate": 6.989822497956501e-06,
+      "loss": 0.182,
+      "step": 11984
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.029764795486539,
+      "learning_rate": 6.9882519589779005e-06,
+      "loss": 0.194,
+      "step": 11985
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9275625368461713,
+      "learning_rate": 6.986681501693139e-06,
+      "loss": 0.1921,
+      "step": 11986
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0466489063396667,
+      "learning_rate": 6.985111126144808e-06,
+      "loss": 0.1705,
+      "step": 11987
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7498953008113175,
+      "learning_rate": 6.983540832375511e-06,
+      "loss": 0.1567,
+      "step": 11988
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3164636095920277,
+      "learning_rate": 6.981970620427836e-06,
+      "loss": 0.1582,
+      "step": 11989
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9719462276135705,
+      "learning_rate": 6.980400490344383e-06,
+      "loss": 0.2122,
+      "step": 11990
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8601857009168751,
+      "learning_rate": 6.9788304421677355e-06,
+      "loss": 0.1746,
+      "step": 11991
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.861866966346215,
+      "learning_rate": 6.97726047594048e-06,
+      "loss": 0.1765,
+      "step": 11992
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9007785415154862,
+      "learning_rate": 6.975690591705211e-06,
+      "loss": 0.162,
+      "step": 11993
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9620423411317283,
+      "learning_rate": 6.974120789504499e-06,
+      "loss": 0.1736,
+      "step": 11994
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.686444458854959,
+      "learning_rate": 6.972551069380935e-06,
+      "loss": 0.1866,
+      "step": 11995
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3545266700952874,
+      "learning_rate": 6.9709814313770905e-06,
+      "loss": 0.1644,
+      "step": 11996
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1660286325084908,
+      "learning_rate": 6.969411875535552e-06,
+      "loss": 0.1799,
+      "step": 11997
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0370267507532889,
+      "learning_rate": 6.967842401898882e-06,
+      "loss": 0.2005,
+      "step": 11998
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7969908814107881,
+      "learning_rate": 6.966273010509663e-06,
+      "loss": 0.1664,
+      "step": 11999
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3401157501995602,
+      "learning_rate": 6.964703701410455e-06,
+      "loss": 0.1846,
+      "step": 12000
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0017537274961412,
+      "learning_rate": 6.963134474643834e-06,
+      "loss": 0.2044,
+      "step": 12001
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9056146251290355,
+      "learning_rate": 6.961565330252358e-06,
+      "loss": 0.1871,
+      "step": 12002
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2843447159022305,
+      "learning_rate": 6.959996268278599e-06,
+      "loss": 0.1831,
+      "step": 12003
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9759637968679855,
+      "learning_rate": 6.958427288765108e-06,
+      "loss": 0.1687,
+      "step": 12004
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1690450666793533,
+      "learning_rate": 6.956858391754453e-06,
+      "loss": 0.2056,
+      "step": 12005
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.412385050354641,
+      "learning_rate": 6.955289577289181e-06,
+      "loss": 0.1845,
+      "step": 12006
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9436818805180692,
+      "learning_rate": 6.953720845411858e-06,
+      "loss": 0.1676,
+      "step": 12007
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3523117395487558,
+      "learning_rate": 6.952152196165025e-06,
+      "loss": 0.1984,
+      "step": 12008
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8942514384754394,
+      "learning_rate": 6.950583629591235e-06,
+      "loss": 0.1808,
+      "step": 12009
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9647460971315064,
+      "learning_rate": 6.94901514573304e-06,
+      "loss": 0.1716,
+      "step": 12010
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7947737276649559,
+      "learning_rate": 6.9474467446329775e-06,
+      "loss": 0.1601,
+      "step": 12011
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2135854992453965,
+      "learning_rate": 6.9458784263335965e-06,
+      "loss": 0.1749,
+      "step": 12012
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8258705553834466,
+      "learning_rate": 6.944310190877433e-06,
+      "loss": 0.1806,
+      "step": 12013
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0624409171900406,
+      "learning_rate": 6.942742038307033e-06,
+      "loss": 0.1654,
+      "step": 12014
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8998384943681688,
+      "learning_rate": 6.941173968664923e-06,
+      "loss": 0.1806,
+      "step": 12015
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0775549450326525,
+      "learning_rate": 6.939605981993647e-06,
+      "loss": 0.1928,
+      "step": 12016
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0758381169200557,
+      "learning_rate": 6.938038078335727e-06,
+      "loss": 0.1669,
+      "step": 12017
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0478686096447207,
+      "learning_rate": 6.936470257733699e-06,
+      "loss": 0.1935,
+      "step": 12018
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0007796945466119,
+      "learning_rate": 6.9349025202300865e-06,
+      "loss": 0.163,
+      "step": 12019
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.062954767058354,
+      "learning_rate": 6.93333486586742e-06,
+      "loss": 0.174,
+      "step": 12020
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1895420775002619,
+      "learning_rate": 6.931767294688214e-06,
+      "loss": 0.1786,
+      "step": 12021
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.150659466077958,
+      "learning_rate": 6.930199806734996e-06,
+      "loss": 0.1693,
+      "step": 12022
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9054373439655421,
+      "learning_rate": 6.92863240205028e-06,
+      "loss": 0.1825,
+      "step": 12023
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0779083796299886,
+      "learning_rate": 6.927065080676587e-06,
+      "loss": 0.1796,
+      "step": 12024
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6111483127425923,
+      "learning_rate": 6.9254978426564256e-06,
+      "loss": 0.1789,
+      "step": 12025
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7561342831981865,
+      "learning_rate": 6.923930688032308e-06,
+      "loss": 0.1653,
+      "step": 12026
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.252568257892285,
+      "learning_rate": 6.922363616846746e-06,
+      "loss": 0.1928,
+      "step": 12027
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3861545130349302,
+      "learning_rate": 6.920796629142242e-06,
+      "loss": 0.1789,
+      "step": 12028
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.126739769588076,
+      "learning_rate": 6.9192297249613074e-06,
+      "loss": 0.1963,
+      "step": 12029
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.376665325309207,
+      "learning_rate": 6.9176629043464364e-06,
+      "loss": 0.1705,
+      "step": 12030
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0270356795638427,
+      "learning_rate": 6.916096167340134e-06,
+      "loss": 0.1797,
+      "step": 12031
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.872751930898111,
+      "learning_rate": 6.9145295139848954e-06,
+      "loss": 0.1788,
+      "step": 12032
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3292159171289397,
+      "learning_rate": 6.9129629443232235e-06,
+      "loss": 0.184,
+      "step": 12033
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9125351804266222,
+      "learning_rate": 6.911396458397602e-06,
+      "loss": 0.1744,
+      "step": 12034
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8528246667288966,
+      "learning_rate": 6.909830056250527e-06,
+      "loss": 0.1622,
+      "step": 12035
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.270843585328278,
+      "learning_rate": 6.9082637379244844e-06,
+      "loss": 0.2054,
+      "step": 12036
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9391457519420248,
+      "learning_rate": 6.906697503461968e-06,
+      "loss": 0.1799,
+      "step": 12037
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7632331087196969,
+      "learning_rate": 6.90513135290545e-06,
+      "loss": 0.1588,
+      "step": 12038
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0727942068343892,
+      "learning_rate": 6.903565286297422e-06,
+      "loss": 0.1679,
+      "step": 12039
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0133257925068448,
+      "learning_rate": 6.901999303680359e-06,
+      "loss": 0.177,
+      "step": 12040
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8623895125847567,
+      "learning_rate": 6.900433405096744e-06,
+      "loss": 0.1623,
+      "step": 12041
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9602670825421511,
+      "learning_rate": 6.898867590589047e-06,
+      "loss": 0.1617,
+      "step": 12042
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9567573291126384,
+      "learning_rate": 6.897301860199738e-06,
+      "loss": 0.1533,
+      "step": 12043
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9042572859837555,
+      "learning_rate": 6.895736213971293e-06,
+      "loss": 0.1654,
+      "step": 12044
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.330908865010921,
+      "learning_rate": 6.8941706519461785e-06,
+      "loss": 0.1989,
+      "step": 12045
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0736449432758612,
+      "learning_rate": 6.892605174166862e-06,
+      "loss": 0.1762,
+      "step": 12046
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8609311260815152,
+      "learning_rate": 6.891039780675803e-06,
+      "loss": 0.1609,
+      "step": 12047
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.502432617527464,
+      "learning_rate": 6.889474471515467e-06,
+      "loss": 0.1671,
+      "step": 12048
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5916558758401025,
+      "learning_rate": 6.887909246728311e-06,
+      "loss": 0.1678,
+      "step": 12049
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7377823963807983,
+      "learning_rate": 6.886344106356795e-06,
+      "loss": 0.1733,
+      "step": 12050
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8830484258834155,
+      "learning_rate": 6.8847790504433664e-06,
+      "loss": 0.162,
+      "step": 12051
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1390348137066064,
+      "learning_rate": 6.883214079030485e-06,
+      "loss": 0.1737,
+      "step": 12052
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8768182692622531,
+      "learning_rate": 6.881649192160596e-06,
+      "loss": 0.1553,
+      "step": 12053
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3875687021614835,
+      "learning_rate": 6.880084389876153e-06,
+      "loss": 0.1863,
+      "step": 12054
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9544048864076122,
+      "learning_rate": 6.878519672219592e-06,
+      "loss": 0.1617,
+      "step": 12055
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0780909741086129,
+      "learning_rate": 6.8769550392333665e-06,
+      "loss": 0.1881,
+      "step": 12056
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.954038343043023,
+      "learning_rate": 6.875390490959907e-06,
+      "loss": 0.1743,
+      "step": 12057
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1336834996926124,
+      "learning_rate": 6.873826027441659e-06,
+      "loss": 0.1784,
+      "step": 12058
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9306362669905568,
+      "learning_rate": 6.872261648721055e-06,
+      "loss": 0.1879,
+      "step": 12059
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.197695678667343,
+      "learning_rate": 6.870697354840534e-06,
+      "loss": 0.1962,
+      "step": 12060
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8779976121881782,
+      "learning_rate": 6.869133145842521e-06,
+      "loss": 0.1757,
+      "step": 12061
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0792172420476411,
+      "learning_rate": 6.8675690217694476e-06,
+      "loss": 0.1962,
+      "step": 12062
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0307635970176123,
+      "learning_rate": 6.866004982663746e-06,
+      "loss": 0.1805,
+      "step": 12063
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8749574919822711,
+      "learning_rate": 6.864441028567831e-06,
+      "loss": 0.1726,
+      "step": 12064
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.898665738913217,
+      "learning_rate": 6.862877159524133e-06,
+      "loss": 0.1558,
+      "step": 12065
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4824044475075975,
+      "learning_rate": 6.861313375575067e-06,
+      "loss": 0.1659,
+      "step": 12066
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9629904483904526,
+      "learning_rate": 6.859749676763056e-06,
+      "loss": 0.1581,
+      "step": 12067
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8272828329013648,
+      "learning_rate": 6.85818606313051e-06,
+      "loss": 0.1689,
+      "step": 12068
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6118105822053284,
+      "learning_rate": 6.856622534719848e-06,
+      "loss": 0.1891,
+      "step": 12069
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7255331847586153,
+      "learning_rate": 6.855059091573472e-06,
+      "loss": 0.2084,
+      "step": 12070
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1749647430370458,
+      "learning_rate": 6.853495733733799e-06,
+      "loss": 0.2064,
+      "step": 12071
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8938854197590528,
+      "learning_rate": 6.851932461243229e-06,
+      "loss": 0.1925,
+      "step": 12072
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0110531702751793,
+      "learning_rate": 6.850369274144174e-06,
+      "loss": 0.1726,
+      "step": 12073
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6937841164262705,
+      "learning_rate": 6.848806172479025e-06,
+      "loss": 0.1823,
+      "step": 12074
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9360324682026385,
+      "learning_rate": 6.847243156290191e-06,
+      "loss": 0.173,
+      "step": 12075
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5232523179973194,
+      "learning_rate": 6.8456802256200596e-06,
+      "loss": 0.1734,
+      "step": 12076
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0482809026447173,
+      "learning_rate": 6.844117380511036e-06,
+      "loss": 0.1811,
+      "step": 12077
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8719802085341933,
+      "learning_rate": 6.842554621005504e-06,
+      "loss": 0.1792,
+      "step": 12078
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.224889727426434,
+      "learning_rate": 6.840991947145854e-06,
+      "loss": 0.1858,
+      "step": 12079
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0220060453992144,
+      "learning_rate": 6.839429358974481e-06,
+      "loss": 0.1755,
+      "step": 12080
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9639989927823014,
+      "learning_rate": 6.837866856533761e-06,
+      "loss": 0.1683,
+      "step": 12081
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2119181742534815,
+      "learning_rate": 6.836304439866084e-06,
+      "loss": 0.1741,
+      "step": 12082
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.02227900587153,
+      "learning_rate": 6.834742109013823e-06,
+      "loss": 0.1968,
+      "step": 12083
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3070559048308237,
+      "learning_rate": 6.833179864019366e-06,
+      "loss": 0.1761,
+      "step": 12084
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1660429668454133,
+      "learning_rate": 6.83161770492508e-06,
+      "loss": 0.1861,
+      "step": 12085
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8861079705636326,
+      "learning_rate": 6.830055631773347e-06,
+      "loss": 0.1627,
+      "step": 12086
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.838978038894428,
+      "learning_rate": 6.82849364460653e-06,
+      "loss": 0.1684,
+      "step": 12087
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8761185023942043,
+      "learning_rate": 6.826931743467004e-06,
+      "loss": 0.1795,
+      "step": 12088
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8362184895394403,
+      "learning_rate": 6.825369928397132e-06,
+      "loss": 0.1841,
+      "step": 12089
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3908870975417034,
+      "learning_rate": 6.8238081994392836e-06,
+      "loss": 0.1791,
+      "step": 12090
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8640804279105561,
+      "learning_rate": 6.822246556635814e-06,
+      "loss": 0.1689,
+      "step": 12091
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0446730178037433,
+      "learning_rate": 6.820685000029087e-06,
+      "loss": 0.1835,
+      "step": 12092
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0395863481397598,
+      "learning_rate": 6.819123529661458e-06,
+      "loss": 0.1669,
+      "step": 12093
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9949242524187777,
+      "learning_rate": 6.817562145575285e-06,
+      "loss": 0.1808,
+      "step": 12094
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2442467941307531,
+      "learning_rate": 6.81600084781292e-06,
+      "loss": 0.1768,
+      "step": 12095
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9768991335555441,
+      "learning_rate": 6.814439636416708e-06,
+      "loss": 0.1734,
+      "step": 12096
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1049467939463726,
+      "learning_rate": 6.812878511429002e-06,
+      "loss": 0.2294,
+      "step": 12097
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8864407710560207,
+      "learning_rate": 6.811317472892145e-06,
+      "loss": 0.1815,
+      "step": 12098
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.648388743148969,
+      "learning_rate": 6.809756520848486e-06,
+      "loss": 0.1769,
+      "step": 12099
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7739231396867304,
+      "learning_rate": 6.808195655340357e-06,
+      "loss": 0.1562,
+      "step": 12100
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9479706936152373,
+      "learning_rate": 6.806634876410103e-06,
+      "loss": 0.1788,
+      "step": 12101
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5049638635512383,
+      "learning_rate": 6.805074184100056e-06,
+      "loss": 0.1851,
+      "step": 12102
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1612459449559867,
+      "learning_rate": 6.803513578452557e-06,
+      "loss": 0.1992,
+      "step": 12103
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8635805305573898,
+      "learning_rate": 6.8019530595099294e-06,
+      "loss": 0.1798,
+      "step": 12104
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9803848651176261,
+      "learning_rate": 6.800392627314506e-06,
+      "loss": 0.1819,
+      "step": 12105
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.934349171865883,
+      "learning_rate": 6.798832281908612e-06,
+      "loss": 0.1826,
+      "step": 12106
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.284915985657181,
+      "learning_rate": 6.797272023334578e-06,
+      "loss": 0.17,
+      "step": 12107
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.967496996176986,
+      "learning_rate": 6.7957118516347156e-06,
+      "loss": 0.1915,
+      "step": 12108
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.028295447950258,
+      "learning_rate": 6.794151766851356e-06,
+      "loss": 0.1713,
+      "step": 12109
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9704153265494159,
+      "learning_rate": 6.792591769026804e-06,
+      "loss": 0.1794,
+      "step": 12110
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5479240050610468,
+      "learning_rate": 6.791031858203385e-06,
+      "loss": 0.17,
+      "step": 12111
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0705127091317566,
+      "learning_rate": 6.789472034423412e-06,
+      "loss": 0.1824,
+      "step": 12112
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0649028556132054,
+      "learning_rate": 6.787912297729184e-06,
+      "loss": 0.1883,
+      "step": 12113
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9102264829131282,
+      "learning_rate": 6.786352648163022e-06,
+      "loss": 0.1838,
+      "step": 12114
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0511014344801506,
+      "learning_rate": 6.7847930857672205e-06,
+      "loss": 0.1784,
+      "step": 12115
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0186815963205775,
+      "learning_rate": 6.783233610584095e-06,
+      "loss": 0.1676,
+      "step": 12116
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8880854123514019,
+      "learning_rate": 6.781674222655934e-06,
+      "loss": 0.1794,
+      "step": 12117
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9295879693635356,
+      "learning_rate": 6.780114922025043e-06,
+      "loss": 0.1856,
+      "step": 12118
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9006839537874907,
+      "learning_rate": 6.778555708733715e-06,
+      "loss": 0.1743,
+      "step": 12119
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9975869625126415,
+      "learning_rate": 6.7769965828242505e-06,
+      "loss": 0.1928,
+      "step": 12120
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1032710418317961,
+      "learning_rate": 6.7754375443389294e-06,
+      "loss": 0.1958,
+      "step": 12121
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0261541172084292,
+      "learning_rate": 6.773878593320052e-06,
+      "loss": 0.1651,
+      "step": 12122
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7787786381676972,
+      "learning_rate": 6.772319729809895e-06,
+      "loss": 0.1636,
+      "step": 12123
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6696086999669617,
+      "learning_rate": 6.770760953850754e-06,
+      "loss": 0.1811,
+      "step": 12124
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4360424505617724,
+      "learning_rate": 6.769202265484899e-06,
+      "loss": 0.1558,
+      "step": 12125
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1170506912862783,
+      "learning_rate": 6.767643664754619e-06,
+      "loss": 0.1812,
+      "step": 12126
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2946728030781767,
+      "learning_rate": 6.766085151702184e-06,
+      "loss": 0.1774,
+      "step": 12127
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9326024358102776,
+      "learning_rate": 6.764526726369873e-06,
+      "loss": 0.1827,
+      "step": 12128
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.076727177272974,
+      "learning_rate": 6.762968388799958e-06,
+      "loss": 0.1913,
+      "step": 12129
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9799196474439371,
+      "learning_rate": 6.7614101390347095e-06,
+      "loss": 0.1742,
+      "step": 12130
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9460898423948457,
+      "learning_rate": 6.759851977116392e-06,
+      "loss": 0.1788,
+      "step": 12131
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9967772399446618,
+      "learning_rate": 6.758293903087272e-06,
+      "loss": 0.1802,
+      "step": 12132
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9750470589847444,
+      "learning_rate": 6.756735916989616e-06,
+      "loss": 0.169,
+      "step": 12133
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9776038603721809,
+      "learning_rate": 6.755178018865678e-06,
+      "loss": 0.1588,
+      "step": 12134
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8506390814787281,
+      "learning_rate": 6.753620208757721e-06,
+      "loss": 0.1761,
+      "step": 12135
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8690918817459329,
+      "learning_rate": 6.7520624867079965e-06,
+      "loss": 0.1851,
+      "step": 12136
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9535446340652028,
+      "learning_rate": 6.7505048527587656e-06,
+      "loss": 0.1634,
+      "step": 12137
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3906783266482012,
+      "learning_rate": 6.748947306952269e-06,
+      "loss": 0.179,
+      "step": 12138
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0931209334496226,
+      "learning_rate": 6.747389849330765e-06,
+      "loss": 0.1685,
+      "step": 12139
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2113979580898313,
+      "learning_rate": 6.745832479936492e-06,
+      "loss": 0.186,
+      "step": 12140
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0226123145274468,
+      "learning_rate": 6.744275198811698e-06,
+      "loss": 0.2114,
+      "step": 12141
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8793336866514591,
+      "learning_rate": 6.742718005998621e-06,
+      "loss": 0.1767,
+      "step": 12142
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0782116318894777,
+      "learning_rate": 6.741160901539506e-06,
+      "loss": 0.1711,
+      "step": 12143
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1707774557584145,
+      "learning_rate": 6.7396038854765825e-06,
+      "loss": 0.1655,
+      "step": 12144
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.104390856186119,
+      "learning_rate": 6.738046957852089e-06,
+      "loss": 0.1661,
+      "step": 12145
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9243443440720327,
+      "learning_rate": 6.736490118708253e-06,
+      "loss": 0.1777,
+      "step": 12146
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7956070831990771,
+      "learning_rate": 6.7349333680873134e-06,
+      "loss": 0.1696,
+      "step": 12147
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0515919332880705,
+      "learning_rate": 6.733376706031486e-06,
+      "loss": 0.1781,
+      "step": 12148
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.857621003636688,
+      "learning_rate": 6.731820132582999e-06,
+      "loss": 0.1746,
+      "step": 12149
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1617687067828353,
+      "learning_rate": 6.730263647784079e-06,
+      "loss": 0.1528,
+      "step": 12150
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0959333069840698,
+      "learning_rate": 6.728707251676939e-06,
+      "loss": 0.1886,
+      "step": 12151
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1623054341465726,
+      "learning_rate": 6.727150944303804e-06,
+      "loss": 0.1776,
+      "step": 12152
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.200230087745242,
+      "learning_rate": 6.7255947257068785e-06,
+      "loss": 0.1949,
+      "step": 12153
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8042497751515356,
+      "learning_rate": 6.724038595928385e-06,
+      "loss": 0.1624,
+      "step": 12154
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3636760843144802,
+      "learning_rate": 6.722482555010528e-06,
+      "loss": 0.1907,
+      "step": 12155
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8964440433985537,
+      "learning_rate": 6.720926602995519e-06,
+      "loss": 0.177,
+      "step": 12156
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.782888558585591,
+      "learning_rate": 6.719370739925557e-06,
+      "loss": 0.1531,
+      "step": 12157
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4599766269420338,
+      "learning_rate": 6.717814965842852e-06,
+      "loss": 0.1639,
+      "step": 12158
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.032260911561945,
+      "learning_rate": 6.716259280789599e-06,
+      "loss": 0.2493,
+      "step": 12159
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4459519392693336,
+      "learning_rate": 6.714703684808004e-06,
+      "loss": 0.1685,
+      "step": 12160
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8809204338200619,
+      "learning_rate": 6.7131481779402505e-06,
+      "loss": 0.178,
+      "step": 12161
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0338654307609594,
+      "learning_rate": 6.7115927602285424e-06,
+      "loss": 0.1912,
+      "step": 12162
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0140484819462559,
+      "learning_rate": 6.710037431715063e-06,
+      "loss": 0.1967,
+      "step": 12163
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0985829362573525,
+      "learning_rate": 6.70848219244201e-06,
+      "loss": 0.1755,
+      "step": 12164
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7797192565191957,
+      "learning_rate": 6.706927042451561e-06,
+      "loss": 0.1654,
+      "step": 12165
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0983960363940297,
+      "learning_rate": 6.7053719817859e-06,
+      "loss": 0.1785,
+      "step": 12166
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.75585709682769,
+      "learning_rate": 6.7038170104872106e-06,
+      "loss": 0.1605,
+      "step": 12167
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2958472121653062,
+      "learning_rate": 6.70226212859767e-06,
+      "loss": 0.217,
+      "step": 12168
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8712282797124832,
+      "learning_rate": 6.700707336159458e-06,
+      "loss": 0.1797,
+      "step": 12169
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9107466535849766,
+      "learning_rate": 6.699152633214743e-06,
+      "loss": 0.18,
+      "step": 12170
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8591545415978744,
+      "learning_rate": 6.697598019805701e-06,
+      "loss": 0.175,
+      "step": 12171
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1046279818870088,
+      "learning_rate": 6.696043495974498e-06,
+      "loss": 0.1778,
+      "step": 12172
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0874204812001573,
+      "learning_rate": 6.6944890617633055e-06,
+      "loss": 0.1929,
+      "step": 12173
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1738304949725218,
+      "learning_rate": 6.6929347172142785e-06,
+      "loss": 0.208,
+      "step": 12174
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1878166769378955,
+      "learning_rate": 6.691380462369588e-06,
+      "loss": 0.1731,
+      "step": 12175
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9056200496531202,
+      "learning_rate": 6.689826297271384e-06,
+      "loss": 0.1787,
+      "step": 12176
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1374102316252948,
+      "learning_rate": 6.6882722219618355e-06,
+      "loss": 0.1701,
+      "step": 12177
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.695640372186796,
+      "learning_rate": 6.686718236483086e-06,
+      "loss": 0.1782,
+      "step": 12178
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3815010096582772,
+      "learning_rate": 6.685164340877295e-06,
+      "loss": 0.1695,
+      "step": 12179
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.083543432973144,
+      "learning_rate": 6.683610535186604e-06,
+      "loss": 0.1866,
+      "step": 12180
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2725443888858214,
+      "learning_rate": 6.682056819453168e-06,
+      "loss": 0.1651,
+      "step": 12181
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8442178166320784,
+      "learning_rate": 6.680503193719129e-06,
+      "loss": 0.1743,
+      "step": 12182
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0029030646090114,
+      "learning_rate": 6.678949658026625e-06,
+      "loss": 0.1923,
+      "step": 12183
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3476567787546982,
+      "learning_rate": 6.677396212417801e-06,
+      "loss": 0.1663,
+      "step": 12184
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0698221213216408,
+      "learning_rate": 6.675842856934789e-06,
+      "loss": 0.1724,
+      "step": 12185
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0245881277985018,
+      "learning_rate": 6.674289591619732e-06,
+      "loss": 0.1664,
+      "step": 12186
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.81864411656269,
+      "learning_rate": 6.672736416514754e-06,
+      "loss": 0.1814,
+      "step": 12187
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1444151316489966,
+      "learning_rate": 6.671183331661991e-06,
+      "loss": 0.1682,
+      "step": 12188
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0227672142427702,
+      "learning_rate": 6.669630337103565e-06,
+      "loss": 0.194,
+      "step": 12189
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1132674106865061,
+      "learning_rate": 6.66807743288161e-06,
+      "loss": 0.1802,
+      "step": 12190
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.422171075788219,
+      "learning_rate": 6.666524619038237e-06,
+      "loss": 0.1802,
+      "step": 12191
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1871758296843817,
+      "learning_rate": 6.664971895615578e-06,
+      "loss": 0.1834,
+      "step": 12192
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.460493666235359,
+      "learning_rate": 6.663419262655739e-06,
+      "loss": 0.1733,
+      "step": 12193
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7359353917510435,
+      "learning_rate": 6.6618667202008435e-06,
+      "loss": 0.1693,
+      "step": 12194
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.112629516499026,
+      "learning_rate": 6.660314268293e-06,
+      "loss": 0.1551,
+      "step": 12195
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0067294238530116,
+      "learning_rate": 6.6587619069743236e-06,
+      "loss": 0.1956,
+      "step": 12196
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1353157521263597,
+      "learning_rate": 6.6572096362869165e-06,
+      "loss": 0.1654,
+      "step": 12197
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0968623503957977,
+      "learning_rate": 6.655657456272888e-06,
+      "loss": 0.2003,
+      "step": 12198
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9659530346298896,
+      "learning_rate": 6.6541053669743375e-06,
+      "loss": 0.1741,
+      "step": 12199
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1324711606541336,
+      "learning_rate": 6.6525533684333724e-06,
+      "loss": 0.1946,
+      "step": 12200
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.06053526847367,
+      "learning_rate": 6.6510014606920845e-06,
+      "loss": 0.1825,
+      "step": 12201
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.936716622629329,
+      "learning_rate": 6.64944964379257e-06,
+      "loss": 0.1448,
+      "step": 12202
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8731618283716971,
+      "learning_rate": 6.647897917776925e-06,
+      "loss": 0.1658,
+      "step": 12203
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9600048699949525,
+      "learning_rate": 6.646346282687235e-06,
+      "loss": 0.179,
+      "step": 12204
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.387098874350567,
+      "learning_rate": 6.644794738565597e-06,
+      "loss": 0.1876,
+      "step": 12205
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8351157077754774,
+      "learning_rate": 6.643243285454086e-06,
+      "loss": 0.1786,
+      "step": 12206
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8903763554787368,
+      "learning_rate": 6.641691923394792e-06,
+      "loss": 0.1825,
+      "step": 12207
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9163699840386397,
+      "learning_rate": 6.640140652429793e-06,
+      "loss": 0.1649,
+      "step": 12208
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1858028194793624,
+      "learning_rate": 6.6385894726011725e-06,
+      "loss": 0.166,
+      "step": 12209
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.400037704545918,
+      "learning_rate": 6.637038383950998e-06,
+      "loss": 0.1737,
+      "step": 12210
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9340838632121498,
+      "learning_rate": 6.63548738652135e-06,
+      "loss": 0.1666,
+      "step": 12211
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4298851897685383,
+      "learning_rate": 6.633936480354294e-06,
+      "loss": 0.1829,
+      "step": 12212
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.153294646208856,
+      "learning_rate": 6.632385665491905e-06,
+      "loss": 0.1828,
+      "step": 12213
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5212721972224903,
+      "learning_rate": 6.630834941976241e-06,
+      "loss": 0.1757,
+      "step": 12214
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8293599964046527,
+      "learning_rate": 6.629284309849373e-06,
+      "loss": 0.1726,
+      "step": 12215
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1039891440162553,
+      "learning_rate": 6.627733769153355e-06,
+      "loss": 0.182,
+      "step": 12216
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9536046109231258,
+      "learning_rate": 6.626183319930253e-06,
+      "loss": 0.1731,
+      "step": 12217
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.04469142650679,
+      "learning_rate": 6.624632962222119e-06,
+      "loss": 0.1766,
+      "step": 12218
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9295249528325686,
+      "learning_rate": 6.6230826960710035e-06,
+      "loss": 0.1728,
+      "step": 12219
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0219966420634277,
+      "learning_rate": 6.621532521518962e-06,
+      "loss": 0.1728,
+      "step": 12220
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3289858108783676,
+      "learning_rate": 6.619982438608039e-06,
+      "loss": 0.1957,
+      "step": 12221
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2373544417696125,
+      "learning_rate": 6.618432447380288e-06,
+      "loss": 0.1997,
+      "step": 12222
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9934359092593344,
+      "learning_rate": 6.616882547877743e-06,
+      "loss": 0.1508,
+      "step": 12223
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8810613838875165,
+      "learning_rate": 6.615332740142454e-06,
+      "loss": 0.1403,
+      "step": 12224
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9879691489052049,
+      "learning_rate": 6.613783024216451e-06,
+      "loss": 0.172,
+      "step": 12225
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9526821220696458,
+      "learning_rate": 6.612233400141781e-06,
+      "loss": 0.1726,
+      "step": 12226
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3838159800652556,
+      "learning_rate": 6.610683867960466e-06,
+      "loss": 0.1858,
+      "step": 12227
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4162296287964915,
+      "learning_rate": 6.6091344277145456e-06,
+      "loss": 0.2036,
+      "step": 12228
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1086836490155574,
+      "learning_rate": 6.6075850794460414e-06,
+      "loss": 0.1688,
+      "step": 12229
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8066030135254134,
+      "learning_rate": 6.60603582319699e-06,
+      "loss": 0.1664,
+      "step": 12230
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4957728270277422,
+      "learning_rate": 6.604486659009404e-06,
+      "loss": 0.1833,
+      "step": 12231
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0987554878137216,
+      "learning_rate": 6.602937586925309e-06,
+      "loss": 0.188,
+      "step": 12232
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1942177727237888,
+      "learning_rate": 6.6013886069867235e-06,
+      "loss": 0.1785,
+      "step": 12233
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8794121473028255,
+      "learning_rate": 6.599839719235668e-06,
+      "loss": 0.1964,
+      "step": 12234
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3678191518713683,
+      "learning_rate": 6.598290923714152e-06,
+      "loss": 0.1797,
+      "step": 12235
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9275715467464166,
+      "learning_rate": 6.596742220464183e-06,
+      "loss": 0.1843,
+      "step": 12236
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7509251617436092,
+      "learning_rate": 6.595193609527774e-06,
+      "loss": 0.1704,
+      "step": 12237
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.261131799563364,
+      "learning_rate": 6.593645090946932e-06,
+      "loss": 0.1765,
+      "step": 12238
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.393015058725475,
+      "learning_rate": 6.592096664763661e-06,
+      "loss": 0.1702,
+      "step": 12239
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.956100047072734,
+      "learning_rate": 6.590548331019957e-06,
+      "loss": 0.1815,
+      "step": 12240
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9956111845289233,
+      "learning_rate": 6.589000089757822e-06,
+      "loss": 0.1566,
+      "step": 12241
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7385208387149487,
+      "learning_rate": 6.587451941019253e-06,
+      "loss": 0.1718,
+      "step": 12242
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7324993874826935,
+      "learning_rate": 6.585903884846245e-06,
+      "loss": 0.1692,
+      "step": 12243
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8834736171907606,
+      "learning_rate": 6.584355921280785e-06,
+      "loss": 0.1764,
+      "step": 12244
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1075738748603206,
+      "learning_rate": 6.582808050364864e-06,
+      "loss": 0.1816,
+      "step": 12245
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8171360429587897,
+      "learning_rate": 6.581260272140466e-06,
+      "loss": 0.1796,
+      "step": 12246
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.085648007323475,
+      "learning_rate": 6.579712586649581e-06,
+      "loss": 0.1715,
+      "step": 12247
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0691469016265858,
+      "learning_rate": 6.5781649939341794e-06,
+      "loss": 0.1849,
+      "step": 12248
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9259858332783487,
+      "learning_rate": 6.5766174940362505e-06,
+      "loss": 0.1606,
+      "step": 12249
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.446395498306882,
+      "learning_rate": 6.575070086997762e-06,
+      "loss": 0.1877,
+      "step": 12250
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9671137700828732,
+      "learning_rate": 6.573522772860692e-06,
+      "loss": 0.1925,
+      "step": 12251
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8860157833790498,
+      "learning_rate": 6.571975551667014e-06,
+      "loss": 0.1821,
+      "step": 12252
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2918582948810673,
+      "learning_rate": 6.570428423458687e-06,
+      "loss": 0.1804,
+      "step": 12253
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3403658730770927,
+      "learning_rate": 6.568881388277685e-06,
+      "loss": 0.1773,
+      "step": 12254
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3153845570882179,
+      "learning_rate": 6.567334446165967e-06,
+      "loss": 0.1775,
+      "step": 12255
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7505597276119826,
+      "learning_rate": 6.565787597165501e-06,
+      "loss": 0.1549,
+      "step": 12256
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7234390957600532,
+      "learning_rate": 6.5642408413182345e-06,
+      "loss": 0.1667,
+      "step": 12257
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9668169901742122,
+      "learning_rate": 6.5626941786661335e-06,
+      "loss": 0.1609,
+      "step": 12258
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9568058195100089,
+      "learning_rate": 6.5611476092511435e-06,
+      "loss": 0.194,
+      "step": 12259
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6035064737348144,
+      "learning_rate": 6.559601133115223e-06,
+      "loss": 0.202,
+      "step": 12260
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.96974975639801,
+      "learning_rate": 6.558054750300313e-06,
+      "loss": 0.1978,
+      "step": 12261
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.79531616090137,
+      "learning_rate": 6.556508460848365e-06,
+      "loss": 0.1568,
+      "step": 12262
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.071490924175172,
+      "learning_rate": 6.554962264801316e-06,
+      "loss": 0.1729,
+      "step": 12263
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.036724040066627,
+      "learning_rate": 6.553416162201114e-06,
+      "loss": 0.1712,
+      "step": 12264
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3034074952249906,
+      "learning_rate": 6.55187015308969e-06,
+      "loss": 0.2014,
+      "step": 12265
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8093630274081415,
+      "learning_rate": 6.550324237508986e-06,
+      "loss": 0.1779,
+      "step": 12266
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.093760841897314,
+      "learning_rate": 6.5487784155009285e-06,
+      "loss": 0.1687,
+      "step": 12267
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9507641187044447,
+      "learning_rate": 6.547232687107453e-06,
+      "loss": 0.1768,
+      "step": 12268
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9106248716503401,
+      "learning_rate": 6.5456870523704845e-06,
+      "loss": 0.1669,
+      "step": 12269
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8055247993376253,
+      "learning_rate": 6.544141511331954e-06,
+      "loss": 0.1783,
+      "step": 12270
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9886344291566992,
+      "learning_rate": 6.542596064033777e-06,
+      "loss": 0.1875,
+      "step": 12271
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0897904038840873,
+      "learning_rate": 6.541050710517875e-06,
+      "loss": 0.1544,
+      "step": 12272
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.99314961842094,
+      "learning_rate": 6.539505450826174e-06,
+      "loss": 0.1726,
+      "step": 12273
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8909882895410705,
+      "learning_rate": 6.537960285000577e-06,
+      "loss": 0.1981,
+      "step": 12274
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2303960200796495,
+      "learning_rate": 6.536415213083007e-06,
+      "loss": 0.1855,
+      "step": 12275
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8916355656067946,
+      "learning_rate": 6.534870235115367e-06,
+      "loss": 0.1833,
+      "step": 12276
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8624614446017977,
+      "learning_rate": 6.533325351139569e-06,
+      "loss": 0.1747,
+      "step": 12277
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.103022254689305,
+      "learning_rate": 6.531780561197514e-06,
+      "loss": 0.1913,
+      "step": 12278
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3641657032827077,
+      "learning_rate": 6.530235865331112e-06,
+      "loss": 0.1624,
+      "step": 12279
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8736367851666924,
+      "learning_rate": 6.528691263582254e-06,
+      "loss": 0.1776,
+      "step": 12280
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9389189226683222,
+      "learning_rate": 6.527146755992844e-06,
+      "loss": 0.1697,
+      "step": 12281
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.536533177111427,
+      "learning_rate": 6.525602342604771e-06,
+      "loss": 0.1688,
+      "step": 12282
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0136640445014258,
+      "learning_rate": 6.524058023459936e-06,
+      "loss": 0.2022,
+      "step": 12283
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9594188684445367,
+      "learning_rate": 6.522513798600219e-06,
+      "loss": 0.1913,
+      "step": 12284
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8637535389066199,
+      "learning_rate": 6.520969668067514e-06,
+      "loss": 0.1587,
+      "step": 12285
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7712698643365918,
+      "learning_rate": 6.5194256319036996e-06,
+      "loss": 0.1583,
+      "step": 12286
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1062185087587382,
+      "learning_rate": 6.517881690150667e-06,
+      "loss": 0.1824,
+      "step": 12287
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8197071505996901,
+      "learning_rate": 6.516337842850291e-06,
+      "loss": 0.1703,
+      "step": 12288
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5756187693803567,
+      "learning_rate": 6.514794090044443e-06,
+      "loss": 0.175,
+      "step": 12289
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7986011584614001,
+      "learning_rate": 6.513250431775003e-06,
+      "loss": 0.1599,
+      "step": 12290
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.876400412985864,
+      "learning_rate": 6.511706868083842e-06,
+      "loss": 0.1826,
+      "step": 12291
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3253744757454275,
+      "learning_rate": 6.510163399012832e-06,
+      "loss": 0.2196,
+      "step": 12292
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8228314003915462,
+      "learning_rate": 6.508620024603833e-06,
+      "loss": 0.1873,
+      "step": 12293
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.182334128744673,
+      "learning_rate": 6.507076744898715e-06,
+      "loss": 0.1665,
+      "step": 12294
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0461347391662688,
+      "learning_rate": 6.505533559939335e-06,
+      "loss": 0.1864,
+      "step": 12295
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.896008704215264,
+      "learning_rate": 6.50399046976756e-06,
+      "loss": 0.2103,
+      "step": 12296
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2823807091290649,
+      "learning_rate": 6.502447474425235e-06,
+      "loss": 0.1633,
+      "step": 12297
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9199393892955969,
+      "learning_rate": 6.5009045739542235e-06,
+      "loss": 0.1988,
+      "step": 12298
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0069335458303679,
+      "learning_rate": 6.499361768396371e-06,
+      "loss": 0.1854,
+      "step": 12299
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0795233172269108,
+      "learning_rate": 6.497819057793531e-06,
+      "loss": 0.1732,
+      "step": 12300
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9021808424439925,
+      "learning_rate": 6.496276442187543e-06,
+      "loss": 0.1833,
+      "step": 12301
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.965671577262932,
+      "learning_rate": 6.49473392162026e-06,
+      "loss": 0.1763,
+      "step": 12302
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0642907663234293,
+      "learning_rate": 6.493191496133513e-06,
+      "loss": 0.1626,
+      "step": 12303
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9872091817808939,
+      "learning_rate": 6.491649165769145e-06,
+      "loss": 0.1797,
+      "step": 12304
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.829784055595061,
+      "learning_rate": 6.4901069305689955e-06,
+      "loss": 0.1834,
+      "step": 12305
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2547817572673163,
+      "learning_rate": 6.488564790574889e-06,
+      "loss": 0.1705,
+      "step": 12306
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2223911861358498,
+      "learning_rate": 6.487022745828663e-06,
+      "loss": 0.1894,
+      "step": 12307
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.015859280347554,
+      "learning_rate": 6.485480796372141e-06,
+      "loss": 0.1863,
+      "step": 12308
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0186987150755753,
+      "learning_rate": 6.483938942247155e-06,
+      "loss": 0.1735,
+      "step": 12309
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9826918667646444,
+      "learning_rate": 6.482397183495519e-06,
+      "loss": 0.1746,
+      "step": 12310
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.171723404689601,
+      "learning_rate": 6.4808555201590614e-06,
+      "loss": 0.1505,
+      "step": 12311
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1341544628466593,
+      "learning_rate": 6.479313952279594e-06,
+      "loss": 0.181,
+      "step": 12312
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9730907582567121,
+      "learning_rate": 6.47777247989894e-06,
+      "loss": 0.1756,
+      "step": 12313
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.165375281334116,
+      "learning_rate": 6.476231103058901e-06,
+      "loss": 0.1869,
+      "step": 12314
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2957891090340417,
+      "learning_rate": 6.474689821801295e-06,
+      "loss": 0.1717,
+      "step": 12315
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8354411998000495,
+      "learning_rate": 6.473148636167925e-06,
+      "loss": 0.1796,
+      "step": 12316
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0039260067099076,
+      "learning_rate": 6.471607546200598e-06,
+      "loss": 0.1761,
+      "step": 12317
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9107910951230012,
+      "learning_rate": 6.470066551941114e-06,
+      "loss": 0.1643,
+      "step": 12318
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.013960752228629,
+      "learning_rate": 6.468525653431279e-06,
+      "loss": 0.1653,
+      "step": 12319
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8127586630711379,
+      "learning_rate": 6.466984850712881e-06,
+      "loss": 0.1688,
+      "step": 12320
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8394294595716062,
+      "learning_rate": 6.4654441438277194e-06,
+      "loss": 0.1725,
+      "step": 12321
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2239762650728694,
+      "learning_rate": 6.463903532817587e-06,
+      "loss": 0.1498,
+      "step": 12322
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5568944801202331,
+      "learning_rate": 6.462363017724267e-06,
+      "loss": 0.182,
+      "step": 12323
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9073127865426454,
+      "learning_rate": 6.460822598589554e-06,
+      "loss": 0.1526,
+      "step": 12324
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2372909375299115,
+      "learning_rate": 6.459282275455223e-06,
+      "loss": 0.1721,
+      "step": 12325
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8625107471525778,
+      "learning_rate": 6.457742048363066e-06,
+      "loss": 0.1461,
+      "step": 12326
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.449375047735124,
+      "learning_rate": 6.456201917354852e-06,
+      "loss": 0.1865,
+      "step": 12327
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.952600062618889,
+      "learning_rate": 6.454661882472364e-06,
+      "loss": 0.1612,
+      "step": 12328
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0525552911764944,
+      "learning_rate": 6.45312194375737e-06,
+      "loss": 0.1621,
+      "step": 12329
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1407915087173905,
+      "learning_rate": 6.451582101251645e-06,
+      "loss": 0.2157,
+      "step": 12330
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9388793084022125,
+      "learning_rate": 6.450042354996954e-06,
+      "loss": 0.1724,
+      "step": 12331
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.195697188482703,
+      "learning_rate": 6.448502705035069e-06,
+      "loss": 0.1737,
+      "step": 12332
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 26.04404863008669,
+      "learning_rate": 6.446963151407743e-06,
+      "loss": 0.1697,
+      "step": 12333
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0686488893519888,
+      "learning_rate": 6.445423694156746e-06,
+      "loss": 0.1445,
+      "step": 12334
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2770191889968054,
+      "learning_rate": 6.44388433332383e-06,
+      "loss": 0.1686,
+      "step": 12335
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9588144049117461,
+      "learning_rate": 6.442345068950755e-06,
+      "loss": 0.1631,
+      "step": 12336
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0817090132952616,
+      "learning_rate": 6.440805901079268e-06,
+      "loss": 0.2015,
+      "step": 12337
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1930738953388942,
+      "learning_rate": 6.4392668297511244e-06,
+      "loss": 0.1689,
+      "step": 12338
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5876164763829481,
+      "learning_rate": 6.4377278550080664e-06,
+      "loss": 0.1798,
+      "step": 12339
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1346038126888687,
+      "learning_rate": 6.436188976891846e-06,
+      "loss": 0.1722,
+      "step": 12340
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8509113581377679,
+      "learning_rate": 6.434650195444199e-06,
+      "loss": 0.1723,
+      "step": 12341
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4524280102048681,
+      "learning_rate": 6.433111510706864e-06,
+      "loss": 0.1895,
+      "step": 12342
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2121512139484962,
+      "learning_rate": 6.431572922721585e-06,
+      "loss": 0.1904,
+      "step": 12343
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0653587248529126,
+      "learning_rate": 6.430034431530088e-06,
+      "loss": 0.1584,
+      "step": 12344
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9241859336426226,
+      "learning_rate": 6.428496037174112e-06,
+      "loss": 0.1727,
+      "step": 12345
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7452389512234971,
+      "learning_rate": 6.42695773969538e-06,
+      "loss": 0.1622,
+      "step": 12346
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2684237230799293,
+      "learning_rate": 6.425419539135622e-06,
+      "loss": 0.1634,
+      "step": 12347
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.237057056550504,
+      "learning_rate": 6.42388143553656e-06,
+      "loss": 0.1799,
+      "step": 12348
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9888821411178348,
+      "learning_rate": 6.422343428939919e-06,
+      "loss": 0.1728,
+      "step": 12349
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4023727794262242,
+      "learning_rate": 6.420805519387412e-06,
+      "loss": 0.1853,
+      "step": 12350
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2704004212987643,
+      "learning_rate": 6.419267706920758e-06,
+      "loss": 0.1814,
+      "step": 12351
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9049368121777028,
+      "learning_rate": 6.417729991581668e-06,
+      "loss": 0.1794,
+      "step": 12352
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0791248063852146,
+      "learning_rate": 6.4161923734118594e-06,
+      "loss": 0.178,
+      "step": 12353
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0154705817518106,
+      "learning_rate": 6.41465485245303e-06,
+      "loss": 0.1718,
+      "step": 12354
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8059650877129417,
+      "learning_rate": 6.413117428746892e-06,
+      "loss": 0.1903,
+      "step": 12355
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8510180569734818,
+      "learning_rate": 6.4115801023351444e-06,
+      "loss": 0.1621,
+      "step": 12356
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0253418963036458,
+      "learning_rate": 6.410042873259494e-06,
+      "loss": 0.147,
+      "step": 12357
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1512407414739831,
+      "learning_rate": 6.408505741561633e-06,
+      "loss": 0.1798,
+      "step": 12358
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9846029617594317,
+      "learning_rate": 6.406968707283253e-06,
+      "loss": 0.1722,
+      "step": 12359
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2493645732998286,
+      "learning_rate": 6.405431770466051e-06,
+      "loss": 0.1677,
+      "step": 12360
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9348559490222239,
+      "learning_rate": 6.403894931151714e-06,
+      "loss": 0.1615,
+      "step": 12361
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9569479373941275,
+      "learning_rate": 6.4023581893819345e-06,
+      "loss": 0.1838,
+      "step": 12362
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7611388126812884,
+      "learning_rate": 6.4008215451983864e-06,
+      "loss": 0.1756,
+      "step": 12363
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2822095789309997,
+      "learning_rate": 6.399284998642761e-06,
+      "loss": 0.2031,
+      "step": 12364
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7669714612070854,
+      "learning_rate": 6.39774854975673e-06,
+      "loss": 0.17,
+      "step": 12365
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9943848290476993,
+      "learning_rate": 6.396212198581978e-06,
+      "loss": 0.1678,
+      "step": 12366
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.708205758868274,
+      "learning_rate": 6.394675945160169e-06,
+      "loss": 0.1662,
+      "step": 12367
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5553428840830763,
+      "learning_rate": 6.39313978953298e-06,
+      "loss": 0.1749,
+      "step": 12368
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8969706376294244,
+      "learning_rate": 6.391603731742078e-06,
+      "loss": 0.1733,
+      "step": 12369
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1488854694275947,
+      "learning_rate": 6.390067771829132e-06,
+      "loss": 0.1677,
+      "step": 12370
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9769116755166073,
+      "learning_rate": 6.3885319098357966e-06,
+      "loss": 0.1584,
+      "step": 12371
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8587655311387853,
+      "learning_rate": 6.386996145803741e-06,
+      "loss": 0.1848,
+      "step": 12372
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6151992229384966,
+      "learning_rate": 6.385460479774616e-06,
+      "loss": 0.1781,
+      "step": 12373
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.161128554908505,
+      "learning_rate": 6.383924911790081e-06,
+      "loss": 0.1978,
+      "step": 12374
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.102166422327023,
+      "learning_rate": 6.3823894418917895e-06,
+      "loss": 0.1716,
+      "step": 12375
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9866300969961047,
+      "learning_rate": 6.380854070121385e-06,
+      "loss": 0.1741,
+      "step": 12376
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8744583736841051,
+      "learning_rate": 6.37931879652052e-06,
+      "loss": 0.1776,
+      "step": 12377
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9033550694190644,
+      "learning_rate": 6.377783621130834e-06,
+      "loss": 0.1815,
+      "step": 12378
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7952117507679426,
+      "learning_rate": 6.376248543993977e-06,
+      "loss": 0.1769,
+      "step": 12379
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0221285316727067,
+      "learning_rate": 6.374713565151579e-06,
+      "loss": 0.1809,
+      "step": 12380
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9361376699622568,
+      "learning_rate": 6.373178684645283e-06,
+      "loss": 0.1831,
+      "step": 12381
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.000278952974784,
+      "learning_rate": 6.371643902516715e-06,
+      "loss": 0.2024,
+      "step": 12382
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8337946910176021,
+      "learning_rate": 6.3701092188075176e-06,
+      "loss": 0.1733,
+      "step": 12383
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.068934840910689,
+      "learning_rate": 6.368574633559308e-06,
+      "loss": 0.1867,
+      "step": 12384
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0862813904471362,
+      "learning_rate": 6.367040146813721e-06,
+      "loss": 0.1759,
+      "step": 12385
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9875804752664659,
+      "learning_rate": 6.365505758612371e-06,
+      "loss": 0.1781,
+      "step": 12386
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0816791746408998,
+      "learning_rate": 6.363971468996883e-06,
+      "loss": 0.1785,
+      "step": 12387
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2488081898099574,
+      "learning_rate": 6.362437278008875e-06,
+      "loss": 0.1764,
+      "step": 12388
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1154057711008427,
+      "learning_rate": 6.360903185689964e-06,
+      "loss": 0.1629,
+      "step": 12389
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7958664729630673,
+      "learning_rate": 6.359369192081756e-06,
+      "loss": 0.164,
+      "step": 12390
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8140188803014982,
+      "learning_rate": 6.357835297225865e-06,
+      "loss": 0.1648,
+      "step": 12391
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9086263255788167,
+      "learning_rate": 6.356301501163901e-06,
+      "loss": 0.1498,
+      "step": 12392
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1784567810825564,
+      "learning_rate": 6.3547678039374595e-06,
+      "loss": 0.1702,
+      "step": 12393
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9426996480766275,
+      "learning_rate": 6.35323420558815e-06,
+      "loss": 0.1861,
+      "step": 12394
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8746147919053582,
+      "learning_rate": 6.351700706157565e-06,
+      "loss": 0.1816,
+      "step": 12395
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9674403403160589,
+      "learning_rate": 6.350167305687309e-06,
+      "loss": 0.1854,
+      "step": 12396
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3616418400384807,
+      "learning_rate": 6.348634004218969e-06,
+      "loss": 0.1557,
+      "step": 12397
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.191532683707872,
+      "learning_rate": 6.3471008017941396e-06,
+      "loss": 0.1812,
+      "step": 12398
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9497595078806705,
+      "learning_rate": 6.345567698454405e-06,
+      "loss": 0.1836,
+      "step": 12399
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.061604308953469,
+      "learning_rate": 6.344034694241353e-06,
+      "loss": 0.1713,
+      "step": 12400
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9765112821432018,
+      "learning_rate": 6.342501789196565e-06,
+      "loss": 0.1707,
+      "step": 12401
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9497694404575935,
+      "learning_rate": 6.340968983361629e-06,
+      "loss": 0.1837,
+      "step": 12402
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1070548884603109,
+      "learning_rate": 6.339436276778108e-06,
+      "loss": 0.1638,
+      "step": 12403
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5135514613785872,
+      "learning_rate": 6.33790366948759e-06,
+      "loss": 0.1823,
+      "step": 12404
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9890899244681153,
+      "learning_rate": 6.3363711615316384e-06,
+      "loss": 0.1621,
+      "step": 12405
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.085404642674482,
+      "learning_rate": 6.334838752951829e-06,
+      "loss": 0.2035,
+      "step": 12406
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.821864051302577,
+      "learning_rate": 6.333306443789723e-06,
+      "loss": 0.1637,
+      "step": 12407
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.018056615727074,
+      "learning_rate": 6.331774234086888e-06,
+      "loss": 0.1798,
+      "step": 12408
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1267528713054624,
+      "learning_rate": 6.330242123884882e-06,
+      "loss": 0.1932,
+      "step": 12409
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9204954399604649,
+      "learning_rate": 6.328710113225271e-06,
+      "loss": 0.1607,
+      "step": 12410
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.134368920284439,
+      "learning_rate": 6.327178202149604e-06,
+      "loss": 0.1712,
+      "step": 12411
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8948304924405651,
+      "learning_rate": 6.325646390699432e-06,
+      "loss": 0.1707,
+      "step": 12412
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.148432251960856,
+      "learning_rate": 6.324114678916312e-06,
+      "loss": 0.1651,
+      "step": 12413
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3243261950780123,
+      "learning_rate": 6.322583066841787e-06,
+      "loss": 0.2044,
+      "step": 12414
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0739468116695186,
+      "learning_rate": 6.321051554517406e-06,
+      "loss": 0.1714,
+      "step": 12415
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8085880976560276,
+      "learning_rate": 6.3195201419847075e-06,
+      "loss": 0.1876,
+      "step": 12416
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2755854149125554,
+      "learning_rate": 6.3179888292852345e-06,
+      "loss": 0.1897,
+      "step": 12417
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8972934946635283,
+      "learning_rate": 6.316457616460521e-06,
+      "loss": 0.175,
+      "step": 12418
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8943975090876413,
+      "learning_rate": 6.314926503552106e-06,
+      "loss": 0.1985,
+      "step": 12419
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0943262305329209,
+      "learning_rate": 6.313395490601513e-06,
+      "loss": 0.182,
+      "step": 12420
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9752335207142642,
+      "learning_rate": 6.311864577650278e-06,
+      "loss": 0.1536,
+      "step": 12421
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.255788513507667,
+      "learning_rate": 6.310333764739922e-06,
+      "loss": 0.1692,
+      "step": 12422
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8946464195463422,
+      "learning_rate": 6.308803051911977e-06,
+      "loss": 0.1734,
+      "step": 12423
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0774757219799522,
+      "learning_rate": 6.307272439207952e-06,
+      "loss": 0.2096,
+      "step": 12424
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8735326158800345,
+      "learning_rate": 6.305741926669376e-06,
+      "loss": 0.1648,
+      "step": 12425
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7804412362106942,
+      "learning_rate": 6.304211514337755e-06,
+      "loss": 0.1792,
+      "step": 12426
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8648967530320314,
+      "learning_rate": 6.302681202254605e-06,
+      "loss": 0.1727,
+      "step": 12427
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.951635997434637,
+      "learning_rate": 6.30115099046144e-06,
+      "loss": 0.1663,
+      "step": 12428
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8540674273278183,
+      "learning_rate": 6.299620878999759e-06,
+      "loss": 0.1723,
+      "step": 12429
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8448351618492159,
+      "learning_rate": 6.298090867911073e-06,
+      "loss": 0.1673,
+      "step": 12430
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8999928356265918,
+      "learning_rate": 6.296560957236879e-06,
+      "loss": 0.1645,
+      "step": 12431
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0263325239331658,
+      "learning_rate": 6.295031147018682e-06,
+      "loss": 0.1971,
+      "step": 12432
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8343603797823569,
+      "learning_rate": 6.293501437297971e-06,
+      "loss": 0.1734,
+      "step": 12433
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2727832738429121,
+      "learning_rate": 6.291971828116244e-06,
+      "loss": 0.1658,
+      "step": 12434
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9658412176768761,
+      "learning_rate": 6.290442319514989e-06,
+      "loss": 0.1793,
+      "step": 12435
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2050351172057916,
+      "learning_rate": 6.288912911535701e-06,
+      "loss": 0.174,
+      "step": 12436
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2826664564756307,
+      "learning_rate": 6.2873836042198546e-06,
+      "loss": 0.1864,
+      "step": 12437
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7933695499610162,
+      "learning_rate": 6.285854397608941e-06,
+      "loss": 0.1513,
+      "step": 12438
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8704491729302477,
+      "learning_rate": 6.284325291744433e-06,
+      "loss": 0.1652,
+      "step": 12439
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8751594756844792,
+      "learning_rate": 6.282796286667814e-06,
+      "loss": 0.179,
+      "step": 12440
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.152411953728966,
+      "learning_rate": 6.281267382420553e-06,
+      "loss": 0.191,
+      "step": 12441
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0783429025431217,
+      "learning_rate": 6.2797385790441275e-06,
+      "loss": 0.1865,
+      "step": 12442
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.165853509016558,
+      "learning_rate": 6.278209876580002e-06,
+      "loss": 0.1671,
+      "step": 12443
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4369216323054037,
+      "learning_rate": 6.2766812750696425e-06,
+      "loss": 0.1707,
+      "step": 12444
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1509816814159848,
+      "learning_rate": 6.275152774554518e-06,
+      "loss": 0.1796,
+      "step": 12445
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0754526489308998,
+      "learning_rate": 6.273624375076079e-06,
+      "loss": 0.178,
+      "step": 12446
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9512043301201804,
+      "learning_rate": 6.272096076675794e-06,
+      "loss": 0.1706,
+      "step": 12447
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9603013115681072,
+      "learning_rate": 6.2705678793951085e-06,
+      "loss": 0.2025,
+      "step": 12448
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0168091469068452,
+      "learning_rate": 6.269039783275486e-06,
+      "loss": 0.1876,
+      "step": 12449
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3743891386632616,
+      "learning_rate": 6.267511788358365e-06,
+      "loss": 0.1839,
+      "step": 12450
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8779440589301847,
+      "learning_rate": 6.265983894685199e-06,
+      "loss": 0.1658,
+      "step": 12451
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9168517762241354,
+      "learning_rate": 6.264456102297431e-06,
+      "loss": 0.1808,
+      "step": 12452
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.478232498156621,
+      "learning_rate": 6.262928411236504e-06,
+      "loss": 0.1785,
+      "step": 12453
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.850537886066106,
+      "learning_rate": 6.261400821543853e-06,
+      "loss": 0.1629,
+      "step": 12454
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.945733684470051,
+      "learning_rate": 6.259873333260917e-06,
+      "loss": 0.175,
+      "step": 12455
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1153704099256443,
+      "learning_rate": 6.258345946429127e-06,
+      "loss": 0.1927,
+      "step": 12456
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0169887154193404,
+      "learning_rate": 6.256818661089914e-06,
+      "loss": 0.1717,
+      "step": 12457
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5532608557851424,
+      "learning_rate": 6.255291477284706e-06,
+      "loss": 0.1806,
+      "step": 12458
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8137842106602222,
+      "learning_rate": 6.253764395054931e-06,
+      "loss": 0.1781,
+      "step": 12459
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.778704510268674,
+      "learning_rate": 6.252237414442006e-06,
+      "loss": 0.1645,
+      "step": 12460
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9806884124431697,
+      "learning_rate": 6.250710535487354e-06,
+      "loss": 0.1584,
+      "step": 12461
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7972196638544355,
+      "learning_rate": 6.249183758232391e-06,
+      "loss": 0.1791,
+      "step": 12462
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8762787480998664,
+      "learning_rate": 6.247657082718528e-06,
+      "loss": 0.1787,
+      "step": 12463
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5548117723420225,
+      "learning_rate": 6.246130508987181e-06,
+      "loss": 0.1589,
+      "step": 12464
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9578502567917115,
+      "learning_rate": 6.244604037079754e-06,
+      "loss": 0.1862,
+      "step": 12465
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1923637270836631,
+      "learning_rate": 6.2430776670376565e-06,
+      "loss": 0.1628,
+      "step": 12466
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.132564443816655,
+      "learning_rate": 6.241551398902288e-06,
+      "loss": 0.1805,
+      "step": 12467
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.297131069014862,
+      "learning_rate": 6.240025232715052e-06,
+      "loss": 0.1896,
+      "step": 12468
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0555852137876773,
+      "learning_rate": 6.2384991685173415e-06,
+      "loss": 0.1655,
+      "step": 12469
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2255988170345553,
+      "learning_rate": 6.236973206350554e-06,
+      "loss": 0.1969,
+      "step": 12470
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.857306568176004,
+      "learning_rate": 6.23544734625608e-06,
+      "loss": 0.1773,
+      "step": 12471
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4411426640641447,
+      "learning_rate": 6.233921588275313e-06,
+      "loss": 0.1704,
+      "step": 12472
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.081017069593761,
+      "learning_rate": 6.232395932449632e-06,
+      "loss": 0.179,
+      "step": 12473
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9309147223998671,
+      "learning_rate": 6.230870378820426e-06,
+      "loss": 0.1675,
+      "step": 12474
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3926729763150256,
+      "learning_rate": 6.22934492742907e-06,
+      "loss": 0.1871,
+      "step": 12475
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.05205076758149,
+      "learning_rate": 6.2278195783169525e-06,
+      "loss": 0.1799,
+      "step": 12476
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9209143818079015,
+      "learning_rate": 6.226294331525437e-06,
+      "loss": 0.1982,
+      "step": 12477
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9332230467328521,
+      "learning_rate": 6.224769187095903e-06,
+      "loss": 0.1717,
+      "step": 12478
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8431590555877424,
+      "learning_rate": 6.223244145069715e-06,
+      "loss": 0.1714,
+      "step": 12479
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9699148561201477,
+      "learning_rate": 6.221719205488248e-06,
+      "loss": 0.1887,
+      "step": 12480
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8557638152178924,
+      "learning_rate": 6.220194368392862e-06,
+      "loss": 0.1801,
+      "step": 12481
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0081281462198912,
+      "learning_rate": 6.218669633824911e-06,
+      "loss": 0.1779,
+      "step": 12482
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9963855583977796,
+      "learning_rate": 6.2171450018257625e-06,
+      "loss": 0.1647,
+      "step": 12483
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.812626859159139,
+      "learning_rate": 6.2156204724367674e-06,
+      "loss": 0.1848,
+      "step": 12484
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1443046290186372,
+      "learning_rate": 6.214096045699285e-06,
+      "loss": 0.1837,
+      "step": 12485
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0078120851075847,
+      "learning_rate": 6.212571721654658e-06,
+      "loss": 0.1919,
+      "step": 12486
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9909980525748541,
+      "learning_rate": 6.211047500344239e-06,
+      "loss": 0.1795,
+      "step": 12487
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9126111150104824,
+      "learning_rate": 6.209523381809366e-06,
+      "loss": 0.1545,
+      "step": 12488
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9009539115573664,
+      "learning_rate": 6.207999366091392e-06,
+      "loss": 0.1921,
+      "step": 12489
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7706876863747594,
+      "learning_rate": 6.206475453231644e-06,
+      "loss": 0.1635,
+      "step": 12490
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0875709472808894,
+      "learning_rate": 6.204951643271466e-06,
+      "loss": 0.1667,
+      "step": 12491
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.060787182075934,
+      "learning_rate": 6.2034279362521866e-06,
+      "loss": 0.1804,
+      "step": 12492
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0740900720219337,
+      "learning_rate": 6.201904332215143e-06,
+      "loss": 0.1743,
+      "step": 12493
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9956019682721083,
+      "learning_rate": 6.200380831201655e-06,
+      "loss": 0.1902,
+      "step": 12494
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8991276352608935,
+      "learning_rate": 6.198857433253056e-06,
+      "loss": 0.1836,
+      "step": 12495
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0371989188888475,
+      "learning_rate": 6.19733413841066e-06,
+      "loss": 0.1809,
+      "step": 12496
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.694860645984785,
+      "learning_rate": 6.1958109467157925e-06,
+      "loss": 0.1632,
+      "step": 12497
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.22687686411913,
+      "learning_rate": 6.1942878582097685e-06,
+      "loss": 0.1845,
+      "step": 12498
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.976504010769984,
+      "learning_rate": 6.192764872933899e-06,
+      "loss": 0.1771,
+      "step": 12499
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0657364619208907,
+      "learning_rate": 6.191241990929498e-06,
+      "loss": 0.1724,
+      "step": 12500
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0860099335201034,
+      "learning_rate": 6.1897192122378714e-06,
+      "loss": 0.1886,
+      "step": 12501
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9996622761610917,
+      "learning_rate": 6.18819653690033e-06,
+      "loss": 0.1871,
+      "step": 12502
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1164634795618917,
+      "learning_rate": 6.18667396495817e-06,
+      "loss": 0.1976,
+      "step": 12503
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8602778593863528,
+      "learning_rate": 6.185151496452695e-06,
+      "loss": 0.1791,
+      "step": 12504
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.145776520803997,
+      "learning_rate": 6.1836291314252e-06,
+      "loss": 0.1649,
+      "step": 12505
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.152848083543941,
+      "learning_rate": 6.182106869916984e-06,
+      "loss": 0.1775,
+      "step": 12506
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6093748241813428,
+      "learning_rate": 6.180584711969331e-06,
+      "loss": 0.1949,
+      "step": 12507
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1685262483998686,
+      "learning_rate": 6.179062657623536e-06,
+      "loss": 0.2048,
+      "step": 12508
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0364475450915975,
+      "learning_rate": 6.17754070692088e-06,
+      "loss": 0.1723,
+      "step": 12509
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.137829764269063,
+      "learning_rate": 6.17601885990265e-06,
+      "loss": 0.1862,
+      "step": 12510
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9867987488422779,
+      "learning_rate": 6.174497116610121e-06,
+      "loss": 0.1683,
+      "step": 12511
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1026191481551983,
+      "learning_rate": 6.1729754770845795e-06,
+      "loss": 0.1665,
+      "step": 12512
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.263035244463893,
+      "learning_rate": 6.171453941367289e-06,
+      "loss": 0.1601,
+      "step": 12513
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9018081110097101,
+      "learning_rate": 6.1699325094995284e-06,
+      "loss": 0.1554,
+      "step": 12514
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8358416616204396,
+      "learning_rate": 6.168411181522569e-06,
+      "loss": 0.1863,
+      "step": 12515
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3526473813417075,
+      "learning_rate": 6.1668899574776665e-06,
+      "loss": 0.1791,
+      "step": 12516
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.074541065702537,
+      "learning_rate": 6.165368837406094e-06,
+      "loss": 0.1742,
+      "step": 12517
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.266699220462641,
+      "learning_rate": 6.1638478213491045e-06,
+      "loss": 0.1811,
+      "step": 12518
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1978464666007855,
+      "learning_rate": 6.162326909347964e-06,
+      "loss": 0.1596,
+      "step": 12519
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9428725976666876,
+      "learning_rate": 6.160806101443919e-06,
+      "loss": 0.2014,
+      "step": 12520
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2711091819725837,
+      "learning_rate": 6.159285397678231e-06,
+      "loss": 0.1745,
+      "step": 12521
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8918084712482156,
+      "learning_rate": 6.157764798092139e-06,
+      "loss": 0.169,
+      "step": 12522
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1336763768588909,
+      "learning_rate": 6.156244302726894e-06,
+      "loss": 0.1734,
+      "step": 12523
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8880467909572677,
+      "learning_rate": 6.154723911623739e-06,
+      "loss": 0.1854,
+      "step": 12524
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2989210515788927,
+      "learning_rate": 6.153203624823918e-06,
+      "loss": 0.1713,
+      "step": 12525
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.845146306048577,
+      "learning_rate": 6.151683442368662e-06,
+      "loss": 0.1709,
+      "step": 12526
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9185256844844901,
+      "learning_rate": 6.150163364299213e-06,
+      "loss": 0.1749,
+      "step": 12527
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.797325566838219,
+      "learning_rate": 6.148643390656797e-06,
+      "loss": 0.1799,
+      "step": 12528
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.075427656446058,
+      "learning_rate": 6.147123521482652e-06,
+      "loss": 0.2035,
+      "step": 12529
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4026009435662508,
+      "learning_rate": 6.145603756817994e-06,
+      "loss": 0.174,
+      "step": 12530
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1020490093254394,
+      "learning_rate": 6.144084096704054e-06,
+      "loss": 0.1807,
+      "step": 12531
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7151474986946946,
+      "learning_rate": 6.142564541182052e-06,
+      "loss": 0.1773,
+      "step": 12532
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8427975962075497,
+      "learning_rate": 6.141045090293203e-06,
+      "loss": 0.1482,
+      "step": 12533
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4389259546810442,
+      "learning_rate": 6.1395257440787246e-06,
+      "loss": 0.1911,
+      "step": 12534
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1012280045114835,
+      "learning_rate": 6.1380065025798275e-06,
+      "loss": 0.1771,
+      "step": 12535
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9123329869617892,
+      "learning_rate": 6.136487365837723e-06,
+      "loss": 0.1756,
+      "step": 12536
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8711802443983307,
+      "learning_rate": 6.134968333893614e-06,
+      "loss": 0.1823,
+      "step": 12537
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8182388385244617,
+      "learning_rate": 6.133449406788712e-06,
+      "loss": 0.1746,
+      "step": 12538
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7628933579636543,
+      "learning_rate": 6.13193058456421e-06,
+      "loss": 0.1636,
+      "step": 12539
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0452787874927696,
+      "learning_rate": 6.13041186726131e-06,
+      "loss": 0.1752,
+      "step": 12540
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1334682158832199,
+      "learning_rate": 6.128893254921204e-06,
+      "loss": 0.1732,
+      "step": 12541
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9651552788366533,
+      "learning_rate": 6.127374747585093e-06,
+      "loss": 0.1817,
+      "step": 12542
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.003239206354539,
+      "learning_rate": 6.125856345294156e-06,
+      "loss": 0.1699,
+      "step": 12543
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0489086733374817,
+      "learning_rate": 6.124338048089586e-06,
+      "loss": 0.1975,
+      "step": 12544
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6214492669460177,
+      "learning_rate": 6.122819856012564e-06,
+      "loss": 0.1545,
+      "step": 12545
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9810971851715937,
+      "learning_rate": 6.121301769104277e-06,
+      "loss": 0.1707,
+      "step": 12546
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9002404769464839,
+      "learning_rate": 6.119783787405893e-06,
+      "loss": 0.1655,
+      "step": 12547
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9440022409126835,
+      "learning_rate": 6.118265910958599e-06,
+      "loss": 0.1719,
+      "step": 12548
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8769209904214279,
+      "learning_rate": 6.116748139803554e-06,
+      "loss": 0.1809,
+      "step": 12549
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.039009549593771,
+      "learning_rate": 6.115230473981939e-06,
+      "loss": 0.18,
+      "step": 12550
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8954688446998333,
+      "learning_rate": 6.113712913534919e-06,
+      "loss": 0.1879,
+      "step": 12551
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8868688380705804,
+      "learning_rate": 6.1121954585036525e-06,
+      "loss": 0.1731,
+      "step": 12552
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9068755428729511,
+      "learning_rate": 6.110678108929304e-06,
+      "loss": 0.1735,
+      "step": 12553
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8247271190545595,
+      "learning_rate": 6.109160864853031e-06,
+      "loss": 0.1609,
+      "step": 12554
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8450076184849792,
+      "learning_rate": 6.107643726315993e-06,
+      "loss": 0.1572,
+      "step": 12555
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8092402376512072,
+      "learning_rate": 6.106126693359334e-06,
+      "loss": 0.1653,
+      "step": 12556
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8671305954230574,
+      "learning_rate": 6.104609766024211e-06,
+      "loss": 0.1621,
+      "step": 12557
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9829882238021675,
+      "learning_rate": 6.103092944351766e-06,
+      "loss": 0.1917,
+      "step": 12558
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9452635175929243,
+      "learning_rate": 6.1015762283831485e-06,
+      "loss": 0.225,
+      "step": 12559
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9586078615730995,
+      "learning_rate": 6.100059618159493e-06,
+      "loss": 0.1717,
+      "step": 12560
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0276062367385321,
+      "learning_rate": 6.098543113721942e-06,
+      "loss": 0.1911,
+      "step": 12561
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.928138926530368,
+      "learning_rate": 6.097026715111627e-06,
+      "loss": 0.1668,
+      "step": 12562
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5221052583071613,
+      "learning_rate": 6.095510422369687e-06,
+      "loss": 0.1712,
+      "step": 12563
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0930100736865827,
+      "learning_rate": 6.093994235537244e-06,
+      "loss": 0.1807,
+      "step": 12564
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8970169138993683,
+      "learning_rate": 6.092478154655431e-06,
+      "loss": 0.1762,
+      "step": 12565
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8136805000756253,
+      "learning_rate": 6.090962179765365e-06,
+      "loss": 0.1645,
+      "step": 12566
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0157663073300447,
+      "learning_rate": 6.089446310908174e-06,
+      "loss": 0.1983,
+      "step": 12567
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4187437187365042,
+      "learning_rate": 6.087930548124973e-06,
+      "loss": 0.1778,
+      "step": 12568
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8998068081604149,
+      "learning_rate": 6.086414891456873e-06,
+      "loss": 0.2043,
+      "step": 12569
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.23602924999552,
+      "learning_rate": 6.084899340944993e-06,
+      "loss": 0.1731,
+      "step": 12570
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0015833512001684,
+      "learning_rate": 6.083383896630437e-06,
+      "loss": 0.181,
+      "step": 12571
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1870337585167698,
+      "learning_rate": 6.081868558554318e-06,
+      "loss": 0.1765,
+      "step": 12572
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.041099892707607,
+      "learning_rate": 6.080353326757732e-06,
+      "loss": 0.1851,
+      "step": 12573
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9109856770529133,
+      "learning_rate": 6.078838201281785e-06,
+      "loss": 0.1775,
+      "step": 12574
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8825147559118742,
+      "learning_rate": 6.077323182167572e-06,
+      "loss": 0.1584,
+      "step": 12575
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9183584383098351,
+      "learning_rate": 6.075808269456191e-06,
+      "loss": 0.1842,
+      "step": 12576
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8299857824314143,
+      "learning_rate": 6.074293463188731e-06,
+      "loss": 0.1685,
+      "step": 12577
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.815838704444408,
+      "learning_rate": 6.072778763406285e-06,
+      "loss": 0.171,
+      "step": 12578
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8010028661546602,
+      "learning_rate": 6.071264170149933e-06,
+      "loss": 0.1635,
+      "step": 12579
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0617631073565157,
+      "learning_rate": 6.069749683460765e-06,
+      "loss": 0.1699,
+      "step": 12580
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8113835607544837,
+      "learning_rate": 6.068235303379857e-06,
+      "loss": 0.1667,
+      "step": 12581
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7552633298934263,
+      "learning_rate": 6.066721029948291e-06,
+      "loss": 0.1869,
+      "step": 12582
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8411484353367596,
+      "learning_rate": 6.065206863207136e-06,
+      "loss": 0.1598,
+      "step": 12583
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1801608714988914,
+      "learning_rate": 6.06369280319747e-06,
+      "loss": 0.1732,
+      "step": 12584
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9325254706005588,
+      "learning_rate": 6.062178849960359e-06,
+      "loss": 0.1748,
+      "step": 12585
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8197909597762826,
+      "learning_rate": 6.060665003536868e-06,
+      "loss": 0.1739,
+      "step": 12586
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8392306554506301,
+      "learning_rate": 6.059151263968061e-06,
+      "loss": 0.1668,
+      "step": 12587
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9631135744680915,
+      "learning_rate": 6.057637631294997e-06,
+      "loss": 0.1712,
+      "step": 12588
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9695296619769307,
+      "learning_rate": 6.0561241055587385e-06,
+      "loss": 0.1689,
+      "step": 12589
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9120100410242605,
+      "learning_rate": 6.054610686800333e-06,
+      "loss": 0.1876,
+      "step": 12590
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7689129787731763,
+      "learning_rate": 6.053097375060839e-06,
+      "loss": 0.1474,
+      "step": 12591
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.12835739084911,
+      "learning_rate": 6.051584170381298e-06,
+      "loss": 0.1648,
+      "step": 12592
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7884327382625383,
+      "learning_rate": 6.050071072802761e-06,
+      "loss": 0.154,
+      "step": 12593
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9505681884985597,
+      "learning_rate": 6.048558082366269e-06,
+      "loss": 0.173,
+      "step": 12594
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1454112915904329,
+      "learning_rate": 6.047045199112865e-06,
+      "loss": 0.1908,
+      "step": 12595
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2481260398457952,
+      "learning_rate": 6.045532423083578e-06,
+      "loss": 0.1882,
+      "step": 12596
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0597575897436244,
+      "learning_rate": 6.04401975431945e-06,
+      "loss": 0.1596,
+      "step": 12597
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9179478487130808,
+      "learning_rate": 6.042507192861509e-06,
+      "loss": 0.1782,
+      "step": 12598
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1493989123689738,
+      "learning_rate": 6.040994738750788e-06,
+      "loss": 0.1578,
+      "step": 12599
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1523802744773493,
+      "learning_rate": 6.039482392028302e-06,
+      "loss": 0.162,
+      "step": 12600
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9364003989788462,
+      "learning_rate": 6.037970152735083e-06,
+      "loss": 0.1816,
+      "step": 12601
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9241871003791564,
+      "learning_rate": 6.036458020912151e-06,
+      "loss": 0.1792,
+      "step": 12602
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8505269578328596,
+      "learning_rate": 6.034945996600512e-06,
+      "loss": 0.1693,
+      "step": 12603
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.070369815758006,
+      "learning_rate": 6.033434079841192e-06,
+      "loss": 0.1704,
+      "step": 12604
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7797409791480164,
+      "learning_rate": 6.031922270675193e-06,
+      "loss": 0.1754,
+      "step": 12605
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9061342005549365,
+      "learning_rate": 6.0304105691435285e-06,
+      "loss": 0.1869,
+      "step": 12606
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2881616547284978,
+      "learning_rate": 6.028898975287199e-06,
+      "loss": 0.1554,
+      "step": 12607
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3061192645383484,
+      "learning_rate": 6.027387489147214e-06,
+      "loss": 0.1809,
+      "step": 12608
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1680155757501567,
+      "learning_rate": 6.025876110764563e-06,
+      "loss": 0.1547,
+      "step": 12609
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8902388411726355,
+      "learning_rate": 6.02436484018025e-06,
+      "loss": 0.1789,
+      "step": 12610
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8750040119051603,
+      "learning_rate": 6.022853677435262e-06,
+      "loss": 0.1615,
+      "step": 12611
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0092482811801622,
+      "learning_rate": 6.021342622570597e-06,
+      "loss": 0.1881,
+      "step": 12612
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0436717957655843,
+      "learning_rate": 6.019831675627235e-06,
+      "loss": 0.184,
+      "step": 12613
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0653181286038373,
+      "learning_rate": 6.018320836646164e-06,
+      "loss": 0.1948,
+      "step": 12614
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.280309670639057,
+      "learning_rate": 6.016810105668365e-06,
+      "loss": 0.1689,
+      "step": 12615
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.231377345198479,
+      "learning_rate": 6.015299482734819e-06,
+      "loss": 0.1624,
+      "step": 12616
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7934969383349629,
+      "learning_rate": 6.013788967886496e-06,
+      "loss": 0.1635,
+      "step": 12617
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9639688222591635,
+      "learning_rate": 6.012278561164377e-06,
+      "loss": 0.2132,
+      "step": 12618
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8849522972008587,
+      "learning_rate": 6.010768262609425e-06,
+      "loss": 0.162,
+      "step": 12619
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0277230971550373,
+      "learning_rate": 6.009258072262607e-06,
+      "loss": 0.1852,
+      "step": 12620
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0009072609625371,
+      "learning_rate": 6.0077479901648935e-06,
+      "loss": 0.149,
+      "step": 12621
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.136258000812448,
+      "learning_rate": 6.006238016357238e-06,
+      "loss": 0.1721,
+      "step": 12622
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8486374392042341,
+      "learning_rate": 6.0047281508806035e-06,
+      "loss": 0.1714,
+      "step": 12623
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9263199151783211,
+      "learning_rate": 6.00321839377594e-06,
+      "loss": 0.1878,
+      "step": 12624
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8905235912213837,
+      "learning_rate": 6.001708745084209e-06,
+      "loss": 0.1812,
+      "step": 12625
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0831038098071641,
+      "learning_rate": 6.000199204846348e-06,
+      "loss": 0.1646,
+      "step": 12626
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.742740153937068,
+      "learning_rate": 5.998689773103314e-06,
+      "loss": 0.1742,
+      "step": 12627
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8225453319335321,
+      "learning_rate": 5.997180449896043e-06,
+      "loss": 0.1609,
+      "step": 12628
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9766197525212346,
+      "learning_rate": 5.995671235265483e-06,
+      "loss": 0.2027,
+      "step": 12629
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8413232819910352,
+      "learning_rate": 5.994162129252561e-06,
+      "loss": 0.2092,
+      "step": 12630
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.6884841068191743,
+      "learning_rate": 5.992653131898223e-06,
+      "loss": 0.1658,
+      "step": 12631
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.094459052333219,
+      "learning_rate": 5.991144243243392e-06,
+      "loss": 0.1712,
+      "step": 12632
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5800840062814814,
+      "learning_rate": 5.989635463329e-06,
+      "loss": 0.1637,
+      "step": 12633
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8334496722945175,
+      "learning_rate": 5.988126792195972e-06,
+      "loss": 0.161,
+      "step": 12634
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0395517635284837,
+      "learning_rate": 5.986618229885234e-06,
+      "loss": 0.1679,
+      "step": 12635
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1216746978677126,
+      "learning_rate": 5.985109776437699e-06,
+      "loss": 0.1818,
+      "step": 12636
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.139123745295039,
+      "learning_rate": 5.983601431894291e-06,
+      "loss": 0.1953,
+      "step": 12637
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8956866938954258,
+      "learning_rate": 5.982093196295924e-06,
+      "loss": 0.1775,
+      "step": 12638
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.939937101816068,
+      "learning_rate": 5.9805850696835e-06,
+      "loss": 0.1668,
+      "step": 12639
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9370917688385507,
+      "learning_rate": 5.979077052097936e-06,
+      "loss": 0.1718,
+      "step": 12640
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9764839246852787,
+      "learning_rate": 5.977569143580132e-06,
+      "loss": 0.1749,
+      "step": 12641
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.857365213898431,
+      "learning_rate": 5.976061344170995e-06,
+      "loss": 0.1691,
+      "step": 12642
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.000499061221936,
+      "learning_rate": 5.974553653911419e-06,
+      "loss": 0.1784,
+      "step": 12643
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8623887963384143,
+      "learning_rate": 5.973046072842305e-06,
+      "loss": 0.1772,
+      "step": 12644
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.749589177560146,
+      "learning_rate": 5.971538601004542e-06,
+      "loss": 0.1734,
+      "step": 12645
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1602944806291715,
+      "learning_rate": 5.970031238439023e-06,
+      "loss": 0.1871,
+      "step": 12646
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8914992617167609,
+      "learning_rate": 5.968523985186632e-06,
+      "loss": 0.1541,
+      "step": 12647
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7662373818745853,
+      "learning_rate": 5.967016841288258e-06,
+      "loss": 0.1553,
+      "step": 12648
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1534363779308747,
+      "learning_rate": 5.965509806784777e-06,
+      "loss": 0.1768,
+      "step": 12649
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.753453688553895,
+      "learning_rate": 5.964002881717073e-06,
+      "loss": 0.1986,
+      "step": 12650
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.079062233365335,
+      "learning_rate": 5.962496066126018e-06,
+      "loss": 0.1727,
+      "step": 12651
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7864402552268221,
+      "learning_rate": 5.960989360052487e-06,
+      "loss": 0.1631,
+      "step": 12652
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3834907670938499,
+      "learning_rate": 5.959482763537344e-06,
+      "loss": 0.2107,
+      "step": 12653
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8888693044438032,
+      "learning_rate": 5.9579762766214624e-06,
+      "loss": 0.166,
+      "step": 12654
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.041141009502544,
+      "learning_rate": 5.956469899345704e-06,
+      "loss": 0.1655,
+      "step": 12655
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.14499034697284,
+      "learning_rate": 5.954963631750923e-06,
+      "loss": 0.1748,
+      "step": 12656
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7490457820595138,
+      "learning_rate": 5.953457473877988e-06,
+      "loss": 0.1484,
+      "step": 12657
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1030787733107046,
+      "learning_rate": 5.9519514257677416e-06,
+      "loss": 0.1658,
+      "step": 12658
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1598241699401937,
+      "learning_rate": 5.950445487461045e-06,
+      "loss": 0.1905,
+      "step": 12659
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3936158789622728,
+      "learning_rate": 5.94893965899874e-06,
+      "loss": 0.1799,
+      "step": 12660
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9178583971284034,
+      "learning_rate": 5.947433940421681e-06,
+      "loss": 0.1729,
+      "step": 12661
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8496088616896511,
+      "learning_rate": 5.9459283317707e-06,
+      "loss": 0.1826,
+      "step": 12662
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.973870769727679,
+      "learning_rate": 5.944422833086645e-06,
+      "loss": 0.169,
+      "step": 12663
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7730557051735637,
+      "learning_rate": 5.942917444410346e-06,
+      "loss": 0.1772,
+      "step": 12664
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8075041620874908,
+      "learning_rate": 5.941412165782645e-06,
+      "loss": 0.1764,
+      "step": 12665
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8686461335552085,
+      "learning_rate": 5.939906997244364e-06,
+      "loss": 0.1511,
+      "step": 12666
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.94060984040844,
+      "learning_rate": 5.938401938836339e-06,
+      "loss": 0.1719,
+      "step": 12667
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0649589557653525,
+      "learning_rate": 5.936896990599388e-06,
+      "loss": 0.2042,
+      "step": 12668
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4109876552900762,
+      "learning_rate": 5.9353921525743394e-06,
+      "loss": 0.1642,
+      "step": 12669
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9844463011964707,
+      "learning_rate": 5.933887424802003e-06,
+      "loss": 0.174,
+      "step": 12670
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8399276919431213,
+      "learning_rate": 5.9323828073232025e-06,
+      "loss": 0.1947,
+      "step": 12671
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3259786476616642,
+      "learning_rate": 5.930878300178751e-06,
+      "loss": 0.1677,
+      "step": 12672
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0698628567492015,
+      "learning_rate": 5.929373903409451e-06,
+      "loss": 0.1983,
+      "step": 12673
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9322445915079713,
+      "learning_rate": 5.9278696170561175e-06,
+      "loss": 0.1894,
+      "step": 12674
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0557079540673695,
+      "learning_rate": 5.926365441159547e-06,
+      "loss": 0.1894,
+      "step": 12675
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8609547739005059,
+      "learning_rate": 5.924861375760547e-06,
+      "loss": 0.1706,
+      "step": 12676
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1008738440055796,
+      "learning_rate": 5.923357420899908e-06,
+      "loss": 0.1648,
+      "step": 12677
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2090115629624052,
+      "learning_rate": 5.921853576618435e-06,
+      "loss": 0.1855,
+      "step": 12678
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6107429435083536,
+      "learning_rate": 5.920349842956909e-06,
+      "loss": 0.1756,
+      "step": 12679
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9722248729245764,
+      "learning_rate": 5.918846219956126e-06,
+      "loss": 0.1689,
+      "step": 12680
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8651383834716352,
+      "learning_rate": 5.917342707656868e-06,
+      "loss": 0.1806,
+      "step": 12681
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0928700656041868,
+      "learning_rate": 5.915839306099924e-06,
+      "loss": 0.1755,
+      "step": 12682
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1338462339538051,
+      "learning_rate": 5.9143360153260655e-06,
+      "loss": 0.1763,
+      "step": 12683
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9910656380198765,
+      "learning_rate": 5.912832835376074e-06,
+      "loss": 0.1552,
+      "step": 12684
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0793574090034495,
+      "learning_rate": 5.911329766290723e-06,
+      "loss": 0.1802,
+      "step": 12685
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9980755060477906,
+      "learning_rate": 5.9098268081107855e-06,
+      "loss": 0.1837,
+      "step": 12686
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.013160068361227,
+      "learning_rate": 5.9083239608770225e-06,
+      "loss": 0.1778,
+      "step": 12687
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8906471754734917,
+      "learning_rate": 5.9068212246302084e-06,
+      "loss": 0.1656,
+      "step": 12688
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8351175530599811,
+      "learning_rate": 5.9053185994110975e-06,
+      "loss": 0.1603,
+      "step": 12689
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2711286630467495,
+      "learning_rate": 5.903816085260447e-06,
+      "loss": 0.1776,
+      "step": 12690
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.928024545479977,
+      "learning_rate": 5.902313682219023e-06,
+      "loss": 0.1774,
+      "step": 12691
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6213176678927754,
+      "learning_rate": 5.9008113903275675e-06,
+      "loss": 0.1622,
+      "step": 12692
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9402086897119339,
+      "learning_rate": 5.899309209626836e-06,
+      "loss": 0.1659,
+      "step": 12693
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0009811336498147,
+      "learning_rate": 5.8978071401575724e-06,
+      "loss": 0.1802,
+      "step": 12694
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9628070707067192,
+      "learning_rate": 5.896305181960524e-06,
+      "loss": 0.1425,
+      "step": 12695
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7368779807655511,
+      "learning_rate": 5.894803335076427e-06,
+      "loss": 0.1902,
+      "step": 12696
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9819880577436897,
+      "learning_rate": 5.8933015995460215e-06,
+      "loss": 0.1744,
+      "step": 12697
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1420654284111416,
+      "learning_rate": 5.8917999754100415e-06,
+      "loss": 0.1839,
+      "step": 12698
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8692782871985733,
+      "learning_rate": 5.890298462709224e-06,
+      "loss": 0.175,
+      "step": 12699
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.945136366281455,
+      "learning_rate": 5.888797061484288e-06,
+      "loss": 0.1802,
+      "step": 12700
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8843237227460644,
+      "learning_rate": 5.887295771775968e-06,
+      "loss": 0.1718,
+      "step": 12701
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2556922365483774,
+      "learning_rate": 5.885794593624978e-06,
+      "loss": 0.1818,
+      "step": 12702
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9307613969627148,
+      "learning_rate": 5.884293527072045e-06,
+      "loss": 0.1816,
+      "step": 12703
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9934432917494341,
+      "learning_rate": 5.88279257215788e-06,
+      "loss": 0.1676,
+      "step": 12704
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9168098448675442,
+      "learning_rate": 5.881291728923202e-06,
+      "loss": 0.1724,
+      "step": 12705
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9078438921079398,
+      "learning_rate": 5.8797909974087166e-06,
+      "loss": 0.1746,
+      "step": 12706
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9363191791158435,
+      "learning_rate": 5.878290377655134e-06,
+      "loss": 0.1869,
+      "step": 12707
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8066846892597944,
+      "learning_rate": 5.876789869703159e-06,
+      "loss": 0.1675,
+      "step": 12708
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1161164446921086,
+      "learning_rate": 5.875289473593489e-06,
+      "loss": 0.1872,
+      "step": 12709
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8283708904113805,
+      "learning_rate": 5.8737891893668255e-06,
+      "loss": 0.1833,
+      "step": 12710
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9024395515456646,
+      "learning_rate": 5.872289017063861e-06,
+      "loss": 0.1809,
+      "step": 12711
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.864448371808424,
+      "learning_rate": 5.8707889567252965e-06,
+      "loss": 0.1474,
+      "step": 12712
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9419293887724072,
+      "learning_rate": 5.869289008391809e-06,
+      "loss": 0.1964,
+      "step": 12713
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.14700717152764,
+      "learning_rate": 5.8677891721040945e-06,
+      "loss": 0.1757,
+      "step": 12714
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8869622133560018,
+      "learning_rate": 5.866289447902829e-06,
+      "loss": 0.2008,
+      "step": 12715
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4162474657019513,
+      "learning_rate": 5.864789835828697e-06,
+      "loss": 0.1503,
+      "step": 12716
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0489769046663078,
+      "learning_rate": 5.863290335922371e-06,
+      "loss": 0.1688,
+      "step": 12717
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1416131733336279,
+      "learning_rate": 5.861790948224535e-06,
+      "loss": 0.2004,
+      "step": 12718
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7944055566838395,
+      "learning_rate": 5.860291672775847e-06,
+      "loss": 0.1544,
+      "step": 12719
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1178239682169353,
+      "learning_rate": 5.858792509616984e-06,
+      "loss": 0.1741,
+      "step": 12720
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9122154797112343,
+      "learning_rate": 5.857293458788607e-06,
+      "loss": 0.1751,
+      "step": 12721
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9993588874074452,
+      "learning_rate": 5.855794520331382e-06,
+      "loss": 0.1545,
+      "step": 12722
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9449544422557149,
+      "learning_rate": 5.854295694285961e-06,
+      "loss": 0.1748,
+      "step": 12723
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0191948675121945,
+      "learning_rate": 5.852796980693005e-06,
+      "loss": 0.1866,
+      "step": 12724
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9340845236893027,
+      "learning_rate": 5.8512983795931665e-06,
+      "loss": 0.1717,
+      "step": 12725
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9371937668222649,
+      "learning_rate": 5.8497998910270915e-06,
+      "loss": 0.1793,
+      "step": 12726
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8167577931936453,
+      "learning_rate": 5.848301515035433e-06,
+      "loss": 0.1629,
+      "step": 12727
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9070997042355996,
+      "learning_rate": 5.846803251658824e-06,
+      "loss": 0.1653,
+      "step": 12728
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2774767895726513,
+      "learning_rate": 5.8453051009379145e-06,
+      "loss": 0.1733,
+      "step": 12729
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.753620675806103,
+      "learning_rate": 5.843807062913338e-06,
+      "loss": 0.1849,
+      "step": 12730
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9565505103888103,
+      "learning_rate": 5.842309137625732e-06,
+      "loss": 0.1838,
+      "step": 12731
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5290954314169458,
+      "learning_rate": 5.840811325115723e-06,
+      "loss": 0.1738,
+      "step": 12732
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.979728759438856,
+      "learning_rate": 5.8393136254239424e-06,
+      "loss": 0.1738,
+      "step": 12733
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8586160518040348,
+      "learning_rate": 5.837816038591016e-06,
+      "loss": 0.1518,
+      "step": 12734
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0986488341282836,
+      "learning_rate": 5.836318564657561e-06,
+      "loss": 0.1572,
+      "step": 12735
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.970494590695129,
+      "learning_rate": 5.8348212036642004e-06,
+      "loss": 0.206,
+      "step": 12736
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1964209550960125,
+      "learning_rate": 5.833323955651555e-06,
+      "loss": 0.1912,
+      "step": 12737
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7420502057662117,
+      "learning_rate": 5.831826820660228e-06,
+      "loss": 0.1594,
+      "step": 12738
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9275542125230787,
+      "learning_rate": 5.8303297987308384e-06,
+      "loss": 0.1778,
+      "step": 12739
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4329471083008936,
+      "learning_rate": 5.828832889903983e-06,
+      "loss": 0.1749,
+      "step": 12740
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0909640951660313,
+      "learning_rate": 5.827336094220278e-06,
+      "loss": 0.1742,
+      "step": 12741
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8744495369447978,
+      "learning_rate": 5.825839411720314e-06,
+      "loss": 0.1972,
+      "step": 12742
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1558371119444353,
+      "learning_rate": 5.824342842444689e-06,
+      "loss": 0.1768,
+      "step": 12743
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0440422547177555,
+      "learning_rate": 5.822846386434e-06,
+      "loss": 0.1891,
+      "step": 12744
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9694946099458643,
+      "learning_rate": 5.82135004372884e-06,
+      "loss": 0.1823,
+      "step": 12745
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8690001007512806,
+      "learning_rate": 5.819853814369798e-06,
+      "loss": 0.1734,
+      "step": 12746
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9975380839997835,
+      "learning_rate": 5.818357698397455e-06,
+      "loss": 0.1795,
+      "step": 12747
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.109368298211608,
+      "learning_rate": 5.816861695852398e-06,
+      "loss": 0.1872,
+      "step": 12748
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9314307134735961,
+      "learning_rate": 5.815365806775201e-06,
+      "loss": 0.1967,
+      "step": 12749
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1428904285367345,
+      "learning_rate": 5.813870031206448e-06,
+      "loss": 0.1875,
+      "step": 12750
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9960408697173141,
+      "learning_rate": 5.812374369186701e-06,
+      "loss": 0.1655,
+      "step": 12751
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2568265105055656,
+      "learning_rate": 5.8108788207565355e-06,
+      "loss": 0.1743,
+      "step": 12752
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8431837679769912,
+      "learning_rate": 5.8093833859565196e-06,
+      "loss": 0.1824,
+      "step": 12753
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1127045480558082,
+      "learning_rate": 5.80788806482722e-06,
+      "loss": 0.1508,
+      "step": 12754
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0746524208411485,
+      "learning_rate": 5.806392857409189e-06,
+      "loss": 0.213,
+      "step": 12755
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9716316543874394,
+      "learning_rate": 5.8048977637429925e-06,
+      "loss": 0.1939,
+      "step": 12756
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3417170022179468,
+      "learning_rate": 5.803402783869178e-06,
+      "loss": 0.1571,
+      "step": 12757
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9285050897401939,
+      "learning_rate": 5.801907917828303e-06,
+      "loss": 0.1817,
+      "step": 12758
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.824448289024012,
+      "learning_rate": 5.800413165660913e-06,
+      "loss": 0.1817,
+      "step": 12759
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.9595828542615563,
+      "learning_rate": 5.798918527407549e-06,
+      "loss": 0.1864,
+      "step": 12760
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9974578852520701,
+      "learning_rate": 5.797424003108758e-06,
+      "loss": 0.1507,
+      "step": 12761
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9464195385603775,
+      "learning_rate": 5.795929592805077e-06,
+      "loss": 0.16,
+      "step": 12762
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1374103732685328,
+      "learning_rate": 5.794435296537049e-06,
+      "loss": 0.1781,
+      "step": 12763
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1827366485411164,
+      "learning_rate": 5.7929411143451955e-06,
+      "loss": 0.1668,
+      "step": 12764
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.168608591288615,
+      "learning_rate": 5.791447046270055e-06,
+      "loss": 0.1834,
+      "step": 12765
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9011955344686472,
+      "learning_rate": 5.78995309235215e-06,
+      "loss": 0.1653,
+      "step": 12766
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4317286725861893,
+      "learning_rate": 5.788459252632008e-06,
+      "loss": 0.1901,
+      "step": 12767
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3077605328902828,
+      "learning_rate": 5.7869655271501415e-06,
+      "loss": 0.1947,
+      "step": 12768
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8625789708204211,
+      "learning_rate": 5.785471915947078e-06,
+      "loss": 0.1683,
+      "step": 12769
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9455906607045814,
+      "learning_rate": 5.783978419063323e-06,
+      "loss": 0.1848,
+      "step": 12770
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2640924961109428,
+      "learning_rate": 5.782485036539391e-06,
+      "loss": 0.1574,
+      "step": 12771
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0937487504731338,
+      "learning_rate": 5.7809917684157915e-06,
+      "loss": 0.1685,
+      "step": 12772
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0232287650417144,
+      "learning_rate": 5.779498614733032e-06,
+      "loss": 0.1958,
+      "step": 12773
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9582307803099454,
+      "learning_rate": 5.778005575531606e-06,
+      "loss": 0.173,
+      "step": 12774
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.08070120123374,
+      "learning_rate": 5.7765126508520216e-06,
+      "loss": 0.1669,
+      "step": 12775
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.898585807582454,
+      "learning_rate": 5.775019840734768e-06,
+      "loss": 0.2019,
+      "step": 12776
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8173973148695457,
+      "learning_rate": 5.773527145220341e-06,
+      "loss": 0.1813,
+      "step": 12777
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1002081814255371,
+      "learning_rate": 5.772034564349227e-06,
+      "loss": 0.166,
+      "step": 12778
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9156571269819482,
+      "learning_rate": 5.770542098161913e-06,
+      "loss": 0.1703,
+      "step": 12779
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9475744155041969,
+      "learning_rate": 5.769049746698889e-06,
+      "loss": 0.1599,
+      "step": 12780
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.027253670713997,
+      "learning_rate": 5.767557510000624e-06,
+      "loss": 0.1714,
+      "step": 12781
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9115685276161519,
+      "learning_rate": 5.7660653881076045e-06,
+      "loss": 0.1624,
+      "step": 12782
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9189978581613581,
+      "learning_rate": 5.7645733810602975e-06,
+      "loss": 0.1615,
+      "step": 12783
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6394894428716966,
+      "learning_rate": 5.76308148889918e-06,
+      "loss": 0.2053,
+      "step": 12784
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3596295229586994,
+      "learning_rate": 5.761589711664714e-06,
+      "loss": 0.1913,
+      "step": 12785
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8407133594687236,
+      "learning_rate": 5.760098049397369e-06,
+      "loss": 0.1632,
+      "step": 12786
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9167228532336161,
+      "learning_rate": 5.7586065021376e-06,
+      "loss": 0.1623,
+      "step": 12787
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0360576474570997,
+      "learning_rate": 5.7571150699258695e-06,
+      "loss": 0.2142,
+      "step": 12788
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1085304668570894,
+      "learning_rate": 5.7556237528026325e-06,
+      "loss": 0.1695,
+      "step": 12789
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0896331639110117,
+      "learning_rate": 5.754132550808345e-06,
+      "loss": 0.1871,
+      "step": 12790
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0283976013604725,
+      "learning_rate": 5.752641463983446e-06,
+      "loss": 0.1893,
+      "step": 12791
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9949917672413399,
+      "learning_rate": 5.751150492368394e-06,
+      "loss": 0.17,
+      "step": 12792
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0405359463683412,
+      "learning_rate": 5.749659636003619e-06,
+      "loss": 0.1585,
+      "step": 12793
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6514389831675638,
+      "learning_rate": 5.748168894929571e-06,
+      "loss": 0.1664,
+      "step": 12794
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9551165337416043,
+      "learning_rate": 5.746678269186682e-06,
+      "loss": 0.1687,
+      "step": 12795
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7885972333474208,
+      "learning_rate": 5.7451877588153805e-06,
+      "loss": 0.1671,
+      "step": 12796
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9364830001912637,
+      "learning_rate": 5.743697363856103e-06,
+      "loss": 0.1723,
+      "step": 12797
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2316441009063175,
+      "learning_rate": 5.742207084349274e-06,
+      "loss": 0.1726,
+      "step": 12798
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9264953776930172,
+      "learning_rate": 5.740716920335321e-06,
+      "loss": 0.1784,
+      "step": 12799
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0204461207639197,
+      "learning_rate": 5.739226871854659e-06,
+      "loss": 0.1939,
+      "step": 12800
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.665770547649148,
+      "learning_rate": 5.737736938947713e-06,
+      "loss": 0.1666,
+      "step": 12801
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8495805423139131,
+      "learning_rate": 5.73624712165489e-06,
+      "loss": 0.1623,
+      "step": 12802
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0323761081415743,
+      "learning_rate": 5.734757420016608e-06,
+      "loss": 0.1841,
+      "step": 12803
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0378100442313927,
+      "learning_rate": 5.733267834073267e-06,
+      "loss": 0.1752,
+      "step": 12804
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1876253510119588,
+      "learning_rate": 5.731778363865278e-06,
+      "loss": 0.1599,
+      "step": 12805
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.041679902340334,
+      "learning_rate": 5.730289009433041e-06,
+      "loss": 0.1814,
+      "step": 12806
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9306180346304497,
+      "learning_rate": 5.7287997708169615e-06,
+      "loss": 0.1777,
+      "step": 12807
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8422275117744921,
+      "learning_rate": 5.7273106480574245e-06,
+      "loss": 0.1842,
+      "step": 12808
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9099700432256908,
+      "learning_rate": 5.725821641194831e-06,
+      "loss": 0.1839,
+      "step": 12809
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0335046237387424,
+      "learning_rate": 5.724332750269563e-06,
+      "loss": 0.1808,
+      "step": 12810
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7342455579777192,
+      "learning_rate": 5.722843975322015e-06,
+      "loss": 0.1691,
+      "step": 12811
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.021904470086725,
+      "learning_rate": 5.721355316392566e-06,
+      "loss": 0.1678,
+      "step": 12812
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8413994432957763,
+      "learning_rate": 5.719866773521592e-06,
+      "loss": 0.1739,
+      "step": 12813
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7656323654203047,
+      "learning_rate": 5.718378346749473e-06,
+      "loss": 0.173,
+      "step": 12814
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8695568552072197,
+      "learning_rate": 5.716890036116582e-06,
+      "loss": 0.1816,
+      "step": 12815
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.637543553189384,
+      "learning_rate": 5.715401841663296e-06,
+      "loss": 0.1538,
+      "step": 12816
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8986297573930118,
+      "learning_rate": 5.713913763429972e-06,
+      "loss": 0.1765,
+      "step": 12817
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.165151118458408,
+      "learning_rate": 5.712425801456984e-06,
+      "loss": 0.1858,
+      "step": 12818
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2275823408117625,
+      "learning_rate": 5.710937955784686e-06,
+      "loss": 0.1808,
+      "step": 12819
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0067028818244466,
+      "learning_rate": 5.709450226453439e-06,
+      "loss": 0.2044,
+      "step": 12820
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7775459709638857,
+      "learning_rate": 5.707962613503595e-06,
+      "loss": 0.1662,
+      "step": 12821
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4527849280844114,
+      "learning_rate": 5.706475116975512e-06,
+      "loss": 0.2129,
+      "step": 12822
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9436189868348397,
+      "learning_rate": 5.704987736909529e-06,
+      "loss": 0.1751,
+      "step": 12823
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0299563360182766,
+      "learning_rate": 5.703500473345995e-06,
+      "loss": 0.1849,
+      "step": 12824
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3373956988323639,
+      "learning_rate": 5.702013326325256e-06,
+      "loss": 0.1731,
+      "step": 12825
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6450668643735773,
+      "learning_rate": 5.700526295887649e-06,
+      "loss": 0.1551,
+      "step": 12826
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4047761777979644,
+      "learning_rate": 5.699039382073508e-06,
+      "loss": 0.159,
+      "step": 12827
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9549652310257992,
+      "learning_rate": 5.69755258492317e-06,
+      "loss": 0.1792,
+      "step": 12828
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8410795021030328,
+      "learning_rate": 5.6960659044769596e-06,
+      "loss": 0.1677,
+      "step": 12829
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8490038369881001,
+      "learning_rate": 5.694579340775202e-06,
+      "loss": 0.1716,
+      "step": 12830
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0215088608250185,
+      "learning_rate": 5.693092893858223e-06,
+      "loss": 0.1724,
+      "step": 12831
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.801362536481239,
+      "learning_rate": 5.691606563766341e-06,
+      "loss": 0.1882,
+      "step": 12832
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1402567424615575,
+      "learning_rate": 5.6901203505398805e-06,
+      "loss": 0.1853,
+      "step": 12833
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.134326759007839,
+      "learning_rate": 5.688634254219143e-06,
+      "loss": 0.1587,
+      "step": 12834
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9233574208843645,
+      "learning_rate": 5.687148274844449e-06,
+      "loss": 0.167,
+      "step": 12835
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7866757718859764,
+      "learning_rate": 5.6856624124560985e-06,
+      "loss": 0.1697,
+      "step": 12836
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0403016153270666,
+      "learning_rate": 5.684176667094403e-06,
+      "loss": 0.1639,
+      "step": 12837
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.605615224634009,
+      "learning_rate": 5.682691038799655e-06,
+      "loss": 0.1811,
+      "step": 12838
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9004022258891372,
+      "learning_rate": 5.68120552761216e-06,
+      "loss": 0.2044,
+      "step": 12839
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9103926230618025,
+      "learning_rate": 5.6797201335722064e-06,
+      "loss": 0.181,
+      "step": 12840
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9655309677333925,
+      "learning_rate": 5.678234856720086e-06,
+      "loss": 0.1693,
+      "step": 12841
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9890053890231713,
+      "learning_rate": 5.67674969709609e-06,
+      "loss": 0.1832,
+      "step": 12842
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.561738973816005,
+      "learning_rate": 5.675264654740506e-06,
+      "loss": 0.1944,
+      "step": 12843
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.112333814532282,
+      "learning_rate": 5.67377972969361e-06,
+      "loss": 0.1898,
+      "step": 12844
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8642933030164457,
+      "learning_rate": 5.672294921995687e-06,
+      "loss": 0.186,
+      "step": 12845
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9086013400978419,
+      "learning_rate": 5.670810231687004e-06,
+      "loss": 0.16,
+      "step": 12846
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.064045502304253,
+      "learning_rate": 5.669325658807843e-06,
+      "loss": 0.1797,
+      "step": 12847
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3122020336260767,
+      "learning_rate": 5.667841203398463e-06,
+      "loss": 0.1961,
+      "step": 12848
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.319368124499809,
+      "learning_rate": 5.666356865499134e-06,
+      "loss": 0.1906,
+      "step": 12849
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.129651978672975,
+      "learning_rate": 5.664872645150126e-06,
+      "loss": 0.1678,
+      "step": 12850
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8886433832662072,
+      "learning_rate": 5.663388542391687e-06,
+      "loss": 0.1729,
+      "step": 12851
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8454505714582762,
+      "learning_rate": 5.661904557264083e-06,
+      "loss": 0.1826,
+      "step": 12852
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.909024630136565,
+      "learning_rate": 5.6604206898075595e-06,
+      "loss": 0.1766,
+      "step": 12853
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7422197954953109,
+      "learning_rate": 5.658936940062373e-06,
+      "loss": 0.1818,
+      "step": 12854
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9900409014516947,
+      "learning_rate": 5.657453308068763e-06,
+      "loss": 0.1645,
+      "step": 12855
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0872223642690009,
+      "learning_rate": 5.655969793866982e-06,
+      "loss": 0.1686,
+      "step": 12856
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1858898783551484,
+      "learning_rate": 5.654486397497262e-06,
+      "loss": 0.172,
+      "step": 12857
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.69732530121704,
+      "learning_rate": 5.653003118999843e-06,
+      "loss": 0.1616,
+      "step": 12858
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0433921844175187,
+      "learning_rate": 5.651519958414961e-06,
+      "loss": 0.1705,
+      "step": 12859
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.003929183034557,
+      "learning_rate": 5.650036915782849e-06,
+      "loss": 0.1941,
+      "step": 12860
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7799695862625812,
+      "learning_rate": 5.648553991143728e-06,
+      "loss": 0.1883,
+      "step": 12861
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9482306796546718,
+      "learning_rate": 5.647071184537829e-06,
+      "loss": 0.1592,
+      "step": 12862
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9391228858816858,
+      "learning_rate": 5.6455884960053655e-06,
+      "loss": 0.1697,
+      "step": 12863
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0291871921144702,
+      "learning_rate": 5.6441059255865645e-06,
+      "loss": 0.1647,
+      "step": 12864
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9809468989705948,
+      "learning_rate": 5.642623473321638e-06,
+      "loss": 0.179,
+      "step": 12865
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0666735024576603,
+      "learning_rate": 5.64114113925079e-06,
+      "loss": 0.1894,
+      "step": 12866
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9172747178146132,
+      "learning_rate": 5.639658923414235e-06,
+      "loss": 0.1778,
+      "step": 12867
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9556114312786075,
+      "learning_rate": 5.638176825852178e-06,
+      "loss": 0.1634,
+      "step": 12868
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8090140618805673,
+      "learning_rate": 5.636694846604825e-06,
+      "loss": 0.1591,
+      "step": 12869
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8509532820931227,
+      "learning_rate": 5.635212985712366e-06,
+      "loss": 0.1727,
+      "step": 12870
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.221724279621428,
+      "learning_rate": 5.633731243215007e-06,
+      "loss": 0.1786,
+      "step": 12871
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9225094148693288,
+      "learning_rate": 5.63224961915293e-06,
+      "loss": 0.1659,
+      "step": 12872
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.789810703620753,
+      "learning_rate": 5.6307681135663315e-06,
+      "loss": 0.1734,
+      "step": 12873
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8656698646274308,
+      "learning_rate": 5.629286726495393e-06,
+      "loss": 0.1606,
+      "step": 12874
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9659152400307496,
+      "learning_rate": 5.627805457980298e-06,
+      "loss": 0.1709,
+      "step": 12875
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0158763651419542,
+      "learning_rate": 5.626324308061226e-06,
+      "loss": 0.183,
+      "step": 12876
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9541796052338082,
+      "learning_rate": 5.624843276778358e-06,
+      "loss": 0.1869,
+      "step": 12877
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7038071148065027,
+      "learning_rate": 5.62336236417186e-06,
+      "loss": 0.1458,
+      "step": 12878
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8309038062333706,
+      "learning_rate": 5.621881570281909e-06,
+      "loss": 0.1536,
+      "step": 12879
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1898632042402766,
+      "learning_rate": 5.6204008951486636e-06,
+      "loss": 0.1907,
+      "step": 12880
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9013225322317349,
+      "learning_rate": 5.618920338812295e-06,
+      "loss": 0.1781,
+      "step": 12881
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8402968879388498,
+      "learning_rate": 5.61743990131296e-06,
+      "loss": 0.1772,
+      "step": 12882
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2585658510983027,
+      "learning_rate": 5.615959582690812e-06,
+      "loss": 0.2075,
+      "step": 12883
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0642373692264568,
+      "learning_rate": 5.614479382986007e-06,
+      "loss": 0.1954,
+      "step": 12884
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8121988348561141,
+      "learning_rate": 5.612999302238696e-06,
+      "loss": 0.1607,
+      "step": 12885
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9981702401051303,
+      "learning_rate": 5.611519340489031e-06,
+      "loss": 0.1698,
+      "step": 12886
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.01944572613643,
+      "learning_rate": 5.610039497777149e-06,
+      "loss": 0.1731,
+      "step": 12887
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2615462260108041,
+      "learning_rate": 5.608559774143196e-06,
+      "loss": 0.1783,
+      "step": 12888
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2403582568494242,
+      "learning_rate": 5.607080169627304e-06,
+      "loss": 0.1756,
+      "step": 12889
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8311748159017751,
+      "learning_rate": 5.6056006842696145e-06,
+      "loss": 0.1722,
+      "step": 12890
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5292143098094482,
+      "learning_rate": 5.60412131811025e-06,
+      "loss": 0.1706,
+      "step": 12891
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1798060723173425,
+      "learning_rate": 5.6026420711893485e-06,
+      "loss": 0.2004,
+      "step": 12892
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9015841379843128,
+      "learning_rate": 5.601162943547023e-06,
+      "loss": 0.16,
+      "step": 12893
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1723385310959136,
+      "learning_rate": 5.599683935223402e-06,
+      "loss": 0.152,
+      "step": 12894
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1839672034771838,
+      "learning_rate": 5.598205046258603e-06,
+      "loss": 0.1623,
+      "step": 12895
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3569216140387987,
+      "learning_rate": 5.596726276692745e-06,
+      "loss": 0.1868,
+      "step": 12896
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9215860827965177,
+      "learning_rate": 5.5952476265659315e-06,
+      "loss": 0.1629,
+      "step": 12897
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2795995011922863,
+      "learning_rate": 5.593769095918278e-06,
+      "loss": 0.189,
+      "step": 12898
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0940029725833822,
+      "learning_rate": 5.592290684789887e-06,
+      "loss": 0.1557,
+      "step": 12899
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.846845827744864,
+      "learning_rate": 5.5908123932208565e-06,
+      "loss": 0.1829,
+      "step": 12900
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0259373068826925,
+      "learning_rate": 5.589334221251289e-06,
+      "loss": 0.1777,
+      "step": 12901
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9493271016656901,
+      "learning_rate": 5.587856168921279e-06,
+      "loss": 0.1728,
+      "step": 12902
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.137369310293687,
+      "learning_rate": 5.586378236270925e-06,
+      "loss": 0.1685,
+      "step": 12903
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8186890126068809,
+      "learning_rate": 5.584900423340306e-06,
+      "loss": 0.177,
+      "step": 12904
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.218637687925129,
+      "learning_rate": 5.5834227301695166e-06,
+      "loss": 0.2081,
+      "step": 12905
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6991846261385979,
+      "learning_rate": 5.581945156798629e-06,
+      "loss": 0.1579,
+      "step": 12906
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0576293401743173,
+      "learning_rate": 5.580467703267736e-06,
+      "loss": 0.1876,
+      "step": 12907
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8104066484939919,
+      "learning_rate": 5.578990369616899e-06,
+      "loss": 0.1455,
+      "step": 12908
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8376631152931486,
+      "learning_rate": 5.577513155886204e-06,
+      "loss": 0.1837,
+      "step": 12909
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8523762348499792,
+      "learning_rate": 5.576036062115709e-06,
+      "loss": 0.1626,
+      "step": 12910
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9641752481361844,
+      "learning_rate": 5.574559088345487e-06,
+      "loss": 0.1724,
+      "step": 12911
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9164492673667283,
+      "learning_rate": 5.573082234615599e-06,
+      "loss": 0.1857,
+      "step": 12912
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.019011994175837,
+      "learning_rate": 5.57160550096611e-06,
+      "loss": 0.1662,
+      "step": 12913
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9286318752863383,
+      "learning_rate": 5.570128887437067e-06,
+      "loss": 0.1789,
+      "step": 12914
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2982218911285703,
+      "learning_rate": 5.568652394068532e-06,
+      "loss": 0.1752,
+      "step": 12915
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1241713257522084,
+      "learning_rate": 5.567176020900549e-06,
+      "loss": 0.1662,
+      "step": 12916
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.113905632611933,
+      "learning_rate": 5.565699767973169e-06,
+      "loss": 0.1792,
+      "step": 12917
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8693497758333975,
+      "learning_rate": 5.564223635326433e-06,
+      "loss": 0.1824,
+      "step": 12918
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2387170423762486,
+      "learning_rate": 5.562747623000379e-06,
+      "loss": 0.1758,
+      "step": 12919
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1359411489145923,
+      "learning_rate": 5.561271731035045e-06,
+      "loss": 0.2008,
+      "step": 12920
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9097943534585743,
+      "learning_rate": 5.559795959470467e-06,
+      "loss": 0.1707,
+      "step": 12921
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0803842029847395,
+      "learning_rate": 5.558320308346677e-06,
+      "loss": 0.1905,
+      "step": 12922
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2140417100881817,
+      "learning_rate": 5.556844777703697e-06,
+      "loss": 0.1703,
+      "step": 12923
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0870153760009866,
+      "learning_rate": 5.5553693675815565e-06,
+      "loss": 0.1776,
+      "step": 12924
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.008402963959758,
+      "learning_rate": 5.55389407802027e-06,
+      "loss": 0.1832,
+      "step": 12925
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.943314946152503,
+      "learning_rate": 5.55241890905986e-06,
+      "loss": 0.1818,
+      "step": 12926
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9804005130777163,
+      "learning_rate": 5.5509438607403355e-06,
+      "loss": 0.1663,
+      "step": 12927
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6517596718905303,
+      "learning_rate": 5.549468933101709e-06,
+      "loss": 0.1787,
+      "step": 12928
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2459658247423453,
+      "learning_rate": 5.547994126183991e-06,
+      "loss": 0.1752,
+      "step": 12929
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9106347033122705,
+      "learning_rate": 5.546519440027186e-06,
+      "loss": 0.1606,
+      "step": 12930
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.641041700811944,
+      "learning_rate": 5.545044874671289e-06,
+      "loss": 0.1646,
+      "step": 12931
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9307740229504946,
+      "learning_rate": 5.543570430156307e-06,
+      "loss": 0.1647,
+      "step": 12932
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8617543708343491,
+      "learning_rate": 5.542096106522224e-06,
+      "loss": 0.1474,
+      "step": 12933
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7869774137012839,
+      "learning_rate": 5.540621903809038e-06,
+      "loss": 0.1718,
+      "step": 12934
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8044176966116101,
+      "learning_rate": 5.539147822056736e-06,
+      "loss": 0.165,
+      "step": 12935
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.885086608903983,
+      "learning_rate": 5.537673861305297e-06,
+      "loss": 0.1713,
+      "step": 12936
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9198934470644922,
+      "learning_rate": 5.536200021594707e-06,
+      "loss": 0.1562,
+      "step": 12937
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.415345934333553,
+      "learning_rate": 5.534726302964944e-06,
+      "loss": 0.1745,
+      "step": 12938
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1173915430681507,
+      "learning_rate": 5.533252705455985e-06,
+      "loss": 0.1775,
+      "step": 12939
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3559267057784843,
+      "learning_rate": 5.531779229107797e-06,
+      "loss": 0.1655,
+      "step": 12940
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8986687901180408,
+      "learning_rate": 5.530305873960351e-06,
+      "loss": 0.1875,
+      "step": 12941
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9091720492697619,
+      "learning_rate": 5.528832640053607e-06,
+      "loss": 0.1675,
+      "step": 12942
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0879286275960187,
+      "learning_rate": 5.527359527427536e-06,
+      "loss": 0.1863,
+      "step": 12943
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1110812546096838,
+      "learning_rate": 5.525886536122085e-06,
+      "loss": 0.1794,
+      "step": 12944
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0303150589230612,
+      "learning_rate": 5.524413666177216e-06,
+      "loss": 0.1943,
+      "step": 12945
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8707871663246685,
+      "learning_rate": 5.522940917632878e-06,
+      "loss": 0.1703,
+      "step": 12946
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8765865925819567,
+      "learning_rate": 5.521468290529023e-06,
+      "loss": 0.1785,
+      "step": 12947
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8522416033424861,
+      "learning_rate": 5.5199957849055905e-06,
+      "loss": 0.1682,
+      "step": 12948
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2651317087868765,
+      "learning_rate": 5.51852340080253e-06,
+      "loss": 0.1779,
+      "step": 12949
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.036138580945707,
+      "learning_rate": 5.517051138259771e-06,
+      "loss": 0.1651,
+      "step": 12950
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5286022891139401,
+      "learning_rate": 5.515578997317257e-06,
+      "loss": 0.1616,
+      "step": 12951
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9511716474931488,
+      "learning_rate": 5.514106978014917e-06,
+      "loss": 0.1965,
+      "step": 12952
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8597682366979487,
+      "learning_rate": 5.512635080392673e-06,
+      "loss": 0.1596,
+      "step": 12953
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0870180016132283,
+      "learning_rate": 5.511163304490456e-06,
+      "loss": 0.1469,
+      "step": 12954
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0480633504288328,
+      "learning_rate": 5.50969165034819e-06,
+      "loss": 0.1846,
+      "step": 12955
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.5381871452442013,
+      "learning_rate": 5.508220118005794e-06,
+      "loss": 0.1666,
+      "step": 12956
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8148043083133922,
+      "learning_rate": 5.5067487075031764e-06,
+      "loss": 0.1735,
+      "step": 12957
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0187217086868456,
+      "learning_rate": 5.505277418880259e-06,
+      "loss": 0.1874,
+      "step": 12958
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4282204424031621,
+      "learning_rate": 5.503806252176941e-06,
+      "loss": 0.169,
+      "step": 12959
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9937963162216736,
+      "learning_rate": 5.502335207433136e-06,
+      "loss": 0.1993,
+      "step": 12960
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.930548165625113,
+      "learning_rate": 5.500864284688739e-06,
+      "loss": 0.1654,
+      "step": 12961
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8048369428436289,
+      "learning_rate": 5.499393483983657e-06,
+      "loss": 0.1646,
+      "step": 12962
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2232545060032005,
+      "learning_rate": 5.497922805357776e-06,
+      "loss": 0.1819,
+      "step": 12963
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8363692909755749,
+      "learning_rate": 5.496452248850994e-06,
+      "loss": 0.153,
+      "step": 12964
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8803241764374603,
+      "learning_rate": 5.494981814503199e-06,
+      "loss": 0.1801,
+      "step": 12965
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.189162147818141,
+      "learning_rate": 5.49351150235428e-06,
+      "loss": 0.181,
+      "step": 12966
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0669972255809883,
+      "learning_rate": 5.492041312444112e-06,
+      "loss": 0.1623,
+      "step": 12967
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7783963693138267,
+      "learning_rate": 5.490571244812582e-06,
+      "loss": 0.1841,
+      "step": 12968
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0361244849682094,
+      "learning_rate": 5.489101299499562e-06,
+      "loss": 0.1925,
+      "step": 12969
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9770182854110955,
+      "learning_rate": 5.487631476544921e-06,
+      "loss": 0.15,
+      "step": 12970
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.600128640473354,
+      "learning_rate": 5.48616177598853e-06,
+      "loss": 0.1938,
+      "step": 12971
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.070133379570775,
+      "learning_rate": 5.484692197870256e-06,
+      "loss": 0.193,
+      "step": 12972
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1204297316426883,
+      "learning_rate": 5.483222742229964e-06,
+      "loss": 0.1566,
+      "step": 12973
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.001429122190217,
+      "learning_rate": 5.4817534091075084e-06,
+      "loss": 0.1688,
+      "step": 12974
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0830367798031313,
+      "learning_rate": 5.480284198542749e-06,
+      "loss": 0.1836,
+      "step": 12975
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0061740479549004,
+      "learning_rate": 5.4788151105755326e-06,
+      "loss": 0.1923,
+      "step": 12976
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.008732350438919,
+      "learning_rate": 5.477346145245717e-06,
+      "loss": 0.1805,
+      "step": 12977
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1567289166163461,
+      "learning_rate": 5.475877302593135e-06,
+      "loss": 0.1672,
+      "step": 12978
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9044833916082834,
+      "learning_rate": 5.4744085826576445e-06,
+      "loss": 0.1732,
+      "step": 12979
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7896497231676506,
+      "learning_rate": 5.472939985479071e-06,
+      "loss": 0.1773,
+      "step": 12980
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1109312907063378,
+      "learning_rate": 5.471471511097257e-06,
+      "loss": 0.1797,
+      "step": 12981
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9273909805501587,
+      "learning_rate": 5.470003159552033e-06,
+      "loss": 0.158,
+      "step": 12982
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9094058333615452,
+      "learning_rate": 5.468534930883234e-06,
+      "loss": 0.1557,
+      "step": 12983
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0902517874089719,
+      "learning_rate": 5.467066825130676e-06,
+      "loss": 0.1632,
+      "step": 12984
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0784363867926698,
+      "learning_rate": 5.465598842334192e-06,
+      "loss": 0.1722,
+      "step": 12985
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0900263158696295,
+      "learning_rate": 5.46413098253359e-06,
+      "loss": 0.168,
+      "step": 12986
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8354048699959953,
+      "learning_rate": 5.462663245768696e-06,
+      "loss": 0.1683,
+      "step": 12987
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8168123012717525,
+      "learning_rate": 5.461195632079317e-06,
+      "loss": 0.1612,
+      "step": 12988
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2364602564754394,
+      "learning_rate": 5.459728141505259e-06,
+      "loss": 0.1733,
+      "step": 12989
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.041956726976596,
+      "learning_rate": 5.458260774086332e-06,
+      "loss": 0.1809,
+      "step": 12990
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8923402773137696,
+      "learning_rate": 5.4567935298623385e-06,
+      "loss": 0.188,
+      "step": 12991
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9256412258996756,
+      "learning_rate": 5.45532640887308e-06,
+      "loss": 0.203,
+      "step": 12992
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9424634298516746,
+      "learning_rate": 5.453859411158347e-06,
+      "loss": 0.1721,
+      "step": 12993
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.748854710324556,
+      "learning_rate": 5.452392536757936e-06,
+      "loss": 0.1591,
+      "step": 12994
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.107341764985314,
+      "learning_rate": 5.450925785711632e-06,
+      "loss": 0.1865,
+      "step": 12995
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.000456094721113,
+      "learning_rate": 5.449459158059226e-06,
+      "loss": 0.1719,
+      "step": 12996
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8533555961879594,
+      "learning_rate": 5.447992653840494e-06,
+      "loss": 0.1715,
+      "step": 12997
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0126316604351642,
+      "learning_rate": 5.4465262730952186e-06,
+      "loss": 0.1931,
+      "step": 12998
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8904401314425945,
+      "learning_rate": 5.445060015863175e-06,
+      "loss": 0.1852,
+      "step": 12999
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.886172356588746,
+      "learning_rate": 5.443593882184139e-06,
+      "loss": 0.1705,
+      "step": 13000
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9153670957628168,
+      "learning_rate": 5.442127872097873e-06,
+      "loss": 0.1782,
+      "step": 13001
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2129372270740568,
+      "learning_rate": 5.440661985644149e-06,
+      "loss": 0.165,
+      "step": 13002
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8697255131527631,
+      "learning_rate": 5.439196222862724e-06,
+      "loss": 0.1648,
+      "step": 13003
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1582811645269684,
+      "learning_rate": 5.437730583793362e-06,
+      "loss": 0.1658,
+      "step": 13004
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3861422946046051,
+      "learning_rate": 5.436265068475815e-06,
+      "loss": 0.1846,
+      "step": 13005
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7321198610302062,
+      "learning_rate": 5.4347996769498315e-06,
+      "loss": 0.1696,
+      "step": 13006
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9401040464943988,
+      "learning_rate": 5.433334409255165e-06,
+      "loss": 0.1754,
+      "step": 13007
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.078890431543232,
+      "learning_rate": 5.431869265431562e-06,
+      "loss": 0.1951,
+      "step": 13008
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0439385899204332,
+      "learning_rate": 5.430404245518766e-06,
+      "loss": 0.1818,
+      "step": 13009
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9609589053298132,
+      "learning_rate": 5.4289393495565076e-06,
+      "loss": 0.1804,
+      "step": 13010
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1166850002485198,
+      "learning_rate": 5.427474577584534e-06,
+      "loss": 0.1579,
+      "step": 13011
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9587875712844113,
+      "learning_rate": 5.426009929642566e-06,
+      "loss": 0.1869,
+      "step": 13012
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0499277335217805,
+      "learning_rate": 5.424545405770341e-06,
+      "loss": 0.1932,
+      "step": 13013
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9020401891335598,
+      "learning_rate": 5.423081006007576e-06,
+      "loss": 0.1855,
+      "step": 13014
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0372741435503523,
+      "learning_rate": 5.421616730394e-06,
+      "loss": 0.1919,
+      "step": 13015
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9728797046954707,
+      "learning_rate": 5.420152578969327e-06,
+      "loss": 0.1659,
+      "step": 13016
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0922675199537346,
+      "learning_rate": 5.4186885517732724e-06,
+      "loss": 0.1897,
+      "step": 13017
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9800919632695411,
+      "learning_rate": 5.417224648845551e-06,
+      "loss": 0.2087,
+      "step": 13018
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.946962467462363,
+      "learning_rate": 5.415760870225873e-06,
+      "loss": 0.1881,
+      "step": 13019
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.648603475752819,
+      "learning_rate": 5.414297215953937e-06,
+      "loss": 0.1628,
+      "step": 13020
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.042906718375941,
+      "learning_rate": 5.41283368606945e-06,
+      "loss": 0.1598,
+      "step": 13021
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1437852574832168,
+      "learning_rate": 5.411370280612109e-06,
+      "loss": 0.1707,
+      "step": 13022
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.109094609556804,
+      "learning_rate": 5.4099069996216055e-06,
+      "loss": 0.1531,
+      "step": 13023
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1181170488361203,
+      "learning_rate": 5.408443843137634e-06,
+      "loss": 0.1735,
+      "step": 13024
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0072555196252104,
+      "learning_rate": 5.406980811199881e-06,
+      "loss": 0.1584,
+      "step": 13025
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9447327044724418,
+      "learning_rate": 5.405517903848039e-06,
+      "loss": 0.1512,
+      "step": 13026
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0406129068479135,
+      "learning_rate": 5.404055121121778e-06,
+      "loss": 0.1772,
+      "step": 13027
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9813397660614563,
+      "learning_rate": 5.402592463060785e-06,
+      "loss": 0.1917,
+      "step": 13028
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9043884637346777,
+      "learning_rate": 5.401129929704727e-06,
+      "loss": 0.1743,
+      "step": 13029
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.7330322694414013,
+      "learning_rate": 5.399667521093285e-06,
+      "loss": 0.1953,
+      "step": 13030
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2142528716335321,
+      "learning_rate": 5.398205237266116e-06,
+      "loss": 0.1697,
+      "step": 13031
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9287096481877403,
+      "learning_rate": 5.396743078262895e-06,
+      "loss": 0.1736,
+      "step": 13032
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2355905776550489,
+      "learning_rate": 5.395281044123273e-06,
+      "loss": 0.1645,
+      "step": 13033
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8630831827981413,
+      "learning_rate": 5.393819134886913e-06,
+      "loss": 0.1682,
+      "step": 13034
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0966432650161853,
+      "learning_rate": 5.392357350593469e-06,
+      "loss": 0.1745,
+      "step": 13035
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8858228134039371,
+      "learning_rate": 5.390895691282596e-06,
+      "loss": 0.1785,
+      "step": 13036
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9077285942791732,
+      "learning_rate": 5.389434156993935e-06,
+      "loss": 0.169,
+      "step": 13037
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9508245618541968,
+      "learning_rate": 5.387972747767136e-06,
+      "loss": 0.1893,
+      "step": 13038
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.127200876715937,
+      "learning_rate": 5.386511463641836e-06,
+      "loss": 0.1729,
+      "step": 13039
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5310594407487175,
+      "learning_rate": 5.38505030465767e-06,
+      "loss": 0.1898,
+      "step": 13040
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.938278714469896,
+      "learning_rate": 5.383589270854279e-06,
+      "loss": 0.1894,
+      "step": 13041
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0370258145919886,
+      "learning_rate": 5.382128362271285e-06,
+      "loss": 0.1794,
+      "step": 13042
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.064224882974358,
+      "learning_rate": 5.380667578948321e-06,
+      "loss": 0.1656,
+      "step": 13043
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9352410513608885,
+      "learning_rate": 5.379206920925009e-06,
+      "loss": 0.1976,
+      "step": 13044
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7886476725295988,
+      "learning_rate": 5.3777463882409744e-06,
+      "loss": 0.1864,
+      "step": 13045
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.079848051901478,
+      "learning_rate": 5.376285980935827e-06,
+      "loss": 0.1585,
+      "step": 13046
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9483693289000759,
+      "learning_rate": 5.374825699049186e-06,
+      "loss": 0.1597,
+      "step": 13047
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.109683010946804,
+      "learning_rate": 5.3733655426206564e-06,
+      "loss": 0.1881,
+      "step": 13048
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9538865416720401,
+      "learning_rate": 5.371905511689852e-06,
+      "loss": 0.1738,
+      "step": 13049
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8736303533856425,
+      "learning_rate": 5.3704456062963674e-06,
+      "loss": 0.1677,
+      "step": 13050
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9141910495008178,
+      "learning_rate": 5.368985826479807e-06,
+      "loss": 0.1821,
+      "step": 13051
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9491642343236588,
+      "learning_rate": 5.367526172279768e-06,
+      "loss": 0.1619,
+      "step": 13052
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1240984076090763,
+      "learning_rate": 5.366066643735847e-06,
+      "loss": 0.1568,
+      "step": 13053
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.389968272416857,
+      "learning_rate": 5.364607240887626e-06,
+      "loss": 0.1866,
+      "step": 13054
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3602336034440015,
+      "learning_rate": 5.3631479637747e-06,
+      "loss": 0.1631,
+      "step": 13055
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0299834130513226,
+      "learning_rate": 5.361688812436642e-06,
+      "loss": 0.1609,
+      "step": 13056
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2674930173145462,
+      "learning_rate": 5.360229786913042e-06,
+      "loss": 0.1657,
+      "step": 13057
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.13060064340207,
+      "learning_rate": 5.3587708872434705e-06,
+      "loss": 0.1746,
+      "step": 13058
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8379752729010114,
+      "learning_rate": 5.357312113467497e-06,
+      "loss": 0.1534,
+      "step": 13059
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9557520113683214,
+      "learning_rate": 5.355853465624695e-06,
+      "loss": 0.1711,
+      "step": 13060
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9239994296030585,
+      "learning_rate": 5.354394943754631e-06,
+      "loss": 0.1737,
+      "step": 13061
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.321773416829171,
+      "learning_rate": 5.352936547896868e-06,
+      "loss": 0.1484,
+      "step": 13062
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9572197883433509,
+      "learning_rate": 5.351478278090962e-06,
+      "loss": 0.1839,
+      "step": 13063
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5778148636154194,
+      "learning_rate": 5.350020134376472e-06,
+      "loss": 0.1724,
+      "step": 13064
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9079286587968833,
+      "learning_rate": 5.348562116792946e-06,
+      "loss": 0.1738,
+      "step": 13065
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9419826033561322,
+      "learning_rate": 5.34710422537994e-06,
+      "loss": 0.1763,
+      "step": 13066
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.116714619089233,
+      "learning_rate": 5.345646460176989e-06,
+      "loss": 0.1757,
+      "step": 13067
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1682603017157274,
+      "learning_rate": 5.344188821223642e-06,
+      "loss": 0.1545,
+      "step": 13068
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9448118760115939,
+      "learning_rate": 5.342731308559435e-06,
+      "loss": 0.163,
+      "step": 13069
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2298396046807265,
+      "learning_rate": 5.341273922223908e-06,
+      "loss": 0.1825,
+      "step": 13070
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1640638526000524,
+      "learning_rate": 5.339816662256587e-06,
+      "loss": 0.1556,
+      "step": 13071
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1082203563147006,
+      "learning_rate": 5.338359528697005e-06,
+      "loss": 0.1632,
+      "step": 13072
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2232992504838813,
+      "learning_rate": 5.3369025215846796e-06,
+      "loss": 0.1789,
+      "step": 13073
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9051337295054136,
+      "learning_rate": 5.3354456409591405e-06,
+      "loss": 0.1649,
+      "step": 13074
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9972396031450846,
+      "learning_rate": 5.333988886859903e-06,
+      "loss": 0.1574,
+      "step": 13075
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7576046547142246,
+      "learning_rate": 5.332532259326476e-06,
+      "loss": 0.1912,
+      "step": 13076
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9144689190373975,
+      "learning_rate": 5.331075758398375e-06,
+      "loss": 0.1796,
+      "step": 13077
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7592614649841753,
+      "learning_rate": 5.329619384115108e-06,
+      "loss": 0.1616,
+      "step": 13078
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0801539664317343,
+      "learning_rate": 5.328163136516184e-06,
+      "loss": 0.1411,
+      "step": 13079
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0975411316744548,
+      "learning_rate": 5.326707015641093e-06,
+      "loss": 0.1754,
+      "step": 13080
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.636575145510722,
+      "learning_rate": 5.325251021529343e-06,
+      "loss": 0.1778,
+      "step": 13081
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4382356623470685,
+      "learning_rate": 5.323795154220419e-06,
+      "loss": 0.1736,
+      "step": 13082
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6580238322303211,
+      "learning_rate": 5.322339413753819e-06,
+      "loss": 0.1732,
+      "step": 13083
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0080504503860495,
+      "learning_rate": 5.3208838001690236e-06,
+      "loss": 0.1907,
+      "step": 13084
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7317639621699024,
+      "learning_rate": 5.319428313505523e-06,
+      "loss": 0.1658,
+      "step": 13085
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4234701788945343,
+      "learning_rate": 5.317972953802789e-06,
+      "loss": 0.205,
+      "step": 13086
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8881240899721385,
+      "learning_rate": 5.316517721100304e-06,
+      "loss": 0.1756,
+      "step": 13087
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0114148632119997,
+      "learning_rate": 5.31506261543754e-06,
+      "loss": 0.1618,
+      "step": 13088
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2493343913754495,
+      "learning_rate": 5.3136076368539706e-06,
+      "loss": 0.1857,
+      "step": 13089
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.011576010409811,
+      "learning_rate": 5.312152785389056e-06,
+      "loss": 0.1718,
+      "step": 13090
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1400992672593633,
+      "learning_rate": 5.310698061082264e-06,
+      "loss": 0.187,
+      "step": 13091
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8391929239457808,
+      "learning_rate": 5.309243463973054e-06,
+      "loss": 0.1662,
+      "step": 13092
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0521217755250742,
+      "learning_rate": 5.307788994100876e-06,
+      "loss": 0.189,
+      "step": 13093
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8679850090640902,
+      "learning_rate": 5.306334651505185e-06,
+      "loss": 0.1759,
+      "step": 13094
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5800351297499289,
+      "learning_rate": 5.304880436225432e-06,
+      "loss": 0.2178,
+      "step": 13095
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1295677367762624,
+      "learning_rate": 5.303426348301066e-06,
+      "loss": 0.157,
+      "step": 13096
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9239596443120076,
+      "learning_rate": 5.3019723877715235e-06,
+      "loss": 0.1811,
+      "step": 13097
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9849840091548583,
+      "learning_rate": 5.300518554676247e-06,
+      "loss": 0.1619,
+      "step": 13098
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9109699828134366,
+      "learning_rate": 5.299064849054667e-06,
+      "loss": 0.1476,
+      "step": 13099
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3790099251749453,
+      "learning_rate": 5.297611270946223e-06,
+      "loss": 0.1761,
+      "step": 13100
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.226670570464689,
+      "learning_rate": 5.296157820390335e-06,
+      "loss": 0.1539,
+      "step": 13101
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7468050795649048,
+      "learning_rate": 5.294704497426435e-06,
+      "loss": 0.1555,
+      "step": 13102
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9152207253205684,
+      "learning_rate": 5.293251302093938e-06,
+      "loss": 0.1565,
+      "step": 13103
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1090519116839588,
+      "learning_rate": 5.291798234432264e-06,
+      "loss": 0.1788,
+      "step": 13104
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0360390926658547,
+      "learning_rate": 5.2903452944808294e-06,
+      "loss": 0.2028,
+      "step": 13105
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7909596590664957,
+      "learning_rate": 5.28889248227905e-06,
+      "loss": 0.1699,
+      "step": 13106
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.618341463656754,
+      "learning_rate": 5.287439797866323e-06,
+      "loss": 0.1906,
+      "step": 13107
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9067444923206541,
+      "learning_rate": 5.2859872412820625e-06,
+      "loss": 0.1802,
+      "step": 13108
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4808531136665637,
+      "learning_rate": 5.284534812565663e-06,
+      "loss": 0.1836,
+      "step": 13109
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8096508917230821,
+      "learning_rate": 5.283082511756519e-06,
+      "loss": 0.173,
+      "step": 13110
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9257012351305746,
+      "learning_rate": 5.281630338894032e-06,
+      "loss": 0.1756,
+      "step": 13111
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.064130585477865,
+      "learning_rate": 5.280178294017586e-06,
+      "loss": 0.1563,
+      "step": 13112
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3097633680081981,
+      "learning_rate": 5.27872637716657e-06,
+      "loss": 0.1749,
+      "step": 13113
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4768187093469949,
+      "learning_rate": 5.277274588380368e-06,
+      "loss": 0.1736,
+      "step": 13114
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9706272896638604,
+      "learning_rate": 5.275822927698362e-06,
+      "loss": 0.2017,
+      "step": 13115
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6035603029810908,
+      "learning_rate": 5.274371395159923e-06,
+      "loss": 0.1456,
+      "step": 13116
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7868879807142999,
+      "learning_rate": 5.27291999080443e-06,
+      "loss": 0.172,
+      "step": 13117
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.16407422123123,
+      "learning_rate": 5.271468714671247e-06,
+      "loss": 0.1899,
+      "step": 13118
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.121016751150056,
+      "learning_rate": 5.2700175667997456e-06,
+      "loss": 0.1558,
+      "step": 13119
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0911689136172678,
+      "learning_rate": 5.2685665472292805e-06,
+      "loss": 0.1773,
+      "step": 13120
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7623012469449829,
+      "learning_rate": 5.267115655999214e-06,
+      "loss": 0.1694,
+      "step": 13121
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2615801018605743,
+      "learning_rate": 5.265664893148904e-06,
+      "loss": 0.1792,
+      "step": 13122
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2063061781367201,
+      "learning_rate": 5.264214258717705e-06,
+      "loss": 0.1627,
+      "step": 13123
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.36952437005172,
+      "learning_rate": 5.2627637527449575e-06,
+      "loss": 0.1762,
+      "step": 13124
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9686427549608618,
+      "learning_rate": 5.2613133752700145e-06,
+      "loss": 0.1918,
+      "step": 13125
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.341735433036232,
+      "learning_rate": 5.2598631263322145e-06,
+      "loss": 0.1809,
+      "step": 13126
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3729701286309794,
+      "learning_rate": 5.25841300597089e-06,
+      "loss": 0.1642,
+      "step": 13127
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9975682349986622,
+      "learning_rate": 5.256963014225385e-06,
+      "loss": 0.1864,
+      "step": 13128
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4672529989764542,
+      "learning_rate": 5.255513151135022e-06,
+      "loss": 0.2066,
+      "step": 13129
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7928599134134791,
+      "learning_rate": 5.2540634167391325e-06,
+      "loss": 0.195,
+      "step": 13130
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0659403602991226,
+      "learning_rate": 5.252613811077042e-06,
+      "loss": 0.1718,
+      "step": 13131
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.20100561882506,
+      "learning_rate": 5.251164334188073e-06,
+      "loss": 0.1737,
+      "step": 13132
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.405450961930424,
+      "learning_rate": 5.249714986111536e-06,
+      "loss": 0.1619,
+      "step": 13133
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3184257732901865,
+      "learning_rate": 5.248265766886752e-06,
+      "loss": 0.1524,
+      "step": 13134
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.032123083377175,
+      "learning_rate": 5.246816676553024e-06,
+      "loss": 0.1774,
+      "step": 13135
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3472235154009833,
+      "learning_rate": 5.245367715149665e-06,
+      "loss": 0.1861,
+      "step": 13136
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.587312343510043,
+      "learning_rate": 5.243918882715973e-06,
+      "loss": 0.1769,
+      "step": 13137
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1996450448267364,
+      "learning_rate": 5.242470179291253e-06,
+      "loss": 0.1543,
+      "step": 13138
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9134390388709754,
+      "learning_rate": 5.241021604914793e-06,
+      "loss": 0.178,
+      "step": 13139
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.897598360509561,
+      "learning_rate": 5.2395731596258925e-06,
+      "loss": 0.1726,
+      "step": 13140
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2008256831395474,
+      "learning_rate": 5.238124843463839e-06,
+      "loss": 0.1792,
+      "step": 13141
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.025255671697507,
+      "learning_rate": 5.236676656467921e-06,
+      "loss": 0.1699,
+      "step": 13142
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1622225323396218,
+      "learning_rate": 5.235228598677413e-06,
+      "loss": 0.1631,
+      "step": 13143
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9028964025869091,
+      "learning_rate": 5.233780670131603e-06,
+      "loss": 0.1655,
+      "step": 13144
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.947513699797315,
+      "learning_rate": 5.232332870869763e-06,
+      "loss": 0.1798,
+      "step": 13145
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9029104736554413,
+      "learning_rate": 5.2308852009311576e-06,
+      "loss": 0.1928,
+      "step": 13146
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.686357294204009,
+      "learning_rate": 5.229437660355061e-06,
+      "loss": 0.1809,
+      "step": 13147
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9317309818471727,
+      "learning_rate": 5.227990249180737e-06,
+      "loss": 0.1723,
+      "step": 13148
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0209870204447664,
+      "learning_rate": 5.226542967447452e-06,
+      "loss": 0.1816,
+      "step": 13149
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.167574722120494,
+      "learning_rate": 5.225095815194453e-06,
+      "loss": 0.1733,
+      "step": 13150
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9041556488319442,
+      "learning_rate": 5.223648792461005e-06,
+      "loss": 0.1856,
+      "step": 13151
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0624582657746249,
+      "learning_rate": 5.222201899286349e-06,
+      "loss": 0.1697,
+      "step": 13152
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0023796482199938,
+      "learning_rate": 5.2207551357097395e-06,
+      "loss": 0.1599,
+      "step": 13153
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8018090044150327,
+      "learning_rate": 5.219308501770415e-06,
+      "loss": 0.1589,
+      "step": 13154
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6515551581906176,
+      "learning_rate": 5.217861997507618e-06,
+      "loss": 0.192,
+      "step": 13155
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8821744422221828,
+      "learning_rate": 5.2164156229605835e-06,
+      "loss": 0.1587,
+      "step": 13156
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 26.084781713598552,
+      "learning_rate": 5.214969378168544e-06,
+      "loss": 0.1689,
+      "step": 13157
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2620184943781705,
+      "learning_rate": 5.213523263170731e-06,
+      "loss": 0.1788,
+      "step": 13158
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9441269351658541,
+      "learning_rate": 5.2120772780063735e-06,
+      "loss": 0.1803,
+      "step": 13159
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8263735589892514,
+      "learning_rate": 5.210631422714686e-06,
+      "loss": 0.177,
+      "step": 13160
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8343292258524222,
+      "learning_rate": 5.209185697334895e-06,
+      "loss": 0.1935,
+      "step": 13161
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7870435131244253,
+      "learning_rate": 5.207740101906215e-06,
+      "loss": 0.1649,
+      "step": 13162
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8376019067077287,
+      "learning_rate": 5.20629463646785e-06,
+      "loss": 0.1656,
+      "step": 13163
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1280017623926835,
+      "learning_rate": 5.2048493010590125e-06,
+      "loss": 0.1756,
+      "step": 13164
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1939937173125184,
+      "learning_rate": 5.20340409571891e-06,
+      "loss": 0.1779,
+      "step": 13165
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0052027725381725,
+      "learning_rate": 5.201959020486746e-06,
+      "loss": 0.1924,
+      "step": 13166
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.968058426905328,
+      "learning_rate": 5.20051407540171e-06,
+      "loss": 0.1891,
+      "step": 13167
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9770179576061162,
+      "learning_rate": 5.199069260503006e-06,
+      "loss": 0.1449,
+      "step": 13168
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2173618399010426,
+      "learning_rate": 5.197624575829815e-06,
+      "loss": 0.2029,
+      "step": 13169
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.1799476798417223,
+      "learning_rate": 5.196180021421332e-06,
+      "loss": 0.1988,
+      "step": 13170
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5495054811517865,
+      "learning_rate": 5.194735597316733e-06,
+      "loss": 0.1916,
+      "step": 13171
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9055829113764653,
+      "learning_rate": 5.193291303555208e-06,
+      "loss": 0.1645,
+      "step": 13172
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.360427470770462,
+      "learning_rate": 5.191847140175923e-06,
+      "loss": 0.1538,
+      "step": 13173
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2417065582611897,
+      "learning_rate": 5.190403107218056e-06,
+      "loss": 0.1882,
+      "step": 13174
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0556889798216127,
+      "learning_rate": 5.188959204720776e-06,
+      "loss": 0.1933,
+      "step": 13175
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.736411415481473,
+      "learning_rate": 5.1875154327232534e-06,
+      "loss": 0.1805,
+      "step": 13176
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9560392244403962,
+      "learning_rate": 5.186071791264642e-06,
+      "loss": 0.1685,
+      "step": 13177
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.854930259475486,
+      "learning_rate": 5.1846282803841095e-06,
+      "loss": 0.1833,
+      "step": 13178
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8505934212998812,
+      "learning_rate": 5.183184900120807e-06,
+      "loss": 0.1485,
+      "step": 13179
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1737667148898814,
+      "learning_rate": 5.181741650513883e-06,
+      "loss": 0.1629,
+      "step": 13180
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2484661380408568,
+      "learning_rate": 5.180298531602491e-06,
+      "loss": 0.1718,
+      "step": 13181
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2998613299876511,
+      "learning_rate": 5.178855543425771e-06,
+      "loss": 0.1617,
+      "step": 13182
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9322637885141669,
+      "learning_rate": 5.177412686022866e-06,
+      "loss": 0.185,
+      "step": 13183
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.215022711148766,
+      "learning_rate": 5.1759699594329135e-06,
+      "loss": 0.1681,
+      "step": 13184
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4889001656465048,
+      "learning_rate": 5.1745273636950545e-06,
+      "loss": 0.1709,
+      "step": 13185
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0293914136113915,
+      "learning_rate": 5.173084898848408e-06,
+      "loss": 0.1693,
+      "step": 13186
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6068083591065103,
+      "learning_rate": 5.17164256493211e-06,
+      "loss": 0.1796,
+      "step": 13187
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0840075598306749,
+      "learning_rate": 5.170200361985277e-06,
+      "loss": 0.1684,
+      "step": 13188
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0189090314842295,
+      "learning_rate": 5.168758290047035e-06,
+      "loss": 0.1501,
+      "step": 13189
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9682687528749686,
+      "learning_rate": 5.167316349156495e-06,
+      "loss": 0.1988,
+      "step": 13190
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0825631032236098,
+      "learning_rate": 5.16587453935277e-06,
+      "loss": 0.1704,
+      "step": 13191
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2869878171587137,
+      "learning_rate": 5.164432860674972e-06,
+      "loss": 0.1723,
+      "step": 13192
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1248074085681445,
+      "learning_rate": 5.162991313162209e-06,
+      "loss": 0.1684,
+      "step": 13193
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7456815987690477,
+      "learning_rate": 5.161549896853577e-06,
+      "loss": 0.2443,
+      "step": 13194
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8367021529866165,
+      "learning_rate": 5.16010861178818e-06,
+      "loss": 0.1567,
+      "step": 13195
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1886731283547356,
+      "learning_rate": 5.158667458005111e-06,
+      "loss": 0.1787,
+      "step": 13196
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9412907591588243,
+      "learning_rate": 5.157226435543456e-06,
+      "loss": 0.1637,
+      "step": 13197
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.008260727572345,
+      "learning_rate": 5.155785544442313e-06,
+      "loss": 0.1919,
+      "step": 13198
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4114423893472183,
+      "learning_rate": 5.154344784740757e-06,
+      "loss": 0.1833,
+      "step": 13199
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8491369164972039,
+      "learning_rate": 5.15290415647787e-06,
+      "loss": 0.1701,
+      "step": 13200
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9168263593528301,
+      "learning_rate": 5.1514636596927325e-06,
+      "loss": 0.1723,
+      "step": 13201
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7817434712222572,
+      "learning_rate": 5.150023294424422e-06,
+      "loss": 0.175,
+      "step": 13202
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8371772526988961,
+      "learning_rate": 5.148583060711999e-06,
+      "loss": 0.1654,
+      "step": 13203
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8143341031199058,
+      "learning_rate": 5.147142958594538e-06,
+      "loss": 0.1649,
+      "step": 13204
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6114298088143855,
+      "learning_rate": 5.145702988111095e-06,
+      "loss": 0.1901,
+      "step": 13205
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0781789312399832,
+      "learning_rate": 5.144263149300737e-06,
+      "loss": 0.1656,
+      "step": 13206
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7097433857989437,
+      "learning_rate": 5.142823442202511e-06,
+      "loss": 0.1549,
+      "step": 13207
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7835889847646613,
+      "learning_rate": 5.141383866855476e-06,
+      "loss": 0.191,
+      "step": 13208
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6006004708707962,
+      "learning_rate": 5.139944423298675e-06,
+      "loss": 0.1878,
+      "step": 13209
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.797442171502045,
+      "learning_rate": 5.138505111571157e-06,
+      "loss": 0.1586,
+      "step": 13210
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8320194440956922,
+      "learning_rate": 5.137065931711962e-06,
+      "loss": 0.1753,
+      "step": 13211
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9070023242298215,
+      "learning_rate": 5.135626883760132e-06,
+      "loss": 0.1673,
+      "step": 13212
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9764642294741012,
+      "learning_rate": 5.134187967754694e-06,
+      "loss": 0.1603,
+      "step": 13213
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9232328856993673,
+      "learning_rate": 5.132749183734684e-06,
+      "loss": 0.1797,
+      "step": 13214
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.220088273132286,
+      "learning_rate": 5.131310531739129e-06,
+      "loss": 0.1743,
+      "step": 13215
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1545015554406468,
+      "learning_rate": 5.129872011807046e-06,
+      "loss": 0.1894,
+      "step": 13216
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.115588263462691,
+      "learning_rate": 5.128433623977461e-06,
+      "loss": 0.1794,
+      "step": 13217
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0539783738675352,
+      "learning_rate": 5.126995368289389e-06,
+      "loss": 0.1848,
+      "step": 13218
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5524460056178087,
+      "learning_rate": 5.125557244781847e-06,
+      "loss": 0.1635,
+      "step": 13219
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1596802547778315,
+      "learning_rate": 5.1241192534938355e-06,
+      "loss": 0.1671,
+      "step": 13220
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0529195047149937,
+      "learning_rate": 5.122681394464368e-06,
+      "loss": 0.1995,
+      "step": 13221
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0346863465392042,
+      "learning_rate": 5.12124366773244e-06,
+      "loss": 0.1613,
+      "step": 13222
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0604730200598846,
+      "learning_rate": 5.119806073337057e-06,
+      "loss": 0.1567,
+      "step": 13223
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8402646199803202,
+      "learning_rate": 5.118368611317205e-06,
+      "loss": 0.1591,
+      "step": 13224
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9687228752658418,
+      "learning_rate": 5.116931281711886e-06,
+      "loss": 0.1733,
+      "step": 13225
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3458085182463966,
+      "learning_rate": 5.115494084560076e-06,
+      "loss": 0.1732,
+      "step": 13226
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8374535462370374,
+      "learning_rate": 5.114057019900764e-06,
+      "loss": 0.1602,
+      "step": 13227
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9107738081284421,
+      "learning_rate": 5.112620087772933e-06,
+      "loss": 0.1593,
+      "step": 13228
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8671955054266972,
+      "learning_rate": 5.111183288215562e-06,
+      "loss": 0.1573,
+      "step": 13229
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3376215180747362,
+      "learning_rate": 5.1097466212676175e-06,
+      "loss": 0.1686,
+      "step": 13230
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8096623233751569,
+      "learning_rate": 5.108310086968075e-06,
+      "loss": 0.1765,
+      "step": 13231
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9195070385921977,
+      "learning_rate": 5.106873685355897e-06,
+      "loss": 0.1723,
+      "step": 13232
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8939499616027785,
+      "learning_rate": 5.105437416470043e-06,
+      "loss": 0.1765,
+      "step": 13233
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8221497410960499,
+      "learning_rate": 5.10400128034948e-06,
+      "loss": 0.1832,
+      "step": 13234
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9561186940936613,
+      "learning_rate": 5.102565277033155e-06,
+      "loss": 0.1513,
+      "step": 13235
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.872333575342079,
+      "learning_rate": 5.101129406560023e-06,
+      "loss": 0.1904,
+      "step": 13236
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6989686795106012,
+      "learning_rate": 5.099693668969033e-06,
+      "loss": 0.1791,
+      "step": 13237
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.59026638608574,
+      "learning_rate": 5.098258064299132e-06,
+      "loss": 0.1643,
+      "step": 13238
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1675096098809408,
+      "learning_rate": 5.096822592589254e-06,
+      "loss": 0.1732,
+      "step": 13239
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0224568286169364,
+      "learning_rate": 5.095387253878346e-06,
+      "loss": 0.2192,
+      "step": 13240
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4791810466091335,
+      "learning_rate": 5.09395204820533e-06,
+      "loss": 0.1682,
+      "step": 13241
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.503518493377402,
+      "learning_rate": 5.092516975609146e-06,
+      "loss": 0.1833,
+      "step": 13242
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9971546165789382,
+      "learning_rate": 5.091082036128712e-06,
+      "loss": 0.173,
+      "step": 13243
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0692436202332236,
+      "learning_rate": 5.0896472298029555e-06,
+      "loss": 0.1666,
+      "step": 13244
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4054718829745738,
+      "learning_rate": 5.088212556670795e-06,
+      "loss": 0.1733,
+      "step": 13245
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8824887512289968,
+      "learning_rate": 5.08677801677115e-06,
+      "loss": 0.1785,
+      "step": 13246
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1406611285731336,
+      "learning_rate": 5.085343610142926e-06,
+      "loss": 0.1765,
+      "step": 13247
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2010164866264619,
+      "learning_rate": 5.083909336825037e-06,
+      "loss": 0.1716,
+      "step": 13248
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1463388525562261,
+      "learning_rate": 5.0824751968563845e-06,
+      "loss": 0.1839,
+      "step": 13249
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8175246259217428,
+      "learning_rate": 5.0810411902758675e-06,
+      "loss": 0.1701,
+      "step": 13250
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3857489048773455,
+      "learning_rate": 5.0796073171223884e-06,
+      "loss": 0.1901,
+      "step": 13251
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0770489342555571,
+      "learning_rate": 5.078173577434836e-06,
+      "loss": 0.156,
+      "step": 13252
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0569178307565479,
+      "learning_rate": 5.076739971252103e-06,
+      "loss": 0.1667,
+      "step": 13253
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8588401381650115,
+      "learning_rate": 5.0753064986130765e-06,
+      "loss": 0.1783,
+      "step": 13254
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0696363407064873,
+      "learning_rate": 5.073873159556643e-06,
+      "loss": 0.1749,
+      "step": 13255
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0143686351956207,
+      "learning_rate": 5.072439954121675e-06,
+      "loss": 0.1777,
+      "step": 13256
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5402276037649723,
+      "learning_rate": 5.0710068823470535e-06,
+      "loss": 0.1656,
+      "step": 13257
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0663757615170488,
+      "learning_rate": 5.069573944271646e-06,
+      "loss": 0.1785,
+      "step": 13258
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0825248239599943,
+      "learning_rate": 5.068141139934328e-06,
+      "loss": 0.1927,
+      "step": 13259
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0449836392755059,
+      "learning_rate": 5.066708469373958e-06,
+      "loss": 0.17,
+      "step": 13260
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2647662251334955,
+      "learning_rate": 5.065275932629401e-06,
+      "loss": 0.1636,
+      "step": 13261
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8529323655944855,
+      "learning_rate": 5.063843529739509e-06,
+      "loss": 0.1719,
+      "step": 13262
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.026661229104386,
+      "learning_rate": 5.062411260743141e-06,
+      "loss": 0.1884,
+      "step": 13263
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2294290210006409,
+      "learning_rate": 5.060979125679147e-06,
+      "loss": 0.1672,
+      "step": 13264
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0393714847360274,
+      "learning_rate": 5.0595471245863745e-06,
+      "loss": 0.1797,
+      "step": 13265
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.209605688386021,
+      "learning_rate": 5.058115257503667e-06,
+      "loss": 0.1614,
+      "step": 13266
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0857832204663802,
+      "learning_rate": 5.056683524469859e-06,
+      "loss": 0.1927,
+      "step": 13267
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9911732545186153,
+      "learning_rate": 5.055251925523792e-06,
+      "loss": 0.1655,
+      "step": 13268
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.624387785149277,
+      "learning_rate": 5.0538204607042925e-06,
+      "loss": 0.1607,
+      "step": 13269
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9864496852717262,
+      "learning_rate": 5.052389130050193e-06,
+      "loss": 0.1615,
+      "step": 13270
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2432216270091936,
+      "learning_rate": 5.050957933600317e-06,
+      "loss": 0.1989,
+      "step": 13271
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0533200560620144,
+      "learning_rate": 5.049526871393491e-06,
+      "loss": 0.171,
+      "step": 13272
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7796060605558551,
+      "learning_rate": 5.048095943468524e-06,
+      "loss": 0.1742,
+      "step": 13273
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9142403208282665,
+      "learning_rate": 5.046665149864238e-06,
+      "loss": 0.1552,
+      "step": 13274
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9049380952582469,
+      "learning_rate": 5.045234490619435e-06,
+      "loss": 0.1755,
+      "step": 13275
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2962081391225797,
+      "learning_rate": 5.043803965772932e-06,
+      "loss": 0.1896,
+      "step": 13276
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.101749651971222,
+      "learning_rate": 5.042373575363522e-06,
+      "loss": 0.1798,
+      "step": 13277
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.951802944183445,
+      "learning_rate": 5.040943319430012e-06,
+      "loss": 0.1939,
+      "step": 13278
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0555114435186326,
+      "learning_rate": 5.03951319801119e-06,
+      "loss": 0.1626,
+      "step": 13279
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.350767178690501,
+      "learning_rate": 5.038083211145854e-06,
+      "loss": 0.1766,
+      "step": 13280
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1201796798396992,
+      "learning_rate": 5.03665335887279e-06,
+      "loss": 0.2076,
+      "step": 13281
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9327089524446213,
+      "learning_rate": 5.035223641230789e-06,
+      "loss": 0.189,
+      "step": 13282
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8261115940786845,
+      "learning_rate": 5.033794058258623e-06,
+      "loss": 0.1659,
+      "step": 13283
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4459582919012035,
+      "learning_rate": 5.0323646099950775e-06,
+      "loss": 0.1576,
+      "step": 13284
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8754760385383721,
+      "learning_rate": 5.030935296478922e-06,
+      "loss": 0.1734,
+      "step": 13285
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0686975505660197,
+      "learning_rate": 5.029506117748924e-06,
+      "loss": 0.1728,
+      "step": 13286
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.282629243662813,
+      "learning_rate": 5.0280770738438535e-06,
+      "loss": 0.1656,
+      "step": 13287
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.013316881874263,
+      "learning_rate": 5.026648164802472e-06,
+      "loss": 0.1848,
+      "step": 13288
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0771000819802043,
+      "learning_rate": 5.025219390663545e-06,
+      "loss": 0.1599,
+      "step": 13289
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3450360397275072,
+      "learning_rate": 5.023790751465818e-06,
+      "loss": 0.1638,
+      "step": 13290
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1537028984901865,
+      "learning_rate": 5.022362247248052e-06,
+      "loss": 0.2023,
+      "step": 13291
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9600143888983569,
+      "learning_rate": 5.020933878048988e-06,
+      "loss": 0.1778,
+      "step": 13292
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.347995113401063,
+      "learning_rate": 5.0195056439073775e-06,
+      "loss": 0.2065,
+      "step": 13293
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8692116195281121,
+      "learning_rate": 5.018077544861954e-06,
+      "loss": 0.1683,
+      "step": 13294
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4525317319667932,
+      "learning_rate": 5.016649580951462e-06,
+      "loss": 0.1714,
+      "step": 13295
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0646484388308648,
+      "learning_rate": 5.015221752214627e-06,
+      "loss": 0.1597,
+      "step": 13296
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0205582628493528,
+      "learning_rate": 5.013794058690185e-06,
+      "loss": 0.1797,
+      "step": 13297
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8514951899742292,
+      "learning_rate": 5.01236650041686e-06,
+      "loss": 0.1743,
+      "step": 13298
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8689347769419848,
+      "learning_rate": 5.010939077433378e-06,
+      "loss": 0.1852,
+      "step": 13299
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9292572474186874,
+      "learning_rate": 5.009511789778454e-06,
+      "loss": 0.1802,
+      "step": 13300
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.692359441421138,
+      "learning_rate": 5.008084637490807e-06,
+      "loss": 0.1784,
+      "step": 13301
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2508345137841275,
+      "learning_rate": 5.006657620609147e-06,
+      "loss": 0.1911,
+      "step": 13302
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9286787172036365,
+      "learning_rate": 5.005230739172175e-06,
+      "loss": 0.1892,
+      "step": 13303
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.867410754178937,
+      "learning_rate": 5.003803993218608e-06,
+      "loss": 0.1789,
+      "step": 13304
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1372677523807884,
+      "learning_rate": 5.002377382787135e-06,
+      "loss": 0.1997,
+      "step": 13305
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8645106706756592,
+      "learning_rate": 5.000950907916457e-06,
+      "loss": 0.1848,
+      "step": 13306
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9028957991781709,
+      "learning_rate": 4.999524568645268e-06,
+      "loss": 0.1791,
+      "step": 13307
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.653303769921554,
+      "learning_rate": 4.998098365012263e-06,
+      "loss": 0.1944,
+      "step": 13308
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.581598586863471,
+      "learning_rate": 4.9966722970561165e-06,
+      "loss": 0.1645,
+      "step": 13309
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0273155328412358,
+      "learning_rate": 4.995246364815522e-06,
+      "loss": 0.1739,
+      "step": 13310
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.233151772861096,
+      "learning_rate": 4.993820568329147e-06,
+      "loss": 0.1814,
+      "step": 13311
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8748578904892358,
+      "learning_rate": 4.992394907635677e-06,
+      "loss": 0.1754,
+      "step": 13312
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3279718652025159,
+      "learning_rate": 4.990969382773773e-06,
+      "loss": 0.1784,
+      "step": 13313
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8481820795080676,
+      "learning_rate": 4.989543993782109e-06,
+      "loss": 0.1647,
+      "step": 13314
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8523082068024254,
+      "learning_rate": 4.9881187406993455e-06,
+      "loss": 0.1654,
+      "step": 13315
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0876009925692094,
+      "learning_rate": 4.98669362356415e-06,
+      "loss": 0.1692,
+      "step": 13316
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2189864873069765,
+      "learning_rate": 4.985268642415167e-06,
+      "loss": 0.2011,
+      "step": 13317
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0037139617465891,
+      "learning_rate": 4.98384379729106e-06,
+      "loss": 0.1682,
+      "step": 13318
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1469562922992695,
+      "learning_rate": 4.982419088230473e-06,
+      "loss": 0.1668,
+      "step": 13319
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8763162550492711,
+      "learning_rate": 4.98099451527205e-06,
+      "loss": 0.188,
+      "step": 13320
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9031762114085561,
+      "learning_rate": 4.9795700784544355e-06,
+      "loss": 0.1561,
+      "step": 13321
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4303154523574806,
+      "learning_rate": 4.978145777816264e-06,
+      "loss": 0.1728,
+      "step": 13322
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.063977785990633,
+      "learning_rate": 4.9767216133961705e-06,
+      "loss": 0.1552,
+      "step": 13323
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1334620235258557,
+      "learning_rate": 4.975297585232788e-06,
+      "loss": 0.1845,
+      "step": 13324
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0121310102918408,
+      "learning_rate": 4.973873693364746e-06,
+      "loss": 0.1838,
+      "step": 13325
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8874720693087358,
+      "learning_rate": 4.972449937830659e-06,
+      "loss": 0.1392,
+      "step": 13326
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8405872199516397,
+      "learning_rate": 4.971026318669156e-06,
+      "loss": 0.2046,
+      "step": 13327
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9140677436739643,
+      "learning_rate": 4.9696028359188444e-06,
+      "loss": 0.1817,
+      "step": 13328
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.122027630302311,
+      "learning_rate": 4.968179489618345e-06,
+      "loss": 0.1768,
+      "step": 13329
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9179887064369858,
+      "learning_rate": 4.966756279806255e-06,
+      "loss": 0.1599,
+      "step": 13330
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2229010455954328,
+      "learning_rate": 4.9653332065211905e-06,
+      "loss": 0.1692,
+      "step": 13331
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9100285797496191,
+      "learning_rate": 4.963910269801743e-06,
+      "loss": 0.1708,
+      "step": 13332
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.108220029993447,
+      "learning_rate": 4.962487469686513e-06,
+      "loss": 0.1887,
+      "step": 13333
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1734441516261298,
+      "learning_rate": 4.961064806214096e-06,
+      "loss": 0.1804,
+      "step": 13334
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.071162594836847,
+      "learning_rate": 4.959642279423085e-06,
+      "loss": 0.1738,
+      "step": 13335
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1985132657468434,
+      "learning_rate": 4.958219889352061e-06,
+      "loss": 0.1652,
+      "step": 13336
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8772492189650118,
+      "learning_rate": 4.956797636039603e-06,
+      "loss": 0.1721,
+      "step": 13337
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2597169048117127,
+      "learning_rate": 4.955375519524299e-06,
+      "loss": 0.2005,
+      "step": 13338
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4108297618878187,
+      "learning_rate": 4.953953539844715e-06,
+      "loss": 0.1738,
+      "step": 13339
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8353302923538877,
+      "learning_rate": 4.952531697039424e-06,
+      "loss": 0.1796,
+      "step": 13340
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.45598031294388,
+      "learning_rate": 4.951109991146999e-06,
+      "loss": 0.1664,
+      "step": 13341
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.890865755028747,
+      "learning_rate": 4.949688422206003e-06,
+      "loss": 0.1665,
+      "step": 13342
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.035408609462023,
+      "learning_rate": 4.9482669902549896e-06,
+      "loss": 0.1791,
+      "step": 13343
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0615178010266604,
+      "learning_rate": 4.946845695332524e-06,
+      "loss": 0.1934,
+      "step": 13344
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1510058838639408,
+      "learning_rate": 4.945424537477149e-06,
+      "loss": 0.1732,
+      "step": 13345
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8626483346119986,
+      "learning_rate": 4.944003516727424e-06,
+      "loss": 0.164,
+      "step": 13346
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9788758357224554,
+      "learning_rate": 4.942582633121885e-06,
+      "loss": 0.1781,
+      "step": 13347
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2100884087094943,
+      "learning_rate": 4.941161886699082e-06,
+      "loss": 0.1607,
+      "step": 13348
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7699738187325713,
+      "learning_rate": 4.939741277497545e-06,
+      "loss": 0.1842,
+      "step": 13349
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.88949621414364,
+      "learning_rate": 4.938320805555811e-06,
+      "loss": 0.1579,
+      "step": 13350
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9083718699578254,
+      "learning_rate": 4.9369004709124115e-06,
+      "loss": 0.1797,
+      "step": 13351
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.122080049815317,
+      "learning_rate": 4.935480273605876e-06,
+      "loss": 0.1666,
+      "step": 13352
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1730433331504413,
+      "learning_rate": 4.93406021367472e-06,
+      "loss": 0.1693,
+      "step": 13353
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1792075254193641,
+      "learning_rate": 4.932640291157471e-06,
+      "loss": 0.1599,
+      "step": 13354
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9867221227698497,
+      "learning_rate": 4.931220506092641e-06,
+      "loss": 0.1657,
+      "step": 13355
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9272440666065497,
+      "learning_rate": 4.929800858518736e-06,
+      "loss": 0.1678,
+      "step": 13356
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3593501320246062,
+      "learning_rate": 4.928381348474274e-06,
+      "loss": 0.1864,
+      "step": 13357
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2641556486628795,
+      "learning_rate": 4.926961975997749e-06,
+      "loss": 0.2181,
+      "step": 13358
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0468224571302156,
+      "learning_rate": 4.925542741127669e-06,
+      "loss": 0.177,
+      "step": 13359
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0279548360084887,
+      "learning_rate": 4.9241236439025275e-06,
+      "loss": 0.1573,
+      "step": 13360
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4210161107545294,
+      "learning_rate": 4.9227046843608224e-06,
+      "loss": 0.1559,
+      "step": 13361
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0368899818961608,
+      "learning_rate": 4.921285862541037e-06,
+      "loss": 0.1776,
+      "step": 13362
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8609128185278291,
+      "learning_rate": 4.919867178481662e-06,
+      "loss": 0.1789,
+      "step": 13363
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0819398376435012,
+      "learning_rate": 4.9184486322211734e-06,
+      "loss": 0.1582,
+      "step": 13364
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.186110252214916,
+      "learning_rate": 4.917030223798057e-06,
+      "loss": 0.1507,
+      "step": 13365
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8508139881204698,
+      "learning_rate": 4.915611953250778e-06,
+      "loss": 0.179,
+      "step": 13366
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8510715693991586,
+      "learning_rate": 4.914193820617813e-06,
+      "loss": 0.1841,
+      "step": 13367
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8557222228620209,
+      "learning_rate": 4.912775825937627e-06,
+      "loss": 0.181,
+      "step": 13368
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0329111875129768,
+      "learning_rate": 4.911357969248688e-06,
+      "loss": 0.1757,
+      "step": 13369
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0033361006799728,
+      "learning_rate": 4.909940250589448e-06,
+      "loss": 0.174,
+      "step": 13370
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9562282908872947,
+      "learning_rate": 4.90852266999837e-06,
+      "loss": 0.1636,
+      "step": 13371
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6260398737326067,
+      "learning_rate": 4.907105227513902e-06,
+      "loss": 0.1776,
+      "step": 13372
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9791958386788515,
+      "learning_rate": 4.905687923174488e-06,
+      "loss": 0.149,
+      "step": 13373
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8983703155733312,
+      "learning_rate": 4.904270757018581e-06,
+      "loss": 0.1563,
+      "step": 13374
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8223754890980843,
+      "learning_rate": 4.902853729084615e-06,
+      "loss": 0.1608,
+      "step": 13375
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8900805982829235,
+      "learning_rate": 4.9014368394110275e-06,
+      "loss": 0.1645,
+      "step": 13376
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.17004263604402,
+      "learning_rate": 4.900020088036254e-06,
+      "loss": 0.157,
+      "step": 13377
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2806837428129272,
+      "learning_rate": 4.898603474998729e-06,
+      "loss": 0.1875,
+      "step": 13378
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.894743260473036,
+      "learning_rate": 4.897187000336867e-06,
+      "loss": 0.1517,
+      "step": 13379
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0252274864395192,
+      "learning_rate": 4.895770664089101e-06,
+      "loss": 0.199,
+      "step": 13380
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1158333518389476,
+      "learning_rate": 4.89435446629384e-06,
+      "loss": 0.1824,
+      "step": 13381
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9805177074634117,
+      "learning_rate": 4.892938406989507e-06,
+      "loss": 0.1725,
+      "step": 13382
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.998938905422329,
+      "learning_rate": 4.891522486214503e-06,
+      "loss": 0.1641,
+      "step": 13383
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9151046573689942,
+      "learning_rate": 4.89010670400724e-06,
+      "loss": 0.1815,
+      "step": 13384
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1280548607462288,
+      "learning_rate": 4.888691060406122e-06,
+      "loss": 0.1789,
+      "step": 13385
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0898656353463203,
+      "learning_rate": 4.887275555449552e-06,
+      "loss": 0.1374,
+      "step": 13386
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0457794371285571,
+      "learning_rate": 4.885860189175917e-06,
+      "loss": 0.1746,
+      "step": 13387
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8664565229805467,
+      "learning_rate": 4.884444961623616e-06,
+      "loss": 0.1848,
+      "step": 13388
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4847731662977808,
+      "learning_rate": 4.8830298728310355e-06,
+      "loss": 0.2005,
+      "step": 13389
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9869762959474485,
+      "learning_rate": 4.881614922836555e-06,
+      "loss": 0.1632,
+      "step": 13390
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1011081086417145,
+      "learning_rate": 4.880200111678563e-06,
+      "loss": 0.1541,
+      "step": 13391
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8733096551389045,
+      "learning_rate": 4.878785439395427e-06,
+      "loss": 0.1746,
+      "step": 13392
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7062110329950042,
+      "learning_rate": 4.8773709060255256e-06,
+      "loss": 0.1809,
+      "step": 13393
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4984101977385287,
+      "learning_rate": 4.8759565116072285e-06,
+      "loss": 0.196,
+      "step": 13394
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9800320483035345,
+      "learning_rate": 4.874542256178903e-06,
+      "loss": 0.1675,
+      "step": 13395
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0454689807754207,
+      "learning_rate": 4.873128139778906e-06,
+      "loss": 0.1897,
+      "step": 13396
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9859535430018859,
+      "learning_rate": 4.8717141624456e-06,
+      "loss": 0.1584,
+      "step": 13397
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0988300467051104,
+      "learning_rate": 4.870300324217334e-06,
+      "loss": 0.1787,
+      "step": 13398
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9376020847074387,
+      "learning_rate": 4.868886625132465e-06,
+      "loss": 0.1755,
+      "step": 13399
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7033458031810654,
+      "learning_rate": 4.867473065229332e-06,
+      "loss": 0.135,
+      "step": 13400
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.063844911772989,
+      "learning_rate": 4.866059644546287e-06,
+      "loss": 0.1879,
+      "step": 13401
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5097869097695407,
+      "learning_rate": 4.864646363121659e-06,
+      "loss": 0.182,
+      "step": 13402
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.369818989617672,
+      "learning_rate": 4.863233220993789e-06,
+      "loss": 0.1657,
+      "step": 13403
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9667709687107531,
+      "learning_rate": 4.861820218201009e-06,
+      "loss": 0.1761,
+      "step": 13404
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1342468133599284,
+      "learning_rate": 4.860407354781647e-06,
+      "loss": 0.1535,
+      "step": 13405
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1973300465476358,
+      "learning_rate": 4.858994630774028e-06,
+      "loss": 0.1678,
+      "step": 13406
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8670715244476761,
+      "learning_rate": 4.857582046216465e-06,
+      "loss": 0.1865,
+      "step": 13407
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9690513997511496,
+      "learning_rate": 4.856169601147285e-06,
+      "loss": 0.1735,
+      "step": 13408
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0865856176368371,
+      "learning_rate": 4.8547572956047894e-06,
+      "loss": 0.1875,
+      "step": 13409
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0067725107937646,
+      "learning_rate": 4.8533451296272934e-06,
+      "loss": 0.1698,
+      "step": 13410
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8611937224662363,
+      "learning_rate": 4.8519331032531015e-06,
+      "loss": 0.1621,
+      "step": 13411
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2179372902706707,
+      "learning_rate": 4.85052121652052e-06,
+      "loss": 0.1737,
+      "step": 13412
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0646786227946774,
+      "learning_rate": 4.849109469467835e-06,
+      "loss": 0.1744,
+      "step": 13413
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3153447789011714,
+      "learning_rate": 4.847697862133351e-06,
+      "loss": 0.1599,
+      "step": 13414
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.139802471031512,
+      "learning_rate": 4.846286394555352e-06,
+      "loss": 0.1461,
+      "step": 13415
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8781096489001784,
+      "learning_rate": 4.844875066772126e-06,
+      "loss": 0.1754,
+      "step": 13416
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9085196484189163,
+      "learning_rate": 4.843463878821955e-06,
+      "loss": 0.1776,
+      "step": 13417
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9693730021548259,
+      "learning_rate": 4.842052830743118e-06,
+      "loss": 0.1638,
+      "step": 13418
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.911581147326447,
+      "learning_rate": 4.840641922573888e-06,
+      "loss": 0.1603,
+      "step": 13419
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1832359251033444,
+      "learning_rate": 4.839231154352535e-06,
+      "loss": 0.181,
+      "step": 13420
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9944004541687458,
+      "learning_rate": 4.837820526117329e-06,
+      "loss": 0.1786,
+      "step": 13421
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9643622262243208,
+      "learning_rate": 4.836410037906537e-06,
+      "loss": 0.1863,
+      "step": 13422
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.798376813456311,
+      "learning_rate": 4.834999689758412e-06,
+      "loss": 0.1683,
+      "step": 13423
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0119099256446815,
+      "learning_rate": 4.833589481711214e-06,
+      "loss": 0.1666,
+      "step": 13424
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1561921433167999,
+      "learning_rate": 4.832179413803193e-06,
+      "loss": 0.1566,
+      "step": 13425
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.135528996676528,
+      "learning_rate": 4.830769486072594e-06,
+      "loss": 0.191,
+      "step": 13426
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0686848952484496,
+      "learning_rate": 4.829359698557669e-06,
+      "loss": 0.201,
+      "step": 13427
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9368366479450847,
+      "learning_rate": 4.827950051296651e-06,
+      "loss": 0.1854,
+      "step": 13428
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.055241925769985,
+      "learning_rate": 4.826540544327778e-06,
+      "loss": 0.1735,
+      "step": 13429
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4929546877284332,
+      "learning_rate": 4.825131177689286e-06,
+      "loss": 0.1832,
+      "step": 13430
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9835157565703031,
+      "learning_rate": 4.8237219514194064e-06,
+      "loss": 0.1614,
+      "step": 13431
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8935105225693067,
+      "learning_rate": 4.8223128655563574e-06,
+      "loss": 0.1619,
+      "step": 13432
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9941601481300318,
+      "learning_rate": 4.820903920138369e-06,
+      "loss": 0.1813,
+      "step": 13433
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2681225658498505,
+      "learning_rate": 4.819495115203651e-06,
+      "loss": 0.1727,
+      "step": 13434
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0083145685072834,
+      "learning_rate": 4.818086450790423e-06,
+      "loss": 0.1838,
+      "step": 13435
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.002803369604112,
+      "learning_rate": 4.816677926936889e-06,
+      "loss": 0.1649,
+      "step": 13436
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8970907695821947,
+      "learning_rate": 4.815269543681259e-06,
+      "loss": 0.1644,
+      "step": 13437
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7536868504864365,
+      "learning_rate": 4.813861301061737e-06,
+      "loss": 0.168,
+      "step": 13438
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.287016327629495,
+      "learning_rate": 4.812453199116522e-06,
+      "loss": 0.1568,
+      "step": 13439
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.365945925124035,
+      "learning_rate": 4.811045237883803e-06,
+      "loss": 0.1613,
+      "step": 13440
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8182433301573657,
+      "learning_rate": 4.80963741740178e-06,
+      "loss": 0.1671,
+      "step": 13441
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2762692421010002,
+      "learning_rate": 4.808229737708635e-06,
+      "loss": 0.183,
+      "step": 13442
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1055784334758063,
+      "learning_rate": 4.806822198842548e-06,
+      "loss": 0.1614,
+      "step": 13443
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6299468631717564,
+      "learning_rate": 4.805414800841706e-06,
+      "loss": 0.182,
+      "step": 13444
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9921407678152764,
+      "learning_rate": 4.804007543744277e-06,
+      "loss": 0.1731,
+      "step": 13445
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.226900393560294,
+      "learning_rate": 4.802600427588437e-06,
+      "loss": 0.1875,
+      "step": 13446
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8771639659433222,
+      "learning_rate": 4.801193452412353e-06,
+      "loss": 0.1828,
+      "step": 13447
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2076103968998635,
+      "learning_rate": 4.799786618254194e-06,
+      "loss": 0.1824,
+      "step": 13448
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.165591143836002,
+      "learning_rate": 4.798379925152113e-06,
+      "loss": 0.1688,
+      "step": 13449
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9737739548462891,
+      "learning_rate": 4.796973373144276e-06,
+      "loss": 0.1849,
+      "step": 13450
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9713615342015935,
+      "learning_rate": 4.795566962268824e-06,
+      "loss": 0.1916,
+      "step": 13451
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.033946714466998,
+      "learning_rate": 4.794160692563917e-06,
+      "loss": 0.2029,
+      "step": 13452
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2749100365224257,
+      "learning_rate": 4.792754564067691e-06,
+      "loss": 0.1633,
+      "step": 13453
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.806561978231977,
+      "learning_rate": 4.791348576818296e-06,
+      "loss": 0.162,
+      "step": 13454
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6382894362878446,
+      "learning_rate": 4.78994273085386e-06,
+      "loss": 0.1802,
+      "step": 13455
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9880680606393522,
+      "learning_rate": 4.788537026212523e-06,
+      "loss": 0.1718,
+      "step": 13456
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3179714810536076,
+      "learning_rate": 4.7871314629324125e-06,
+      "loss": 0.1637,
+      "step": 13457
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8938350389586178,
+      "learning_rate": 4.78572604105166e-06,
+      "loss": 0.1625,
+      "step": 13458
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2134710981285515,
+      "learning_rate": 4.784320760608384e-06,
+      "loss": 0.1706,
+      "step": 13459
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4017759179084095,
+      "learning_rate": 4.782915621640697e-06,
+      "loss": 0.1752,
+      "step": 13460
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1175978308846175,
+      "learning_rate": 4.781510624186723e-06,
+      "loss": 0.1664,
+      "step": 13461
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0035509318285942,
+      "learning_rate": 4.780105768284563e-06,
+      "loss": 0.1715,
+      "step": 13462
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7894408453332361,
+      "learning_rate": 4.778701053972329e-06,
+      "loss": 0.1549,
+      "step": 13463
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.307745294733831,
+      "learning_rate": 4.777296481288125e-06,
+      "loss": 0.1649,
+      "step": 13464
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.050140125564269,
+      "learning_rate": 4.775892050270051e-06,
+      "loss": 0.1903,
+      "step": 13465
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8613949065020686,
+      "learning_rate": 4.774487760956198e-06,
+      "loss": 0.169,
+      "step": 13466
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7835237230319492,
+      "learning_rate": 4.773083613384663e-06,
+      "loss": 0.1403,
+      "step": 13467
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3627385906428595,
+      "learning_rate": 4.771679607593526e-06,
+      "loss": 0.1839,
+      "step": 13468
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0347857730772498,
+      "learning_rate": 4.770275743620879e-06,
+      "loss": 0.183,
+      "step": 13469
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9220173253856068,
+      "learning_rate": 4.768872021504795e-06,
+      "loss": 0.1678,
+      "step": 13470
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.468227027681512,
+      "learning_rate": 4.767468441283355e-06,
+      "loss": 0.1754,
+      "step": 13471
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.953328021389925,
+      "learning_rate": 4.766065002994626e-06,
+      "loss": 0.2112,
+      "step": 13472
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8236174244301655,
+      "learning_rate": 4.764661706676679e-06,
+      "loss": 0.1659,
+      "step": 13473
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.07470258531206,
+      "learning_rate": 4.763258552367579e-06,
+      "loss": 0.159,
+      "step": 13474
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4178247476696861,
+      "learning_rate": 4.761855540105391e-06,
+      "loss": 0.1845,
+      "step": 13475
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0258835616788802,
+      "learning_rate": 4.760452669928167e-06,
+      "loss": 0.1675,
+      "step": 13476
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8569243751405146,
+      "learning_rate": 4.759049941873957e-06,
+      "loss": 0.1865,
+      "step": 13477
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9533173851533268,
+      "learning_rate": 4.757647355980816e-06,
+      "loss": 0.1824,
+      "step": 13478
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.775607340301293,
+      "learning_rate": 4.756244912286782e-06,
+      "loss": 0.1543,
+      "step": 13479
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8975477515315292,
+      "learning_rate": 4.754842610829908e-06,
+      "loss": 0.1811,
+      "step": 13480
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8611261096224713,
+      "learning_rate": 4.753440451648218e-06,
+      "loss": 0.1522,
+      "step": 13481
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2630671689125583,
+      "learning_rate": 4.752038434779752e-06,
+      "loss": 0.1542,
+      "step": 13482
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3134315148743594,
+      "learning_rate": 4.750636560262542e-06,
+      "loss": 0.1737,
+      "step": 13483
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6569673148435342,
+      "learning_rate": 4.749234828134614e-06,
+      "loss": 0.1738,
+      "step": 13484
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.843456888349568,
+      "learning_rate": 4.7478332384339834e-06,
+      "loss": 0.1788,
+      "step": 13485
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1360067662500801,
+      "learning_rate": 4.746431791198678e-06,
+      "loss": 0.1787,
+      "step": 13486
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9441431557652228,
+      "learning_rate": 4.745030486466702e-06,
+      "loss": 0.1614,
+      "step": 13487
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.217989906596319,
+      "learning_rate": 4.743629324276076e-06,
+      "loss": 0.1607,
+      "step": 13488
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2598036556797265,
+      "learning_rate": 4.742228304664797e-06,
+      "loss": 0.1863,
+      "step": 13489
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9377746170865727,
+      "learning_rate": 4.740827427670871e-06,
+      "loss": 0.1719,
+      "step": 13490
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7620754681797829,
+      "learning_rate": 4.7394266933322995e-06,
+      "loss": 0.155,
+      "step": 13491
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.446752341984604,
+      "learning_rate": 4.73802610168708e-06,
+      "loss": 0.1637,
+      "step": 13492
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3227834294404783,
+      "learning_rate": 4.736625652773195e-06,
+      "loss": 0.1758,
+      "step": 13493
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2196491773894043,
+      "learning_rate": 4.735225346628641e-06,
+      "loss": 0.1779,
+      "step": 13494
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.972599455443251,
+      "learning_rate": 4.733825183291396e-06,
+      "loss": 0.1582,
+      "step": 13495
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1493289407215332,
+      "learning_rate": 4.7324251627994375e-06,
+      "loss": 0.1861,
+      "step": 13496
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.213066587740853,
+      "learning_rate": 4.731025285190748e-06,
+      "loss": 0.1705,
+      "step": 13497
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9963341122388324,
+      "learning_rate": 4.729625550503291e-06,
+      "loss": 0.1576,
+      "step": 13498
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.928854673356536,
+      "learning_rate": 4.728225958775038e-06,
+      "loss": 0.1631,
+      "step": 13499
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1516730033725135,
+      "learning_rate": 4.726826510043953e-06,
+      "loss": 0.1566,
+      "step": 13500
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0721723247090889,
+      "learning_rate": 4.725427204348002e-06,
+      "loss": 0.1706,
+      "step": 13501
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0402834885592727,
+      "learning_rate": 4.724028041725132e-06,
+      "loss": 0.1372,
+      "step": 13502
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2585969099620888,
+      "learning_rate": 4.722629022213303e-06,
+      "loss": 0.1784,
+      "step": 13503
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.211687513437,
+      "learning_rate": 4.721230145850456e-06,
+      "loss": 0.1699,
+      "step": 13504
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8774157171124775,
+      "learning_rate": 4.7198314126745424e-06,
+      "loss": 0.1476,
+      "step": 13505
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1104455447501893,
+      "learning_rate": 4.718432822723498e-06,
+      "loss": 0.1737,
+      "step": 13506
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8279150288922937,
+      "learning_rate": 4.7170343760352595e-06,
+      "loss": 0.1626,
+      "step": 13507
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3927560254857045,
+      "learning_rate": 4.715636072647763e-06,
+      "loss": 0.1732,
+      "step": 13508
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9894871996707575,
+      "learning_rate": 4.714237912598941e-06,
+      "loss": 0.1672,
+      "step": 13509
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.001968051076667,
+      "learning_rate": 4.7128398959267095e-06,
+      "loss": 0.1654,
+      "step": 13510
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2122031088329057,
+      "learning_rate": 4.711442022668998e-06,
+      "loss": 0.1674,
+      "step": 13511
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7976776552459861,
+      "learning_rate": 4.710044292863721e-06,
+      "loss": 0.1469,
+      "step": 13512
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9277944585899132,
+      "learning_rate": 4.7086467065487875e-06,
+      "loss": 0.1499,
+      "step": 13513
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2169821076074399,
+      "learning_rate": 4.707249263762115e-06,
+      "loss": 0.1595,
+      "step": 13514
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6496269445057359,
+      "learning_rate": 4.7058519645416004e-06,
+      "loss": 0.1722,
+      "step": 13515
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0799766772357269,
+      "learning_rate": 4.7044548089251505e-06,
+      "loss": 0.161,
+      "step": 13516
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7183745714562393,
+      "learning_rate": 4.703057796950663e-06,
+      "loss": 0.1632,
+      "step": 13517
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2890048452136667,
+      "learning_rate": 4.701660928656036e-06,
+      "loss": 0.1841,
+      "step": 13518
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.299031005709721,
+      "learning_rate": 4.7002642040791526e-06,
+      "loss": 0.1627,
+      "step": 13519
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0199202670304646,
+      "learning_rate": 4.698867623257905e-06,
+      "loss": 0.1509,
+      "step": 13520
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0061548078404354,
+      "learning_rate": 4.697471186230168e-06,
+      "loss": 0.1723,
+      "step": 13521
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6402715115473485,
+      "learning_rate": 4.696074893033828e-06,
+      "loss": 0.1612,
+      "step": 13522
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9338752918297215,
+      "learning_rate": 4.694678743706754e-06,
+      "loss": 0.1874,
+      "step": 13523
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3107696335607468,
+      "learning_rate": 4.693282738286822e-06,
+      "loss": 0.1601,
+      "step": 13524
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0287906491985137,
+      "learning_rate": 4.6918868768118906e-06,
+      "loss": 0.1608,
+      "step": 13525
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3655937539532998,
+      "learning_rate": 4.690491159319829e-06,
+      "loss": 0.1741,
+      "step": 13526
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0816495827850183,
+      "learning_rate": 4.689095585848494e-06,
+      "loss": 0.1555,
+      "step": 13527
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2177066010221815,
+      "learning_rate": 4.687700156435745e-06,
+      "loss": 0.1835,
+      "step": 13528
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8894544378116847,
+      "learning_rate": 4.686304871119429e-06,
+      "loss": 0.1577,
+      "step": 13529
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4856797398142272,
+      "learning_rate": 4.68490972993739e-06,
+      "loss": 0.1922,
+      "step": 13530
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.842186615856397,
+      "learning_rate": 4.683514732927479e-06,
+      "loss": 0.1726,
+      "step": 13531
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0661252290878187,
+      "learning_rate": 4.682119880127526e-06,
+      "loss": 0.181,
+      "step": 13532
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1548888139770868,
+      "learning_rate": 4.680725171575373e-06,
+      "loss": 0.1751,
+      "step": 13533
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9473059849013943,
+      "learning_rate": 4.679330607308849e-06,
+      "loss": 0.195,
+      "step": 13534
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4304758217482247,
+      "learning_rate": 4.677936187365787e-06,
+      "loss": 0.1815,
+      "step": 13535
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9854967453610628,
+      "learning_rate": 4.676541911784004e-06,
+      "loss": 0.1773,
+      "step": 13536
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2275356983406849,
+      "learning_rate": 4.675147780601324e-06,
+      "loss": 0.1621,
+      "step": 13537
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8949481564151959,
+      "learning_rate": 4.673753793855559e-06,
+      "loss": 0.173,
+      "step": 13538
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0624308576100148,
+      "learning_rate": 4.672359951584526e-06,
+      "loss": 0.1774,
+      "step": 13539
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9060232614534275,
+      "learning_rate": 4.670966253826027e-06,
+      "loss": 0.1626,
+      "step": 13540
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8764413985940274,
+      "learning_rate": 4.669572700617872e-06,
+      "loss": 0.1696,
+      "step": 13541
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1743336547666958,
+      "learning_rate": 4.6681792919978565e-06,
+      "loss": 0.1732,
+      "step": 13542
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9143877752093164,
+      "learning_rate": 4.666786028003778e-06,
+      "loss": 0.1569,
+      "step": 13543
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0888796487708368,
+      "learning_rate": 4.66539290867343e-06,
+      "loss": 0.157,
+      "step": 13544
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.25089617021715,
+      "learning_rate": 4.6639999340446045e-06,
+      "loss": 0.1631,
+      "step": 13545
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4660317954025943,
+      "learning_rate": 4.662607104155081e-06,
+      "loss": 0.1864,
+      "step": 13546
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7761331672489337,
+      "learning_rate": 4.661214419042639e-06,
+      "loss": 0.1929,
+      "step": 13547
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0568586982203227,
+      "learning_rate": 4.65982187874506e-06,
+      "loss": 0.1751,
+      "step": 13548
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8537179892331328,
+      "learning_rate": 4.658429483300111e-06,
+      "loss": 0.1648,
+      "step": 13549
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3181985905550566,
+      "learning_rate": 4.6570372327455686e-06,
+      "loss": 0.1788,
+      "step": 13550
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8328073860100116,
+      "learning_rate": 4.6556451271191875e-06,
+      "loss": 0.1745,
+      "step": 13551
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2966089598592383,
+      "learning_rate": 4.6542531664587355e-06,
+      "loss": 0.1703,
+      "step": 13552
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1042830729817357,
+      "learning_rate": 4.652861350801967e-06,
+      "loss": 0.1756,
+      "step": 13553
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8945608077369172,
+      "learning_rate": 4.651469680186641e-06,
+      "loss": 0.1402,
+      "step": 13554
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.672402165651304,
+      "learning_rate": 4.650078154650498e-06,
+      "loss": 0.1837,
+      "step": 13555
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.071037699984741,
+      "learning_rate": 4.648686774231291e-06,
+      "loss": 0.1725,
+      "step": 13556
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8953953526421431,
+      "learning_rate": 4.647295538966754e-06,
+      "loss": 0.1629,
+      "step": 13557
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0473247996790758,
+      "learning_rate": 4.645904448894632e-06,
+      "loss": 0.1598,
+      "step": 13558
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9166293227738351,
+      "learning_rate": 4.644513504052649e-06,
+      "loss": 0.1726,
+      "step": 13559
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8631086924575336,
+      "learning_rate": 4.643122704478541e-06,
+      "loss": 0.1433,
+      "step": 13560
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6481431810406317,
+      "learning_rate": 4.641732050210032e-06,
+      "loss": 0.1535,
+      "step": 13561
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.250059379509954,
+      "learning_rate": 4.640341541284847e-06,
+      "loss": 0.1824,
+      "step": 13562
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8370577808863652,
+      "learning_rate": 4.638951177740697e-06,
+      "loss": 0.1811,
+      "step": 13563
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0688233014826511,
+      "learning_rate": 4.637560959615302e-06,
+      "loss": 0.1542,
+      "step": 13564
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.165288677744048,
+      "learning_rate": 4.636170886946371e-06,
+      "loss": 0.1886,
+      "step": 13565
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9040930500201682,
+      "learning_rate": 4.634780959771602e-06,
+      "loss": 0.1591,
+      "step": 13566
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0498273932508568,
+      "learning_rate": 4.633391178128707e-06,
+      "loss": 0.1724,
+      "step": 13567
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5616589553427944,
+      "learning_rate": 4.632001542055375e-06,
+      "loss": 0.1837,
+      "step": 13568
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1889390923160186,
+      "learning_rate": 4.630612051589305e-06,
+      "loss": 0.1698,
+      "step": 13569
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1318211001505287,
+      "learning_rate": 4.629222706768186e-06,
+      "loss": 0.1799,
+      "step": 13570
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3790972285522614,
+      "learning_rate": 4.6278335076297085e-06,
+      "loss": 0.1619,
+      "step": 13571
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7173679176378776,
+      "learning_rate": 4.626444454211547e-06,
+      "loss": 0.1688,
+      "step": 13572
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1035216925111442,
+      "learning_rate": 4.6250555465513866e-06,
+      "loss": 0.1538,
+      "step": 13573
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9786599408298793,
+      "learning_rate": 4.623666784686895e-06,
+      "loss": 0.1866,
+      "step": 13574
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.421386305458509,
+      "learning_rate": 4.6222781686557485e-06,
+      "loss": 0.1817,
+      "step": 13575
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4270252922303626,
+      "learning_rate": 4.620889698495606e-06,
+      "loss": 0.1885,
+      "step": 13576
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2922180581849296,
+      "learning_rate": 4.619501374244138e-06,
+      "loss": 0.1724,
+      "step": 13577
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8167681017098267,
+      "learning_rate": 4.618113195938997e-06,
+      "loss": 0.1858,
+      "step": 13578
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.004602648431441,
+      "learning_rate": 4.616725163617838e-06,
+      "loss": 0.1783,
+      "step": 13579
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.984136134542311,
+      "learning_rate": 4.615337277318313e-06,
+      "loss": 0.1699,
+      "step": 13580
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8063277723505651,
+      "learning_rate": 4.613949537078074e-06,
+      "loss": 0.1626,
+      "step": 13581
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6604415687659848,
+      "learning_rate": 4.612561942934757e-06,
+      "loss": 0.1848,
+      "step": 13582
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.60353535338271,
+      "learning_rate": 4.611174494925998e-06,
+      "loss": 0.1684,
+      "step": 13583
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5032736790328078,
+      "learning_rate": 4.609787193089438e-06,
+      "loss": 0.1871,
+      "step": 13584
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9766809056291623,
+      "learning_rate": 4.608400037462702e-06,
+      "loss": 0.1659,
+      "step": 13585
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8504763258567766,
+      "learning_rate": 4.607013028083419e-06,
+      "loss": 0.1633,
+      "step": 13586
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9512887618479131,
+      "learning_rate": 4.605626164989212e-06,
+      "loss": 0.1586,
+      "step": 13587
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9440239043274677,
+      "learning_rate": 4.604239448217704e-06,
+      "loss": 0.1931,
+      "step": 13588
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.94407399820329,
+      "learning_rate": 4.602852877806502e-06,
+      "loss": 0.1728,
+      "step": 13589
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8781022672939047,
+      "learning_rate": 4.601466453793224e-06,
+      "loss": 0.16,
+      "step": 13590
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.02587555658374,
+      "learning_rate": 4.60008017621547e-06,
+      "loss": 0.1712,
+      "step": 13591
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8299062732482788,
+      "learning_rate": 4.598694045110851e-06,
+      "loss": 0.1637,
+      "step": 13592
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6775253264414962,
+      "learning_rate": 4.597308060516956e-06,
+      "loss": 0.178,
+      "step": 13593
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0173767614248344,
+      "learning_rate": 4.595922222471388e-06,
+      "loss": 0.1702,
+      "step": 13594
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.15134970328035,
+      "learning_rate": 4.5945365310117325e-06,
+      "loss": 0.2012,
+      "step": 13595
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2736440643117173,
+      "learning_rate": 4.593150986175578e-06,
+      "loss": 0.1611,
+      "step": 13596
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9160802750515318,
+      "learning_rate": 4.59176558800051e-06,
+      "loss": 0.1854,
+      "step": 13597
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0618565771701247,
+      "learning_rate": 4.590380336524108e-06,
+      "loss": 0.1863,
+      "step": 13598
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4095872357508041,
+      "learning_rate": 4.588995231783946e-06,
+      "loss": 0.1574,
+      "step": 13599
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4633830074142442,
+      "learning_rate": 4.5876102738175895e-06,
+      "loss": 0.1711,
+      "step": 13600
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1496702197396802,
+      "learning_rate": 4.586225462662615e-06,
+      "loss": 0.1694,
+      "step": 13601
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2995727514266986,
+      "learning_rate": 4.584840798356574e-06,
+      "loss": 0.1774,
+      "step": 13602
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2066523787933985,
+      "learning_rate": 4.583456280937035e-06,
+      "loss": 0.178,
+      "step": 13603
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0492721475579148,
+      "learning_rate": 4.582071910441549e-06,
+      "loss": 0.159,
+      "step": 13604
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9537120737237095,
+      "learning_rate": 4.580687686907673e-06,
+      "loss": 0.1438,
+      "step": 13605
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1519038592765245,
+      "learning_rate": 4.579303610372945e-06,
+      "loss": 0.165,
+      "step": 13606
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9498047286022816,
+      "learning_rate": 4.577919680874917e-06,
+      "loss": 0.1709,
+      "step": 13607
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.250374359466655,
+      "learning_rate": 4.57653589845112e-06,
+      "loss": 0.1443,
+      "step": 13608
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.918021914540636,
+      "learning_rate": 4.575152263139096e-06,
+      "loss": 0.1468,
+      "step": 13609
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.416579009119526,
+      "learning_rate": 4.573768774976371e-06,
+      "loss": 0.1792,
+      "step": 13610
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2171797058734255,
+      "learning_rate": 4.572385434000477e-06,
+      "loss": 0.1675,
+      "step": 13611
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3487422467380303,
+      "learning_rate": 4.5710022402489316e-06,
+      "loss": 0.1774,
+      "step": 13612
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8390122329378098,
+      "learning_rate": 4.569619193759257e-06,
+      "loss": 0.1526,
+      "step": 13613
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0920505118260386,
+      "learning_rate": 4.5682362945689684e-06,
+      "loss": 0.1821,
+      "step": 13614
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2520630500194392,
+      "learning_rate": 4.5668535427155816e-06,
+      "loss": 0.1818,
+      "step": 13615
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9270273436084522,
+      "learning_rate": 4.565470938236598e-06,
+      "loss": 0.1543,
+      "step": 13616
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0954459642480405,
+      "learning_rate": 4.5640884811695185e-06,
+      "loss": 0.1644,
+      "step": 13617
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.939343554571579,
+      "learning_rate": 4.5627061715518504e-06,
+      "loss": 0.1832,
+      "step": 13618
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9068057167860933,
+      "learning_rate": 4.561324009421081e-06,
+      "loss": 0.1756,
+      "step": 13619
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.21653027952336,
+      "learning_rate": 4.5599419948147075e-06,
+      "loss": 0.1592,
+      "step": 13620
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9564302335647423,
+      "learning_rate": 4.558560127770212e-06,
+      "loss": 0.1713,
+      "step": 13621
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8784787604845598,
+      "learning_rate": 4.55717840832508e-06,
+      "loss": 0.155,
+      "step": 13622
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8785266321659584,
+      "learning_rate": 4.5557968365167905e-06,
+      "loss": 0.1812,
+      "step": 13623
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9026291345633957,
+      "learning_rate": 4.5544154123828246e-06,
+      "loss": 0.1774,
+      "step": 13624
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1734358448640543,
+      "learning_rate": 4.553034135960643e-06,
+      "loss": 0.1724,
+      "step": 13625
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3709598780484173,
+      "learning_rate": 4.551653007287722e-06,
+      "loss": 0.155,
+      "step": 13626
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0501572052064885,
+      "learning_rate": 4.550272026401518e-06,
+      "loss": 0.1617,
+      "step": 13627
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.106664806576892,
+      "learning_rate": 4.548891193339496e-06,
+      "loss": 0.1843,
+      "step": 13628
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4397585434339168,
+      "learning_rate": 4.5475105081391045e-06,
+      "loss": 0.1819,
+      "step": 13629
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.77510083667638,
+      "learning_rate": 4.546129970837799e-06,
+      "loss": 0.1791,
+      "step": 13630
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8667437504260953,
+      "learning_rate": 4.544749581473026e-06,
+      "loss": 0.1788,
+      "step": 13631
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0402234284927807,
+      "learning_rate": 4.543369340082232e-06,
+      "loss": 0.1657,
+      "step": 13632
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2141343623878411,
+      "learning_rate": 4.54198924670285e-06,
+      "loss": 0.1783,
+      "step": 13633
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9137213467530735,
+      "learning_rate": 4.540609301372321e-06,
+      "loss": 0.1905,
+      "step": 13634
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0482645074998238,
+      "learning_rate": 4.539229504128073e-06,
+      "loss": 0.1715,
+      "step": 13635
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3001918628637983,
+      "learning_rate": 4.53784985500753e-06,
+      "loss": 0.1828,
+      "step": 13636
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4116670746126074,
+      "learning_rate": 4.536470354048121e-06,
+      "loss": 0.1869,
+      "step": 13637
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8045872138382979,
+      "learning_rate": 4.535091001287259e-06,
+      "loss": 0.1484,
+      "step": 13638
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0206620077990305,
+      "learning_rate": 4.533711796762362e-06,
+      "loss": 0.1993,
+      "step": 13639
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4379884999398949,
+      "learning_rate": 4.532332740510842e-06,
+      "loss": 0.1645,
+      "step": 13640
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6563946729632992,
+      "learning_rate": 4.530953832570109e-06,
+      "loss": 0.1724,
+      "step": 13641
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1491454430323254,
+      "learning_rate": 4.5295750729775565e-06,
+      "loss": 0.1714,
+      "step": 13642
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.805474854958942,
+      "learning_rate": 4.528196461770596e-06,
+      "loss": 0.1534,
+      "step": 13643
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9306763153096663,
+      "learning_rate": 4.526817998986609e-06,
+      "loss": 0.1653,
+      "step": 13644
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0826607768088683,
+      "learning_rate": 4.5254396846629975e-06,
+      "loss": 0.1562,
+      "step": 13645
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0587769923842847,
+      "learning_rate": 4.5240615188371404e-06,
+      "loss": 0.1681,
+      "step": 13646
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0287482429144468,
+      "learning_rate": 4.522683501546428e-06,
+      "loss": 0.1685,
+      "step": 13647
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9475189022165657,
+      "learning_rate": 4.52130563282823e-06,
+      "loss": 0.1533,
+      "step": 13648
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9559323826822024,
+      "learning_rate": 4.519927912719927e-06,
+      "loss": 0.2016,
+      "step": 13649
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9539906601319694,
+      "learning_rate": 4.518550341258888e-06,
+      "loss": 0.2025,
+      "step": 13650
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7717629165163703,
+      "learning_rate": 4.517172918482485e-06,
+      "loss": 0.1606,
+      "step": 13651
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5207736020799518,
+      "learning_rate": 4.515795644428076e-06,
+      "loss": 0.191,
+      "step": 13652
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.051008527005148,
+      "learning_rate": 4.514418519133017e-06,
+      "loss": 0.1737,
+      "step": 13653
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3983662117819151,
+      "learning_rate": 4.513041542634668e-06,
+      "loss": 0.1852,
+      "step": 13654
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1120072938260044,
+      "learning_rate": 4.511664714970374e-06,
+      "loss": 0.2028,
+      "step": 13655
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0610598459717966,
+      "learning_rate": 4.510288036177485e-06,
+      "loss": 0.1831,
+      "step": 13656
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2201168795069803,
+      "learning_rate": 4.508911506293343e-06,
+      "loss": 0.1497,
+      "step": 13657
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.019804582026071,
+      "learning_rate": 4.50753512535529e-06,
+      "loss": 0.1689,
+      "step": 13658
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5116796780430295,
+      "learning_rate": 4.5061588934006525e-06,
+      "loss": 0.169,
+      "step": 13659
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9856392022738341,
+      "learning_rate": 4.50478281046677e-06,
+      "loss": 0.152,
+      "step": 13660
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6570900891806821,
+      "learning_rate": 4.50340687659096e-06,
+      "loss": 0.1617,
+      "step": 13661
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.283983702559894,
+      "learning_rate": 4.502031091810553e-06,
+      "loss": 0.1772,
+      "step": 13662
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8647829813983922,
+      "learning_rate": 4.500655456162859e-06,
+      "loss": 0.1605,
+      "step": 13663
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8856314954674944,
+      "learning_rate": 4.4992799696852e-06,
+      "loss": 0.1602,
+      "step": 13664
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8324634660473998,
+      "learning_rate": 4.497904632414879e-06,
+      "loss": 0.1672,
+      "step": 13665
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2142499210913824,
+      "learning_rate": 4.496529444389206e-06,
+      "loss": 0.1883,
+      "step": 13666
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0431450995646872,
+      "learning_rate": 4.495154405645482e-06,
+      "loss": 0.1901,
+      "step": 13667
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0691190749628912,
+      "learning_rate": 4.493779516221009e-06,
+      "loss": 0.1844,
+      "step": 13668
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7628766475524442,
+      "learning_rate": 4.492404776153078e-06,
+      "loss": 0.1729,
+      "step": 13669
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9033434819234581,
+      "learning_rate": 4.491030185478976e-06,
+      "loss": 0.1811,
+      "step": 13670
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.296385993849624,
+      "learning_rate": 4.489655744235994e-06,
+      "loss": 0.1705,
+      "step": 13671
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2469644716777546,
+      "learning_rate": 4.488281452461407e-06,
+      "loss": 0.1694,
+      "step": 13672
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8311193064293749,
+      "learning_rate": 4.4869073101925024e-06,
+      "loss": 0.1622,
+      "step": 13673
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8013546615744547,
+      "learning_rate": 4.4855333174665425e-06,
+      "loss": 0.1701,
+      "step": 13674
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.628093865865405,
+      "learning_rate": 4.484159474320804e-06,
+      "loss": 0.1751,
+      "step": 13675
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3232699202577978,
+      "learning_rate": 4.482785780792551e-06,
+      "loss": 0.169,
+      "step": 13676
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5385823443408615,
+      "learning_rate": 4.481412236919049e-06,
+      "loss": 0.1629,
+      "step": 13677
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9031756800976332,
+      "learning_rate": 4.480038842737548e-06,
+      "loss": 0.1479,
+      "step": 13678
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.257354476225968,
+      "learning_rate": 4.47866559828531e-06,
+      "loss": 0.1869,
+      "step": 13679
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9714070942255482,
+      "learning_rate": 4.477292503599574e-06,
+      "loss": 0.1822,
+      "step": 13680
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1593533989119276,
+      "learning_rate": 4.475919558717596e-06,
+      "loss": 0.1904,
+      "step": 13681
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5104136020876684,
+      "learning_rate": 4.474546763676607e-06,
+      "loss": 0.1784,
+      "step": 13682
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2655799392585738,
+      "learning_rate": 4.47317411851385e-06,
+      "loss": 0.1744,
+      "step": 13683
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2385869323659042,
+      "learning_rate": 4.471801623266558e-06,
+      "loss": 0.1445,
+      "step": 13684
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7638791201707295,
+      "learning_rate": 4.470429277971961e-06,
+      "loss": 0.1627,
+      "step": 13685
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0250060836714399,
+      "learning_rate": 4.469057082667283e-06,
+      "loss": 0.196,
+      "step": 13686
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.077268257744968,
+      "learning_rate": 4.46768503738974e-06,
+      "loss": 0.158,
+      "step": 13687
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9390479740377202,
+      "learning_rate": 4.466313142176557e-06,
+      "loss": 0.17,
+      "step": 13688
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.799024685884593,
+      "learning_rate": 4.464941397064938e-06,
+      "loss": 0.1583,
+      "step": 13689
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.49710107221133,
+      "learning_rate": 4.4635698020921016e-06,
+      "loss": 0.183,
+      "step": 13690
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5396832424734,
+      "learning_rate": 4.462198357295242e-06,
+      "loss": 0.1738,
+      "step": 13691
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8603738965306661,
+      "learning_rate": 4.460827062711564e-06,
+      "loss": 0.1618,
+      "step": 13692
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8633705184933675,
+      "learning_rate": 4.459455918378266e-06,
+      "loss": 0.1624,
+      "step": 13693
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.315956016828563,
+      "learning_rate": 4.458084924332543e-06,
+      "loss": 0.1703,
+      "step": 13694
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.860348062193069,
+      "learning_rate": 4.456714080611575e-06,
+      "loss": 0.1672,
+      "step": 13695
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9190060904210984,
+      "learning_rate": 4.455343387252555e-06,
+      "loss": 0.1822,
+      "step": 13696
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7940459239761528,
+      "learning_rate": 4.453972844292654e-06,
+      "loss": 0.1737,
+      "step": 13697
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.097721016626529,
+      "learning_rate": 4.452602451769058e-06,
+      "loss": 0.1628,
+      "step": 13698
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0759386647310396,
+      "learning_rate": 4.4512322097189295e-06,
+      "loss": 0.167,
+      "step": 13699
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3701935230414497,
+      "learning_rate": 4.449862118179444e-06,
+      "loss": 0.1584,
+      "step": 13700
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9541634410686334,
+      "learning_rate": 4.44849217718776e-06,
+      "loss": 0.1795,
+      "step": 13701
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9153389222149987,
+      "learning_rate": 4.447122386781038e-06,
+      "loss": 0.1872,
+      "step": 13702
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9417906727918101,
+      "learning_rate": 4.445752746996438e-06,
+      "loss": 0.1725,
+      "step": 13703
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1049319774043718,
+      "learning_rate": 4.4443832578711055e-06,
+      "loss": 0.1989,
+      "step": 13704
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9056218269926907,
+      "learning_rate": 4.443013919442194e-06,
+      "loss": 0.1826,
+      "step": 13705
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.984890726452985,
+      "learning_rate": 4.4416447317468405e-06,
+      "loss": 0.1716,
+      "step": 13706
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0207518181035065,
+      "learning_rate": 4.440275694822192e-06,
+      "loss": 0.166,
+      "step": 13707
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0303227419266743,
+      "learning_rate": 4.438906808705374e-06,
+      "loss": 0.1637,
+      "step": 13708
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0069749895417077,
+      "learning_rate": 4.437538073433524e-06,
+      "loss": 0.1834,
+      "step": 13709
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.11647684150671,
+      "learning_rate": 4.436169489043768e-06,
+      "loss": 0.2011,
+      "step": 13710
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.034624038537078,
+      "learning_rate": 4.434801055573232e-06,
+      "loss": 0.1871,
+      "step": 13711
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5391380793012959,
+      "learning_rate": 4.433432773059028e-06,
+      "loss": 0.155,
+      "step": 13712
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0803455955807053,
+      "learning_rate": 4.432064641538279e-06,
+      "loss": 0.1773,
+      "step": 13713
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9100790739475726,
+      "learning_rate": 4.430696661048086e-06,
+      "loss": 0.1683,
+      "step": 13714
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0718263998332211,
+      "learning_rate": 4.429328831625565e-06,
+      "loss": 0.1694,
+      "step": 13715
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1892483023347649,
+      "learning_rate": 4.427961153307811e-06,
+      "loss": 0.1673,
+      "step": 13716
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.207942887979543,
+      "learning_rate": 4.426593626131928e-06,
+      "loss": 0.1533,
+      "step": 13717
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.888967603029378,
+      "learning_rate": 4.425226250135005e-06,
+      "loss": 0.1736,
+      "step": 13718
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6935255519516728,
+      "learning_rate": 4.4238590253541335e-06,
+      "loss": 0.183,
+      "step": 13719
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.4630156900562032,
+      "learning_rate": 4.422491951826402e-06,
+      "loss": 0.1638,
+      "step": 13720
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2770553329985392,
+      "learning_rate": 4.421125029588895e-06,
+      "loss": 0.166,
+      "step": 13721
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5433447524383987,
+      "learning_rate": 4.419758258678687e-06,
+      "loss": 0.173,
+      "step": 13722
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0737467851992872,
+      "learning_rate": 4.418391639132847e-06,
+      "loss": 0.1707,
+      "step": 13723
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9946446709389992,
+      "learning_rate": 4.4170251709884526e-06,
+      "loss": 0.1785,
+      "step": 13724
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4318930340410951,
+      "learning_rate": 4.4156588542825625e-06,
+      "loss": 0.1586,
+      "step": 13725
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.510405383262583,
+      "learning_rate": 4.414292689052241e-06,
+      "loss": 0.1779,
+      "step": 13726
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9444011738801847,
+      "learning_rate": 4.412926675334546e-06,
+      "loss": 0.1636,
+      "step": 13727
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2076138811368429,
+      "learning_rate": 4.411560813166535e-06,
+      "loss": 0.1708,
+      "step": 13728
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.370268298456576,
+      "learning_rate": 4.410195102585247e-06,
+      "loss": 0.1799,
+      "step": 13729
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9673169171945579,
+      "learning_rate": 4.408829543627737e-06,
+      "loss": 0.1746,
+      "step": 13730
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8898649802675859,
+      "learning_rate": 4.407464136331039e-06,
+      "loss": 0.169,
+      "step": 13731
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8779739505312871,
+      "learning_rate": 4.406098880732195e-06,
+      "loss": 0.1844,
+      "step": 13732
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0552422789470997,
+      "learning_rate": 4.404733776868231e-06,
+      "loss": 0.1786,
+      "step": 13733
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3731759447314973,
+      "learning_rate": 4.403368824776183e-06,
+      "loss": 0.1718,
+      "step": 13734
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8969742110147156,
+      "learning_rate": 4.402004024493069e-06,
+      "loss": 0.1735,
+      "step": 13735
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8454532379142259,
+      "learning_rate": 4.4006393760559105e-06,
+      "loss": 0.1647,
+      "step": 13736
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.719264383388367,
+      "learning_rate": 4.399274879501726e-06,
+      "loss": 0.1765,
+      "step": 13737
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1110548054408023,
+      "learning_rate": 4.39791053486753e-06,
+      "loss": 0.1862,
+      "step": 13738
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8295730102929992,
+      "learning_rate": 4.396546342190327e-06,
+      "loss": 0.1518,
+      "step": 13739
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.904612012615748,
+      "learning_rate": 4.3951823015071186e-06,
+      "loss": 0.1683,
+      "step": 13740
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0040002316631793,
+      "learning_rate": 4.393818412854909e-06,
+      "loss": 0.1811,
+      "step": 13741
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2752164591376416,
+      "learning_rate": 4.392454676270687e-06,
+      "loss": 0.1604,
+      "step": 13742
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0096690577838259,
+      "learning_rate": 4.391091091791455e-06,
+      "loss": 0.1636,
+      "step": 13743
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3945369984716742,
+      "learning_rate": 4.389727659454189e-06,
+      "loss": 0.1741,
+      "step": 13744
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8191888837385805,
+      "learning_rate": 4.388364379295878e-06,
+      "loss": 0.1619,
+      "step": 13745
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4411158575624718,
+      "learning_rate": 4.387001251353499e-06,
+      "loss": 0.1612,
+      "step": 13746
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6639282345204542,
+      "learning_rate": 4.3856382756640315e-06,
+      "loss": 0.1821,
+      "step": 13747
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9922552014129387,
+      "learning_rate": 4.3842754522644394e-06,
+      "loss": 0.1863,
+      "step": 13748
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1021385598823878,
+      "learning_rate": 4.382912781191697e-06,
+      "loss": 0.1895,
+      "step": 13749
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0179098567527967,
+      "learning_rate": 4.381550262482759e-06,
+      "loss": 0.1586,
+      "step": 13750
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.999589172516189,
+      "learning_rate": 4.380187896174591e-06,
+      "loss": 0.191,
+      "step": 13751
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0142183508137548,
+      "learning_rate": 4.37882568230414e-06,
+      "loss": 0.1714,
+      "step": 13752
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7396820691494044,
+      "learning_rate": 4.37746362090836e-06,
+      "loss": 0.1879,
+      "step": 13753
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7150075562181797,
+      "learning_rate": 4.376101712024197e-06,
+      "loss": 0.1967,
+      "step": 13754
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9169604172984819,
+      "learning_rate": 4.374739955688595e-06,
+      "loss": 0.1517,
+      "step": 13755
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5748245219247363,
+      "learning_rate": 4.373378351938491e-06,
+      "loss": 0.1449,
+      "step": 13756
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9426798859246521,
+      "learning_rate": 4.372016900810813e-06,
+      "loss": 0.1728,
+      "step": 13757
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9292767642651203,
+      "learning_rate": 4.370655602342497e-06,
+      "loss": 0.1665,
+      "step": 13758
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9681826868018988,
+      "learning_rate": 4.369294456570463e-06,
+      "loss": 0.1758,
+      "step": 13759
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1272505686047567,
+      "learning_rate": 4.3679334635316395e-06,
+      "loss": 0.183,
+      "step": 13760
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3020722492673074,
+      "learning_rate": 4.366572623262934e-06,
+      "loss": 0.1595,
+      "step": 13761
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7390029556224693,
+      "learning_rate": 4.365211935801264e-06,
+      "loss": 0.1916,
+      "step": 13762
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.075142392697466,
+      "learning_rate": 4.363851401183539e-06,
+      "loss": 0.1826,
+      "step": 13763
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.937147420005713,
+      "learning_rate": 4.3624910194466675e-06,
+      "loss": 0.1843,
+      "step": 13764
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.87810639949517,
+      "learning_rate": 4.361130790627541e-06,
+      "loss": 0.1532,
+      "step": 13765
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.308672685903403,
+      "learning_rate": 4.3597707147630645e-06,
+      "loss": 0.1758,
+      "step": 13766
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.514058633514923,
+      "learning_rate": 4.358410791890122e-06,
+      "loss": 0.1718,
+      "step": 13767
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2272785034178764,
+      "learning_rate": 4.357051022045608e-06,
+      "loss": 0.171,
+      "step": 13768
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8642054250805404,
+      "learning_rate": 4.355691405266401e-06,
+      "loss": 0.1507,
+      "step": 13769
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0457190185214238,
+      "learning_rate": 4.354331941589387e-06,
+      "loss": 0.1679,
+      "step": 13770
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2799911203581829,
+      "learning_rate": 4.352972631051435e-06,
+      "loss": 0.1879,
+      "step": 13771
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6635745340448762,
+      "learning_rate": 4.351613473689419e-06,
+      "loss": 0.1617,
+      "step": 13772
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8447529424528541,
+      "learning_rate": 4.350254469540209e-06,
+      "loss": 0.1517,
+      "step": 13773
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5466512179947478,
+      "learning_rate": 4.348895618640663e-06,
+      "loss": 0.1748,
+      "step": 13774
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8568594087610589,
+      "learning_rate": 4.347536921027646e-06,
+      "loss": 0.1654,
+      "step": 13775
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1145052751272784,
+      "learning_rate": 4.346178376738006e-06,
+      "loss": 0.193,
+      "step": 13776
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8522218692305981,
+      "learning_rate": 4.344819985808601e-06,
+      "loss": 0.1649,
+      "step": 13777
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0989297052512519,
+      "learning_rate": 4.343461748276267e-06,
+      "loss": 0.1849,
+      "step": 13778
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2810290352563842,
+      "learning_rate": 4.342103664177856e-06,
+      "loss": 0.172,
+      "step": 13779
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.72999235370836,
+      "learning_rate": 4.3407457335502e-06,
+      "loss": 0.1643,
+      "step": 13780
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9451376817683268,
+      "learning_rate": 4.339387956430141e-06,
+      "loss": 0.1646,
+      "step": 13781
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.321991051912222,
+      "learning_rate": 4.3380303328545e-06,
+      "loss": 0.1706,
+      "step": 13782
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1961186962090051,
+      "learning_rate": 4.336672862860107e-06,
+      "loss": 0.185,
+      "step": 13783
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9379183664300133,
+      "learning_rate": 4.335315546483781e-06,
+      "loss": 0.1697,
+      "step": 13784
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9396690803575051,
+      "learning_rate": 4.333958383762345e-06,
+      "loss": 0.1493,
+      "step": 13785
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.143837166113807,
+      "learning_rate": 4.332601374732602e-06,
+      "loss": 0.1744,
+      "step": 13786
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9022516815447942,
+      "learning_rate": 4.331244519431371e-06,
+      "loss": 0.1565,
+      "step": 13787
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.034186295116794,
+      "learning_rate": 4.329887817895451e-06,
+      "loss": 0.1756,
+      "step": 13788
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9630885441659184,
+      "learning_rate": 4.328531270161642e-06,
+      "loss": 0.1712,
+      "step": 13789
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0721431552611012,
+      "learning_rate": 4.327174876266743e-06,
+      "loss": 0.1594,
+      "step": 13790
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.225581281392638,
+      "learning_rate": 4.325818636247549e-06,
+      "loss": 0.1556,
+      "step": 13791
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0627692078353024,
+      "learning_rate": 4.324462550140847e-06,
+      "loss": 0.1749,
+      "step": 13792
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9279764998021093,
+      "learning_rate": 4.323106617983414e-06,
+      "loss": 0.1599,
+      "step": 13793
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1209255773185283,
+      "learning_rate": 4.321750839812038e-06,
+      "loss": 0.1773,
+      "step": 13794
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2147773049589803,
+      "learning_rate": 4.320395215663488e-06,
+      "loss": 0.1608,
+      "step": 13795
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.355886040723802,
+      "learning_rate": 4.319039745574543e-06,
+      "loss": 0.1587,
+      "step": 13796
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.889548101640906,
+      "learning_rate": 4.317684429581961e-06,
+      "loss": 0.1651,
+      "step": 13797
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9541565186730563,
+      "learning_rate": 4.316329267722509e-06,
+      "loss": 0.1481,
+      "step": 13798
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9623375218408832,
+      "learning_rate": 4.314974260032948e-06,
+      "loss": 0.1656,
+      "step": 13799
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8715958990471303,
+      "learning_rate": 4.313619406550034e-06,
+      "loss": 0.167,
+      "step": 13800
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9084593575369279,
+      "learning_rate": 4.3122647073105114e-06,
+      "loss": 0.182,
+      "step": 13801
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3108811802318745,
+      "learning_rate": 4.310910162351134e-06,
+      "loss": 0.1538,
+      "step": 13802
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3121975048987127,
+      "learning_rate": 4.309555771708637e-06,
+      "loss": 0.1871,
+      "step": 13803
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0499623681533017,
+      "learning_rate": 4.308201535419762e-06,
+      "loss": 0.1691,
+      "step": 13804
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0081547694953512,
+      "learning_rate": 4.306847453521241e-06,
+      "loss": 0.1726,
+      "step": 13805
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5058682236514365,
+      "learning_rate": 4.305493526049803e-06,
+      "loss": 0.1697,
+      "step": 13806
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9462720026788233,
+      "learning_rate": 4.304139753042174e-06,
+      "loss": 0.186,
+      "step": 13807
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8945622051574326,
+      "learning_rate": 4.3027861345350805e-06,
+      "loss": 0.1733,
+      "step": 13808
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9428319169479271,
+      "learning_rate": 4.301432670565235e-06,
+      "loss": 0.1587,
+      "step": 13809
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8628655152606859,
+      "learning_rate": 4.300079361169347e-06,
+      "loss": 0.1919,
+      "step": 13810
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9129736896773106,
+      "learning_rate": 4.2987262063841316e-06,
+      "loss": 0.1778,
+      "step": 13811
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0113974153294598,
+      "learning_rate": 4.297373206246286e-06,
+      "loss": 0.1683,
+      "step": 13812
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0648951024839919,
+      "learning_rate": 4.296020360792518e-06,
+      "loss": 0.1514,
+      "step": 13813
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.073923935323422,
+      "learning_rate": 4.2946676700595155e-06,
+      "loss": 0.16,
+      "step": 13814
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3478803045424044,
+      "learning_rate": 4.293315134083975e-06,
+      "loss": 0.195,
+      "step": 13815
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5146714760895041,
+      "learning_rate": 4.291962752902584e-06,
+      "loss": 0.1863,
+      "step": 13816
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.2258691029557385,
+      "learning_rate": 4.2906105265520295e-06,
+      "loss": 0.1881,
+      "step": 13817
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4088017704480014,
+      "learning_rate": 4.289258455068983e-06,
+      "loss": 0.1969,
+      "step": 13818
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9642670080882169,
+      "learning_rate": 4.287906538490128e-06,
+      "loss": 0.1634,
+      "step": 13819
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0240549553314455,
+      "learning_rate": 4.286554776852125e-06,
+      "loss": 0.1666,
+      "step": 13820
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1796958409744949,
+      "learning_rate": 4.285203170191652e-06,
+      "loss": 0.1694,
+      "step": 13821
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9443459190506037,
+      "learning_rate": 4.283851718545362e-06,
+      "loss": 0.1679,
+      "step": 13822
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9646991551786035,
+      "learning_rate": 4.282500421949917e-06,
+      "loss": 0.1596,
+      "step": 13823
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2808704572171308,
+      "learning_rate": 4.2811492804419695e-06,
+      "loss": 0.1662,
+      "step": 13824
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.347312322023975,
+      "learning_rate": 4.279798294058176e-06,
+      "loss": 0.1772,
+      "step": 13825
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.784693189124559,
+      "learning_rate": 4.278447462835176e-06,
+      "loss": 0.186,
+      "step": 13826
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4600284370723755,
+      "learning_rate": 4.277096786809608e-06,
+      "loss": 0.1915,
+      "step": 13827
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2121081339526243,
+      "learning_rate": 4.275746266018117e-06,
+      "loss": 0.1782,
+      "step": 13828
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0950423022369236,
+      "learning_rate": 4.274395900497328e-06,
+      "loss": 0.1945,
+      "step": 13829
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1221579992549777,
+      "learning_rate": 4.273045690283878e-06,
+      "loss": 0.1806,
+      "step": 13830
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0292814625435251,
+      "learning_rate": 4.2716956354143826e-06,
+      "loss": 0.1599,
+      "step": 13831
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0910953828917187,
+      "learning_rate": 4.2703457359254665e-06,
+      "loss": 0.1721,
+      "step": 13832
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9996097948773663,
+      "learning_rate": 4.268995991853746e-06,
+      "loss": 0.1676,
+      "step": 13833
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1543389514703013,
+      "learning_rate": 4.267646403235836e-06,
+      "loss": 0.1964,
+      "step": 13834
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5130531821660502,
+      "learning_rate": 4.266296970108339e-06,
+      "loss": 0.1613,
+      "step": 13835
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8845460390820609,
+      "learning_rate": 4.264947692507863e-06,
+      "loss": 0.1708,
+      "step": 13836
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9841852132957352,
+      "learning_rate": 4.263598570471003e-06,
+      "loss": 0.1861,
+      "step": 13837
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9041726518279823,
+      "learning_rate": 4.262249604034356e-06,
+      "loss": 0.1699,
+      "step": 13838
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0345479895144427,
+      "learning_rate": 4.260900793234511e-06,
+      "loss": 0.1663,
+      "step": 13839
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8941284391965324,
+      "learning_rate": 4.259552138108061e-06,
+      "loss": 0.1577,
+      "step": 13840
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9192186532010812,
+      "learning_rate": 4.258203638691578e-06,
+      "loss": 0.1976,
+      "step": 13841
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9715839954624031,
+      "learning_rate": 4.256855295021646e-06,
+      "loss": 0.1523,
+      "step": 13842
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9026243535838973,
+      "learning_rate": 4.255507107134842e-06,
+      "loss": 0.1917,
+      "step": 13843
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8797153394563227,
+      "learning_rate": 4.2541590750677285e-06,
+      "loss": 0.1531,
+      "step": 13844
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9451326989379703,
+      "learning_rate": 4.252811198856878e-06,
+      "loss": 0.1708,
+      "step": 13845
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0833224058149244,
+      "learning_rate": 4.251463478538846e-06,
+      "loss": 0.2052,
+      "step": 13846
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3103829188590623,
+      "learning_rate": 4.250115914150194e-06,
+      "loss": 0.1823,
+      "step": 13847
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.94943048980084,
+      "learning_rate": 4.2487685057274695e-06,
+      "loss": 0.1938,
+      "step": 13848
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.559673614053583,
+      "learning_rate": 4.247421253307225e-06,
+      "loss": 0.1897,
+      "step": 13849
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9774268887965165,
+      "learning_rate": 4.246074156926002e-06,
+      "loss": 0.1596,
+      "step": 13850
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4407846971193052,
+      "learning_rate": 4.244727216620348e-06,
+      "loss": 0.1621,
+      "step": 13851
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0361937161045285,
+      "learning_rate": 4.2433804324267895e-06,
+      "loss": 0.1804,
+      "step": 13852
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0412298692371762,
+      "learning_rate": 4.242033804381864e-06,
+      "loss": 0.1917,
+      "step": 13853
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9114379786274113,
+      "learning_rate": 4.240687332522094e-06,
+      "loss": 0.1873,
+      "step": 13854
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6741798793588893,
+      "learning_rate": 4.239341016884008e-06,
+      "loss": 0.1743,
+      "step": 13855
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9870271064861393,
+      "learning_rate": 4.237994857504121e-06,
+      "loss": 0.1615,
+      "step": 13856
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4861495315901574,
+      "learning_rate": 4.236648854418951e-06,
+      "loss": 0.1603,
+      "step": 13857
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.035190561600465,
+      "learning_rate": 4.2353030076650025e-06,
+      "loss": 0.1593,
+      "step": 13858
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1910756738313253,
+      "learning_rate": 4.233957317278786e-06,
+      "loss": 0.16,
+      "step": 13859
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5194077790731053,
+      "learning_rate": 4.232611783296804e-06,
+      "loss": 0.1933,
+      "step": 13860
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9646802011211995,
+      "learning_rate": 4.2312664057555556e-06,
+      "loss": 0.1781,
+      "step": 13861
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4234201229578838,
+      "learning_rate": 4.229921184691531e-06,
+      "loss": 0.1787,
+      "step": 13862
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4329485564112754,
+      "learning_rate": 4.228576120141218e-06,
+      "loss": 0.1733,
+      "step": 13863
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7625806077048197,
+      "learning_rate": 4.2272312121411065e-06,
+      "loss": 0.1707,
+      "step": 13864
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0778457587652996,
+      "learning_rate": 4.225886460727671e-06,
+      "loss": 0.1932,
+      "step": 13865
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.398154670499273,
+      "learning_rate": 4.224541865937395e-06,
+      "loss": 0.158,
+      "step": 13866
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.156600814601921,
+      "learning_rate": 4.2231974278067436e-06,
+      "loss": 0.1709,
+      "step": 13867
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1124745144642676,
+      "learning_rate": 4.221853146372188e-06,
+      "loss": 0.195,
+      "step": 13868
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9143329997002765,
+      "learning_rate": 4.220509021670193e-06,
+      "loss": 0.1533,
+      "step": 13869
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2854654668501229,
+      "learning_rate": 4.21916505373722e-06,
+      "loss": 0.1698,
+      "step": 13870
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7644388099664177,
+      "learning_rate": 4.2178212426097175e-06,
+      "loss": 0.1726,
+      "step": 13871
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4484549246924208,
+      "learning_rate": 4.216477588324144e-06,
+      "loss": 0.1723,
+      "step": 13872
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8021444948614584,
+      "learning_rate": 4.215134090916939e-06,
+      "loss": 0.1833,
+      "step": 13873
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9335939453621243,
+      "learning_rate": 4.213790750424553e-06,
+      "loss": 0.1612,
+      "step": 13874
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0977518688232104,
+      "learning_rate": 4.212447566883415e-06,
+      "loss": 0.1641,
+      "step": 13875
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.220894291615603,
+      "learning_rate": 4.211104540329964e-06,
+      "loss": 0.1739,
+      "step": 13876
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9058485613346287,
+      "learning_rate": 4.209761670800631e-06,
+      "loss": 0.1364,
+      "step": 13877
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.94948120515998,
+      "learning_rate": 4.208418958331841e-06,
+      "loss": 0.1656,
+      "step": 13878
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0817260499443833,
+      "learning_rate": 4.207076402960015e-06,
+      "loss": 0.1685,
+      "step": 13879
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4737839935785986,
+      "learning_rate": 4.205734004721565e-06,
+      "loss": 0.1674,
+      "step": 13880
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2886994072038556,
+      "learning_rate": 4.204391763652911e-06,
+      "loss": 0.1618,
+      "step": 13881
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0956350599041518,
+      "learning_rate": 4.2030496797904526e-06,
+      "loss": 0.1589,
+      "step": 13882
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5778971792081893,
+      "learning_rate": 4.2017077531706056e-06,
+      "loss": 0.1754,
+      "step": 13883
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.280024551622798,
+      "learning_rate": 4.200365983829757e-06,
+      "loss": 0.1679,
+      "step": 13884
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0989316304641512,
+      "learning_rate": 4.19902437180431e-06,
+      "loss": 0.179,
+      "step": 13885
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0262960765847735,
+      "learning_rate": 4.197682917130654e-06,
+      "loss": 0.1701,
+      "step": 13886
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3649534886612573,
+      "learning_rate": 4.196341619845182e-06,
+      "loss": 0.1682,
+      "step": 13887
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9768879481607011,
+      "learning_rate": 4.195000479984264e-06,
+      "loss": 0.1555,
+      "step": 13888
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2674514431589723,
+      "learning_rate": 4.193659497584293e-06,
+      "loss": 0.1718,
+      "step": 13889
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9301726568541404,
+      "learning_rate": 4.192318672681631e-06,
+      "loss": 0.1667,
+      "step": 13890
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6447104967367812,
+      "learning_rate": 4.190978005312657e-06,
+      "loss": 0.1883,
+      "step": 13891
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0557492480294897,
+      "learning_rate": 4.189637495513729e-06,
+      "loss": 0.1608,
+      "step": 13892
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1176874894671438,
+      "learning_rate": 4.188297143321215e-06,
+      "loss": 0.1801,
+      "step": 13893
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9404386916814441,
+      "learning_rate": 4.186956948771467e-06,
+      "loss": 0.1617,
+      "step": 13894
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1853230444085585,
+      "learning_rate": 4.1856169119008384e-06,
+      "loss": 0.19,
+      "step": 13895
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2015223083588877,
+      "learning_rate": 4.184277032745685e-06,
+      "loss": 0.1627,
+      "step": 13896
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8969995228287397,
+      "learning_rate": 4.18293731134234e-06,
+      "loss": 0.1409,
+      "step": 13897
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7125517882768655,
+      "learning_rate": 4.181597747727154e-06,
+      "loss": 0.1916,
+      "step": 13898
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.090728912294368,
+      "learning_rate": 4.180258341936454e-06,
+      "loss": 0.1636,
+      "step": 13899
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5086909915665916,
+      "learning_rate": 4.178919094006578e-06,
+      "loss": 0.1764,
+      "step": 13900
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0281672216497417,
+      "learning_rate": 4.1775800039738465e-06,
+      "loss": 0.1694,
+      "step": 13901
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9611875987842856,
+      "learning_rate": 4.176241071874587e-06,
+      "loss": 0.1667,
+      "step": 13902
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1189889394442891,
+      "learning_rate": 4.174902297745118e-06,
+      "loss": 0.1845,
+      "step": 13903
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0347532611945578,
+      "learning_rate": 4.173563681621756e-06,
+      "loss": 0.1723,
+      "step": 13904
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4840919450107504,
+      "learning_rate": 4.1722252235408045e-06,
+      "loss": 0.1786,
+      "step": 13905
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7848343757219844,
+      "learning_rate": 4.170886923538576e-06,
+      "loss": 0.1658,
+      "step": 13906
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1563399247561783,
+      "learning_rate": 4.169548781651367e-06,
+      "loss": 0.1772,
+      "step": 13907
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.231024214373564,
+      "learning_rate": 4.168210797915479e-06,
+      "loss": 0.1483,
+      "step": 13908
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2235509447303758,
+      "learning_rate": 4.1668729723671994e-06,
+      "loss": 0.163,
+      "step": 13909
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9748722624842394,
+      "learning_rate": 4.165535305042822e-06,
+      "loss": 0.1544,
+      "step": 13910
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1483576747353885,
+      "learning_rate": 4.164197795978628e-06,
+      "loss": 0.1721,
+      "step": 13911
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.011189049498867,
+      "learning_rate": 4.162860445210897e-06,
+      "loss": 0.1866,
+      "step": 13912
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3078625368117252,
+      "learning_rate": 4.16152325277591e-06,
+      "loss": 0.1789,
+      "step": 13913
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9965484103475903,
+      "learning_rate": 4.16018621870993e-06,
+      "loss": 0.2024,
+      "step": 13914
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.021616309600682,
+      "learning_rate": 4.158849343049233e-06,
+      "loss": 0.1891,
+      "step": 13915
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1905640579670889,
+      "learning_rate": 4.157512625830074e-06,
+      "loss": 0.1679,
+      "step": 13916
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9737795035987967,
+      "learning_rate": 4.156176067088717e-06,
+      "loss": 0.1768,
+      "step": 13917
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9511517663436382,
+      "learning_rate": 4.154839666861413e-06,
+      "loss": 0.1708,
+      "step": 13918
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1916910017235933,
+      "learning_rate": 4.153503425184415e-06,
+      "loss": 0.1807,
+      "step": 13919
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2415657986375965,
+      "learning_rate": 4.152167342093965e-06,
+      "loss": 0.1531,
+      "step": 13920
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4514494878822788,
+      "learning_rate": 4.150831417626304e-06,
+      "loss": 0.1882,
+      "step": 13921
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9120073512123998,
+      "learning_rate": 4.149495651817673e-06,
+      "loss": 0.1664,
+      "step": 13922
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9726956622707574,
+      "learning_rate": 4.148160044704306e-06,
+      "loss": 0.1741,
+      "step": 13923
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9938501765351079,
+      "learning_rate": 4.1468245963224245e-06,
+      "loss": 0.1629,
+      "step": 13924
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.102826239226895,
+      "learning_rate": 4.1454893067082605e-06,
+      "loss": 0.1704,
+      "step": 13925
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8498416122497809,
+      "learning_rate": 4.1441541758980256e-06,
+      "loss": 0.1589,
+      "step": 13926
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5947741621857945,
+      "learning_rate": 4.142819203927942e-06,
+      "loss": 0.2137,
+      "step": 13927
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.994381854280893,
+      "learning_rate": 4.141484390834216e-06,
+      "loss": 0.1654,
+      "step": 13928
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9255592604899666,
+      "learning_rate": 4.140149736653056e-06,
+      "loss": 0.1751,
+      "step": 13929
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5580640598570554,
+      "learning_rate": 4.138815241420666e-06,
+      "loss": 0.1859,
+      "step": 13930
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8700202663936708,
+      "learning_rate": 4.137480905173248e-06,
+      "loss": 0.1566,
+      "step": 13931
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5028060839405217,
+      "learning_rate": 4.13614672794699e-06,
+      "loss": 0.1481,
+      "step": 13932
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2787088758138214,
+      "learning_rate": 4.13481270977808e-06,
+      "loss": 0.1806,
+      "step": 13933
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0328407095066583,
+      "learning_rate": 4.133478850702711e-06,
+      "loss": 0.1802,
+      "step": 13934
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9213280279429286,
+      "learning_rate": 4.1321451507570555e-06,
+      "loss": 0.1738,
+      "step": 13935
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.13141968709934,
+      "learning_rate": 4.130811609977297e-06,
+      "loss": 0.1698,
+      "step": 13936
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9087622452263413,
+      "learning_rate": 4.1294782283996024e-06,
+      "loss": 0.189,
+      "step": 13937
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7743718454839085,
+      "learning_rate": 4.128145006060141e-06,
+      "loss": 0.1608,
+      "step": 13938
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.008167328082802,
+      "learning_rate": 4.12681194299508e-06,
+      "loss": 0.1724,
+      "step": 13939
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3185350228359154,
+      "learning_rate": 4.12547903924058e-06,
+      "loss": 0.1587,
+      "step": 13940
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2549184563659543,
+      "learning_rate": 4.124146294832788e-06,
+      "loss": 0.1831,
+      "step": 13941
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.252213411815056,
+      "learning_rate": 4.122813709807864e-06,
+      "loss": 0.177,
+      "step": 13942
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0130254442875402,
+      "learning_rate": 4.121481284201946e-06,
+      "loss": 0.1875,
+      "step": 13943
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0870914766560396,
+      "learning_rate": 4.120149018051184e-06,
+      "loss": 0.1803,
+      "step": 13944
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0470809970557806,
+      "learning_rate": 4.118816911391709e-06,
+      "loss": 0.1668,
+      "step": 13945
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.943477282672622,
+      "learning_rate": 4.117484964259657e-06,
+      "loss": 0.1635,
+      "step": 13946
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.101166936528451,
+      "learning_rate": 4.116153176691158e-06,
+      "loss": 0.1645,
+      "step": 13947
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.306797221498104,
+      "learning_rate": 4.1148215487223385e-06,
+      "loss": 0.1637,
+      "step": 13948
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9415598236086736,
+      "learning_rate": 4.1134900803893185e-06,
+      "loss": 0.1946,
+      "step": 13949
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.200965620690179,
+      "learning_rate": 4.1121587717282085e-06,
+      "loss": 0.1801,
+      "step": 13950
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.141433023565065,
+      "learning_rate": 4.110827622775128e-06,
+      "loss": 0.1821,
+      "step": 13951
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.305752632035842,
+      "learning_rate": 4.1094966335661765e-06,
+      "loss": 0.1645,
+      "step": 13952
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1922977922503717,
+      "learning_rate": 4.108165804137466e-06,
+      "loss": 0.1694,
+      "step": 13953
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0802621937995167,
+      "learning_rate": 4.106835134525087e-06,
+      "loss": 0.1686,
+      "step": 13954
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0049350043646883,
+      "learning_rate": 4.105504624765137e-06,
+      "loss": 0.176,
+      "step": 13955
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.762105872660763,
+      "learning_rate": 4.104174274893709e-06,
+      "loss": 0.168,
+      "step": 13956
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.30451208977044,
+      "learning_rate": 4.102844084946889e-06,
+      "loss": 0.1608,
+      "step": 13957
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.106594812369558,
+      "learning_rate": 4.101514054960752e-06,
+      "loss": 0.1842,
+      "step": 13958
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0997050675526354,
+      "learning_rate": 4.1001841849713845e-06,
+      "loss": 0.1939,
+      "step": 13959
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0702076106090137,
+      "learning_rate": 4.098854475014849e-06,
+      "loss": 0.2071,
+      "step": 13960
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9761023790628669,
+      "learning_rate": 4.097524925127224e-06,
+      "loss": 0.1787,
+      "step": 13961
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0409487600060539,
+      "learning_rate": 4.096195535344565e-06,
+      "loss": 0.1671,
+      "step": 13962
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1941185796143974,
+      "learning_rate": 4.0948663057029395e-06,
+      "loss": 0.1842,
+      "step": 13963
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0081634661008763,
+      "learning_rate": 4.093537236238394e-06,
+      "loss": 0.1734,
+      "step": 13964
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1712157900084783,
+      "learning_rate": 4.092208326986986e-06,
+      "loss": 0.1735,
+      "step": 13965
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.493318392656283,
+      "learning_rate": 4.090879577984763e-06,
+      "loss": 0.1534,
+      "step": 13966
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.76818294481959,
+      "learning_rate": 4.089550989267763e-06,
+      "loss": 0.1598,
+      "step": 13967
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3472354685096652,
+      "learning_rate": 4.0882225608720295e-06,
+      "loss": 0.1624,
+      "step": 13968
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9148780436687676,
+      "learning_rate": 4.086894292833589e-06,
+      "loss": 0.1638,
+      "step": 13969
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7247586560020707,
+      "learning_rate": 4.085566185188478e-06,
+      "loss": 0.1865,
+      "step": 13970
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8492698486093321,
+      "learning_rate": 4.084238237972715e-06,
+      "loss": 0.169,
+      "step": 13971
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1954560908592897,
+      "learning_rate": 4.082910451222325e-06,
+      "loss": 0.1825,
+      "step": 13972
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0239596161957574,
+      "learning_rate": 4.0815828249733226e-06,
+      "loss": 0.1693,
+      "step": 13973
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9171027277912834,
+      "learning_rate": 4.080255359261723e-06,
+      "loss": 0.1612,
+      "step": 13974
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9608743201810395,
+      "learning_rate": 4.078928054123529e-06,
+      "loss": 0.1791,
+      "step": 13975
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9961691965383497,
+      "learning_rate": 4.077600909594748e-06,
+      "loss": 0.1706,
+      "step": 13976
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.336294130271254,
+      "learning_rate": 4.0762739257113734e-06,
+      "loss": 0.1851,
+      "step": 13977
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0139466768340508,
+      "learning_rate": 4.074947102509408e-06,
+      "loss": 0.1655,
+      "step": 13978
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2893579384169815,
+      "learning_rate": 4.073620440024832e-06,
+      "loss": 0.1941,
+      "step": 13979
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9832599128423222,
+      "learning_rate": 4.072293938293641e-06,
+      "loss": 0.1756,
+      "step": 13980
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0238164537959138,
+      "learning_rate": 4.070967597351808e-06,
+      "loss": 0.1858,
+      "step": 13981
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0183251618475806,
+      "learning_rate": 4.069641417235314e-06,
+      "loss": 0.1804,
+      "step": 13982
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9810496013205331,
+      "learning_rate": 4.068315397980135e-06,
+      "loss": 0.1718,
+      "step": 13983
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.082129763356115,
+      "learning_rate": 4.066989539622232e-06,
+      "loss": 0.1605,
+      "step": 13984
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7518739008925989,
+      "learning_rate": 4.065663842197576e-06,
+      "loss": 0.1535,
+      "step": 13985
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1024784812536153,
+      "learning_rate": 4.06433830574212e-06,
+      "loss": 0.1693,
+      "step": 13986
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4415271687351996,
+      "learning_rate": 4.0630129302918266e-06,
+      "loss": 0.145,
+      "step": 13987
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.139274381955693,
+      "learning_rate": 4.06168771588264e-06,
+      "loss": 0.1895,
+      "step": 13988
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.269776868117072,
+      "learning_rate": 4.0603626625505125e-06,
+      "loss": 0.1632,
+      "step": 13989
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9230653132887967,
+      "learning_rate": 4.059037770331379e-06,
+      "loss": 0.168,
+      "step": 13990
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0253563738907223,
+      "learning_rate": 4.057713039261182e-06,
+      "loss": 0.1611,
+      "step": 13991
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.298699424162711,
+      "learning_rate": 4.056388469375853e-06,
+      "loss": 0.1808,
+      "step": 13992
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1550292532198296,
+      "learning_rate": 4.055064060711328e-06,
+      "loss": 0.1677,
+      "step": 13993
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1411457383366315,
+      "learning_rate": 4.0537398133035225e-06,
+      "loss": 0.1628,
+      "step": 13994
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9678233487865195,
+      "learning_rate": 4.0524157271883635e-06,
+      "loss": 0.1843,
+      "step": 13995
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0057253640193635,
+      "learning_rate": 4.05109180240176e-06,
+      "loss": 0.1664,
+      "step": 13996
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8006213175726598,
+      "learning_rate": 4.049768038979631e-06,
+      "loss": 0.1781,
+      "step": 13997
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2773225084954483,
+      "learning_rate": 4.0484444369578764e-06,
+      "loss": 0.1591,
+      "step": 13998
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.18123877918416,
+      "learning_rate": 4.047120996372403e-06,
+      "loss": 0.1706,
+      "step": 13999
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6529702566124738,
+      "learning_rate": 4.045797717259109e-06,
+      "loss": 0.1905,
+      "step": 14000
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3302918737464733,
+      "learning_rate": 4.044474599653891e-06,
+      "loss": 0.1764,
+      "step": 14001
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9162525803200723,
+      "learning_rate": 4.0431516435926365e-06,
+      "loss": 0.1754,
+      "step": 14002
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.316548792491148,
+      "learning_rate": 4.0418288491112255e-06,
+      "loss": 0.1725,
+      "step": 14003
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9532510067195563,
+      "learning_rate": 4.0405062162455474e-06,
+      "loss": 0.184,
+      "step": 14004
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2904459935716295,
+      "learning_rate": 4.0391837450314715e-06,
+      "loss": 0.1592,
+      "step": 14005
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.831784096015568,
+      "learning_rate": 4.037861435504877e-06,
+      "loss": 0.1922,
+      "step": 14006
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9590948888956858,
+      "learning_rate": 4.0365392877016255e-06,
+      "loss": 0.1639,
+      "step": 14007
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7950257845294156,
+      "learning_rate": 4.03521730165758e-06,
+      "loss": 0.1488,
+      "step": 14008
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8696092538942773,
+      "learning_rate": 4.033895477408604e-06,
+      "loss": 0.1593,
+      "step": 14009
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1425549882601498,
+      "learning_rate": 4.032573814990553e-06,
+      "loss": 0.1876,
+      "step": 14010
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9434310996016758,
+      "learning_rate": 4.03125231443927e-06,
+      "loss": 0.1984,
+      "step": 14011
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0730244990456292,
+      "learning_rate": 4.02993097579061e-06,
+      "loss": 0.181,
+      "step": 14012
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.322629645251612,
+      "learning_rate": 4.028609799080405e-06,
+      "loss": 0.1739,
+      "step": 14013
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5905595740156597,
+      "learning_rate": 4.0272887843445005e-06,
+      "loss": 0.1468,
+      "step": 14014
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4081861558909625,
+      "learning_rate": 4.025967931618722e-06,
+      "loss": 0.1725,
+      "step": 14015
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9231296337146756,
+      "learning_rate": 4.024647240938904e-06,
+      "loss": 0.1671,
+      "step": 14016
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0159252032635526,
+      "learning_rate": 4.0233267123408626e-06,
+      "loss": 0.1688,
+      "step": 14017
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9687980046675957,
+      "learning_rate": 4.022006345860422e-06,
+      "loss": 0.1691,
+      "step": 14018
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.656629955728825,
+      "learning_rate": 4.020686141533401e-06,
+      "loss": 0.1753,
+      "step": 14019
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0112391428165952,
+      "learning_rate": 4.019366099395602e-06,
+      "loss": 0.1519,
+      "step": 14020
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8022414046323324,
+      "learning_rate": 4.01804621948284e-06,
+      "loss": 0.163,
+      "step": 14021
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8621081733595416,
+      "learning_rate": 4.0167265018309075e-06,
+      "loss": 0.162,
+      "step": 14022
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9482586410500872,
+      "learning_rate": 4.01540694647561e-06,
+      "loss": 0.1671,
+      "step": 14023
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8599979611209102,
+      "learning_rate": 4.014087553452734e-06,
+      "loss": 0.1569,
+      "step": 14024
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.224541117890549,
+      "learning_rate": 4.012768322798072e-06,
+      "loss": 0.1801,
+      "step": 14025
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1503432737597172,
+      "learning_rate": 4.011449254547406e-06,
+      "loss": 0.1693,
+      "step": 14026
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8289679623687345,
+      "learning_rate": 4.010130348736522e-06,
+      "loss": 0.1473,
+      "step": 14027
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9223129355750407,
+      "learning_rate": 4.008811605401186e-06,
+      "loss": 0.1674,
+      "step": 14028
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1534141512590803,
+      "learning_rate": 4.007493024577177e-06,
+      "loss": 0.1977,
+      "step": 14029
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8860937852779077,
+      "learning_rate": 4.006174606300255e-06,
+      "loss": 0.1612,
+      "step": 14030
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8518361547364689,
+      "learning_rate": 4.00485635060619e-06,
+      "loss": 0.1679,
+      "step": 14031
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9889257574197703,
+      "learning_rate": 4.0035382575307306e-06,
+      "loss": 0.1586,
+      "step": 14032
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1061626270404616,
+      "learning_rate": 4.0022203271096375e-06,
+      "loss": 0.1824,
+      "step": 14033
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9897553297519481,
+      "learning_rate": 4.000902559378654e-06,
+      "loss": 0.1967,
+      "step": 14034
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8693727300851246,
+      "learning_rate": 3.999584954373528e-06,
+      "loss": 0.1628,
+      "step": 14035
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0355612812061994,
+      "learning_rate": 3.998267512130001e-06,
+      "loss": 0.1976,
+      "step": 14036
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8755614439478262,
+      "learning_rate": 3.996950232683804e-06,
+      "loss": 0.1737,
+      "step": 14037
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8701663683866392,
+      "learning_rate": 3.995633116070675e-06,
+      "loss": 0.1635,
+      "step": 14038
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.364662616775321,
+      "learning_rate": 3.994316162326333e-06,
+      "loss": 0.1468,
+      "step": 14039
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0885834864885342,
+      "learning_rate": 3.992999371486508e-06,
+      "loss": 0.1716,
+      "step": 14040
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1753816593544923,
+      "learning_rate": 3.99168274358691e-06,
+      "loss": 0.1937,
+      "step": 14041
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0768423153042457,
+      "learning_rate": 3.990366278663258e-06,
+      "loss": 0.1579,
+      "step": 14042
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9510173102034292,
+      "learning_rate": 3.989049976751259e-06,
+      "loss": 0.1694,
+      "step": 14043
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0030856247148077,
+      "learning_rate": 3.987733837886622e-06,
+      "loss": 0.1723,
+      "step": 14044
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9019917277961534,
+      "learning_rate": 3.986417862105043e-06,
+      "loss": 0.174,
+      "step": 14045
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9420535261557448,
+      "learning_rate": 3.985102049442221e-06,
+      "loss": 0.1612,
+      "step": 14046
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9970341470186542,
+      "learning_rate": 3.983786399933842e-06,
+      "loss": 0.1801,
+      "step": 14047
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1198941603246906,
+      "learning_rate": 3.982470913615602e-06,
+      "loss": 0.1944,
+      "step": 14048
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.031101344226233,
+      "learning_rate": 3.981155590523173e-06,
+      "loss": 0.1442,
+      "step": 14049
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4822485364807862,
+      "learning_rate": 3.979840430692242e-06,
+      "loss": 0.2225,
+      "step": 14050
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9894481492357582,
+      "learning_rate": 3.978525434158476e-06,
+      "loss": 0.1638,
+      "step": 14051
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0964076308612705,
+      "learning_rate": 3.977210600957548e-06,
+      "loss": 0.1754,
+      "step": 14052
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6665709589666882,
+      "learning_rate": 3.975895931125126e-06,
+      "loss": 0.1697,
+      "step": 14053
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.000045807372957,
+      "learning_rate": 3.9745814246968654e-06,
+      "loss": 0.1786,
+      "step": 14054
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8953795174978475,
+      "learning_rate": 3.973267081708425e-06,
+      "loss": 0.1689,
+      "step": 14055
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.3286514942564485,
+      "learning_rate": 3.971952902195453e-06,
+      "loss": 0.1663,
+      "step": 14056
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1203493039554666,
+      "learning_rate": 3.970638886193603e-06,
+      "loss": 0.1955,
+      "step": 14057
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8888136001531051,
+      "learning_rate": 3.969325033738509e-06,
+      "loss": 0.1964,
+      "step": 14058
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.795777988010517,
+      "learning_rate": 3.968011344865819e-06,
+      "loss": 0.1531,
+      "step": 14059
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7770520518118822,
+      "learning_rate": 3.9666978196111575e-06,
+      "loss": 0.1632,
+      "step": 14060
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8263785189625037,
+      "learning_rate": 3.965384458010157e-06,
+      "loss": 0.1834,
+      "step": 14061
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.854320039391229,
+      "learning_rate": 3.964071260098446e-06,
+      "loss": 0.1508,
+      "step": 14062
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3807512839982112,
+      "learning_rate": 3.962758225911646e-06,
+      "loss": 0.1575,
+      "step": 14063
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1555728871664732,
+      "learning_rate": 3.961445355485366e-06,
+      "loss": 0.1758,
+      "step": 14064
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8123345828753771,
+      "learning_rate": 3.960132648855226e-06,
+      "loss": 0.1758,
+      "step": 14065
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9460660456671849,
+      "learning_rate": 3.958820106056826e-06,
+      "loss": 0.156,
+      "step": 14066
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0004887583666053,
+      "learning_rate": 3.957507727125775e-06,
+      "loss": 0.1608,
+      "step": 14067
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3035912785558115,
+      "learning_rate": 3.956195512097664e-06,
+      "loss": 0.1651,
+      "step": 14068
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0046179848207974,
+      "learning_rate": 3.954883461008091e-06,
+      "loss": 0.1829,
+      "step": 14069
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.990047289247749,
+      "learning_rate": 3.953571573892646e-06,
+      "loss": 0.1758,
+      "step": 14070
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8181189145002394,
+      "learning_rate": 3.9522598507869166e-06,
+      "loss": 0.1539,
+      "step": 14071
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8046299088957247,
+      "learning_rate": 3.95094829172648e-06,
+      "loss": 0.1725,
+      "step": 14072
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5116421607576094,
+      "learning_rate": 3.949636896746911e-06,
+      "loss": 0.1514,
+      "step": 14073
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1942340373605995,
+      "learning_rate": 3.948325665883785e-06,
+      "loss": 0.1796,
+      "step": 14074
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0544986056562093,
+      "learning_rate": 3.947014599172664e-06,
+      "loss": 0.1786,
+      "step": 14075
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.336394823926345,
+      "learning_rate": 3.945703696649117e-06,
+      "loss": 0.1519,
+      "step": 14076
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.944020747222741,
+      "learning_rate": 3.944392958348696e-06,
+      "loss": 0.1733,
+      "step": 14077
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8344168292689214,
+      "learning_rate": 3.943082384306958e-06,
+      "loss": 0.1558,
+      "step": 14078
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9462982215603182,
+      "learning_rate": 3.941771974559453e-06,
+      "loss": 0.1849,
+      "step": 14079
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1478756672396018,
+      "learning_rate": 3.940461729141728e-06,
+      "loss": 0.1716,
+      "step": 14080
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.291699795537521,
+      "learning_rate": 3.939151648089317e-06,
+      "loss": 0.138,
+      "step": 14081
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4757223942660185,
+      "learning_rate": 3.937841731437765e-06,
+      "loss": 0.176,
+      "step": 14082
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9795969693563553,
+      "learning_rate": 3.936531979222593e-06,
+      "loss": 0.1532,
+      "step": 14083
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7695703125447169,
+      "learning_rate": 3.935222391479339e-06,
+      "loss": 0.1525,
+      "step": 14084
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8807274886889117,
+      "learning_rate": 3.933912968243515e-06,
+      "loss": 0.1697,
+      "step": 14085
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.17677908764498,
+      "learning_rate": 3.9326037095506486e-06,
+      "loss": 0.1619,
+      "step": 14086
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8695497729024724,
+      "learning_rate": 3.931294615436245e-06,
+      "loss": 0.1745,
+      "step": 14087
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.015175994237581,
+      "learning_rate": 3.929985685935819e-06,
+      "loss": 0.1685,
+      "step": 14088
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6135670353009806,
+      "learning_rate": 3.928676921084877e-06,
+      "loss": 0.177,
+      "step": 14089
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0624513862242058,
+      "learning_rate": 3.9273683209189115e-06,
+      "loss": 0.1712,
+      "step": 14090
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9178673325700115,
+      "learning_rate": 3.926059885473429e-06,
+      "loss": 0.1617,
+      "step": 14091
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0849630238704142,
+      "learning_rate": 3.9247516147839105e-06,
+      "loss": 0.1735,
+      "step": 14092
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9923777691551956,
+      "learning_rate": 3.923443508885851e-06,
+      "loss": 0.1583,
+      "step": 14093
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2803548963051437,
+      "learning_rate": 3.922135567814726e-06,
+      "loss": 0.1708,
+      "step": 14094
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9130041610458673,
+      "learning_rate": 3.920827791606018e-06,
+      "loss": 0.1616,
+      "step": 14095
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0466901162138669,
+      "learning_rate": 3.919520180295199e-06,
+      "loss": 0.1615,
+      "step": 14096
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0599024632729042,
+      "learning_rate": 3.918212733917742e-06,
+      "loss": 0.1568,
+      "step": 14097
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1164968866332436,
+      "learning_rate": 3.9169054525091045e-06,
+      "loss": 0.1734,
+      "step": 14098
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9097229462116942,
+      "learning_rate": 3.915598336104754e-06,
+      "loss": 0.1647,
+      "step": 14099
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9607652166214575,
+      "learning_rate": 3.914291384740139e-06,
+      "loss": 0.1484,
+      "step": 14100
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1161152426737857,
+      "learning_rate": 3.912984598450716e-06,
+      "loss": 0.1711,
+      "step": 14101
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0838039029190736,
+      "learning_rate": 3.9116779772719274e-06,
+      "loss": 0.1828,
+      "step": 14102
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.396753142997566,
+      "learning_rate": 3.91037152123922e-06,
+      "loss": 0.1637,
+      "step": 14103
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.299828907754022,
+      "learning_rate": 3.9090652303880265e-06,
+      "loss": 0.1842,
+      "step": 14104
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5028157315567607,
+      "learning_rate": 3.907759104753782e-06,
+      "loss": 0.1568,
+      "step": 14105
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1985317318332738,
+      "learning_rate": 3.90645314437192e-06,
+      "loss": 0.1913,
+      "step": 14106
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.251572368953231,
+      "learning_rate": 3.905147349277857e-06,
+      "loss": 0.1689,
+      "step": 14107
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3102618168442373,
+      "learning_rate": 3.9038417195070196e-06,
+      "loss": 0.1952,
+      "step": 14108
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0520810466721868,
+      "learning_rate": 3.902536255094816e-06,
+      "loss": 0.1655,
+      "step": 14109
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7655259367920346,
+      "learning_rate": 3.901230956076665e-06,
+      "loss": 0.165,
+      "step": 14110
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2413659929656236,
+      "learning_rate": 3.899925822487965e-06,
+      "loss": 0.1652,
+      "step": 14111
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9601156709495291,
+      "learning_rate": 3.898620854364126e-06,
+      "loss": 0.182,
+      "step": 14112
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0182150466816644,
+      "learning_rate": 3.897316051740536e-06,
+      "loss": 0.1802,
+      "step": 14113
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9465330640666286,
+      "learning_rate": 3.896011414652593e-06,
+      "loss": 0.1558,
+      "step": 14114
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8495037191629528,
+      "learning_rate": 3.894706943135686e-06,
+      "loss": 0.1619,
+      "step": 14115
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.749177420998462,
+      "learning_rate": 3.893402637225201e-06,
+      "loss": 0.1738,
+      "step": 14116
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0153253780913176,
+      "learning_rate": 3.892098496956511e-06,
+      "loss": 0.1617,
+      "step": 14117
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2844861536349312,
+      "learning_rate": 3.890794522364998e-06,
+      "loss": 0.1844,
+      "step": 14118
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1477401726044283,
+      "learning_rate": 3.8894907134860236e-06,
+      "loss": 0.1761,
+      "step": 14119
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9576356068272893,
+      "learning_rate": 3.888187070354964e-06,
+      "loss": 0.1847,
+      "step": 14120
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8866710011707144,
+      "learning_rate": 3.886883593007171e-06,
+      "loss": 0.1629,
+      "step": 14121
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.853715620160077,
+      "learning_rate": 3.885580281478007e-06,
+      "loss": 0.1636,
+      "step": 14122
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0522413642603576,
+      "learning_rate": 3.8842771358028254e-06,
+      "loss": 0.1828,
+      "step": 14123
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7705310953095824,
+      "learning_rate": 3.882974156016968e-06,
+      "loss": 0.146,
+      "step": 14124
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9935591406214743,
+      "learning_rate": 3.881671342155786e-06,
+      "loss": 0.1927,
+      "step": 14125
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0289400849388903,
+      "learning_rate": 3.880368694254612e-06,
+      "loss": 0.1944,
+      "step": 14126
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.075369092240746,
+      "learning_rate": 3.879066212348786e-06,
+      "loss": 0.1874,
+      "step": 14127
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1412571748621356,
+      "learning_rate": 3.877763896473629e-06,
+      "loss": 0.1736,
+      "step": 14128
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8994203625034884,
+      "learning_rate": 3.876461746664478e-06,
+      "loss": 0.1513,
+      "step": 14129
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8771349317185486,
+      "learning_rate": 3.875159762956644e-06,
+      "loss": 0.178,
+      "step": 14130
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0027658517118203,
+      "learning_rate": 3.873857945385447e-06,
+      "loss": 0.1679,
+      "step": 14131
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.87315016577878,
+      "learning_rate": 3.8725562939862e-06,
+      "loss": 0.1711,
+      "step": 14132
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8900586829521311,
+      "learning_rate": 3.871254808794213e-06,
+      "loss": 0.1691,
+      "step": 14133
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8966203204273094,
+      "learning_rate": 3.869953489844781e-06,
+      "loss": 0.1407,
+      "step": 14134
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.998730627617709,
+      "learning_rate": 3.868652337173211e-06,
+      "loss": 0.1883,
+      "step": 14135
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7719003043359163,
+      "learning_rate": 3.8673513508147885e-06,
+      "loss": 0.161,
+      "step": 14136
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8156765100824103,
+      "learning_rate": 3.866050530804811e-06,
+      "loss": 0.141,
+      "step": 14137
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8042549971576102,
+      "learning_rate": 3.864749877178556e-06,
+      "loss": 0.1637,
+      "step": 14138
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8790603683540212,
+      "learning_rate": 3.86344938997131e-06,
+      "loss": 0.1742,
+      "step": 14139
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.54209637857875,
+      "learning_rate": 3.862149069218343e-06,
+      "loss": 0.176,
+      "step": 14140
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0170286994576827,
+      "learning_rate": 3.8608489149549286e-06,
+      "loss": 0.1591,
+      "step": 14141
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9255760660296493,
+      "learning_rate": 3.8595489272163375e-06,
+      "loss": 0.181,
+      "step": 14142
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9384850020253076,
+      "learning_rate": 3.858249106037826e-06,
+      "loss": 0.1547,
+      "step": 14143
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1219354082933488,
+      "learning_rate": 3.856949451454658e-06,
+      "loss": 0.1847,
+      "step": 14144
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3725720945269748,
+      "learning_rate": 3.855649963502078e-06,
+      "loss": 0.2026,
+      "step": 14145
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.826362384690696,
+      "learning_rate": 3.854350642215344e-06,
+      "loss": 0.2124,
+      "step": 14146
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0420657943593228,
+      "learning_rate": 3.853051487629693e-06,
+      "loss": 0.1555,
+      "step": 14147
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1153510092050685,
+      "learning_rate": 3.851752499780368e-06,
+      "loss": 0.1651,
+      "step": 14148
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9459742492520735,
+      "learning_rate": 3.8504536787026025e-06,
+      "loss": 0.1678,
+      "step": 14149
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0397145526569862,
+      "learning_rate": 3.8491550244316326e-06,
+      "loss": 0.171,
+      "step": 14150
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9655573592566931,
+      "learning_rate": 3.847856537002677e-06,
+      "loss": 0.1736,
+      "step": 14151
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9220548799507059,
+      "learning_rate": 3.846558216450962e-06,
+      "loss": 0.1524,
+      "step": 14152
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0456814200250961,
+      "learning_rate": 3.845260062811701e-06,
+      "loss": 0.1764,
+      "step": 14153
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9412885631888973,
+      "learning_rate": 3.843962076120111e-06,
+      "loss": 0.1722,
+      "step": 14154
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8769782231890678,
+      "learning_rate": 3.842664256411393e-06,
+      "loss": 0.1711,
+      "step": 14155
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9941968030192654,
+      "learning_rate": 3.841366603720761e-06,
+      "loss": 0.1581,
+      "step": 14156
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.81692305819075,
+      "learning_rate": 3.840069118083403e-06,
+      "loss": 0.1729,
+      "step": 14157
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5179566558436612,
+      "learning_rate": 3.838771799534518e-06,
+      "loss": 0.1832,
+      "step": 14158
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9901563215065169,
+      "learning_rate": 3.837474648109298e-06,
+      "loss": 0.1637,
+      "step": 14159
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6951338937784872,
+      "learning_rate": 3.836177663842925e-06,
+      "loss": 0.1659,
+      "step": 14160
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9560269248745517,
+      "learning_rate": 3.834880846770584e-06,
+      "loss": 0.1683,
+      "step": 14161
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1981393142829408,
+      "learning_rate": 3.833584196927443e-06,
+      "loss": 0.1925,
+      "step": 14162
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.737439399910469,
+      "learning_rate": 3.8322877143486835e-06,
+      "loss": 0.1732,
+      "step": 14163
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9970214572536786,
+      "learning_rate": 3.830991399069466e-06,
+      "loss": 0.1662,
+      "step": 14164
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9038070805127413,
+      "learning_rate": 3.829695251124953e-06,
+      "loss": 0.1659,
+      "step": 14165
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8343905841105865,
+      "learning_rate": 3.828399270550306e-06,
+      "loss": 0.1765,
+      "step": 14166
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7514551033421455,
+      "learning_rate": 3.827103457380681e-06,
+      "loss": 0.1569,
+      "step": 14167
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1434878349475175,
+      "learning_rate": 3.82580781165122e-06,
+      "loss": 0.1675,
+      "step": 14168
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0057817293164968,
+      "learning_rate": 3.824512333397073e-06,
+      "loss": 0.185,
+      "step": 14169
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2729546710261415,
+      "learning_rate": 3.823217022653376e-06,
+      "loss": 0.1975,
+      "step": 14170
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.81564027380783,
+      "learning_rate": 3.821921879455268e-06,
+      "loss": 0.1631,
+      "step": 14171
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8976442491961399,
+      "learning_rate": 3.820626903837875e-06,
+      "loss": 0.1569,
+      "step": 14172
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9607703608651237,
+      "learning_rate": 3.81933209583633e-06,
+      "loss": 0.1783,
+      "step": 14173
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.98830563137594,
+      "learning_rate": 3.818037455485748e-06,
+      "loss": 0.195,
+      "step": 14174
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.967548832430743,
+      "learning_rate": 3.816742982821249e-06,
+      "loss": 0.1873,
+      "step": 14175
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9725562983600868,
+      "learning_rate": 3.815448677877949e-06,
+      "loss": 0.1637,
+      "step": 14176
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9534466925619756,
+      "learning_rate": 3.8141545406909486e-06,
+      "loss": 0.2168,
+      "step": 14177
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9022686390988373,
+      "learning_rate": 3.8128605712953606e-06,
+      "loss": 0.1771,
+      "step": 14178
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7895997669461096,
+      "learning_rate": 3.811566769726275e-06,
+      "loss": 0.1788,
+      "step": 14179
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.026042586429769,
+      "learning_rate": 3.810273136018793e-06,
+      "loss": 0.1561,
+      "step": 14180
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7893571675039319,
+      "learning_rate": 3.8089796702079996e-06,
+      "loss": 0.1781,
+      "step": 14181
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6919757359686876,
+      "learning_rate": 3.8076863723289847e-06,
+      "loss": 0.1631,
+      "step": 14182
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.862175568391283,
+      "learning_rate": 3.8063932424168236e-06,
+      "loss": 0.1673,
+      "step": 14183
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8485517752972715,
+      "learning_rate": 3.8051002805065964e-06,
+      "loss": 0.1934,
+      "step": 14184
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9233190973713077,
+      "learning_rate": 3.803807486633373e-06,
+      "loss": 0.1653,
+      "step": 14185
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1319349294113454,
+      "learning_rate": 3.802514860832225e-06,
+      "loss": 0.1909,
+      "step": 14186
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8653577480393039,
+      "learning_rate": 3.8012224031382084e-06,
+      "loss": 0.1779,
+      "step": 14187
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1929076075384082,
+      "learning_rate": 3.7999301135863875e-06,
+      "loss": 0.1731,
+      "step": 14188
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9413376467983017,
+      "learning_rate": 3.7986379922118087e-06,
+      "loss": 0.1551,
+      "step": 14189
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9052220902235042,
+      "learning_rate": 3.797346039049529e-06,
+      "loss": 0.1828,
+      "step": 14190
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9246863882687407,
+      "learning_rate": 3.7960542541345836e-06,
+      "loss": 0.1973,
+      "step": 14191
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8486898931706738,
+      "learning_rate": 3.7947626375020173e-06,
+      "loss": 0.1641,
+      "step": 14192
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2499576012040685,
+      "learning_rate": 3.793471189186869e-06,
+      "loss": 0.1742,
+      "step": 14193
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8500478491305098,
+      "learning_rate": 3.792179909224162e-06,
+      "loss": 0.1686,
+      "step": 14194
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9980383111334427,
+      "learning_rate": 3.7908887976489284e-06,
+      "loss": 0.2013,
+      "step": 14195
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.557717958905487,
+      "learning_rate": 3.789597854496183e-06,
+      "loss": 0.1621,
+      "step": 14196
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.172579101340141,
+      "learning_rate": 3.7883070798009503e-06,
+      "loss": 0.1838,
+      "step": 14197
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5335676877341493,
+      "learning_rate": 3.7870164735982363e-06,
+      "loss": 0.1741,
+      "step": 14198
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.236872824666607,
+      "learning_rate": 3.7857260359230543e-06,
+      "loss": 0.1668,
+      "step": 14199
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0325130618195033,
+      "learning_rate": 3.7844357668104005e-06,
+      "loss": 0.1567,
+      "step": 14200
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1342609581997496,
+      "learning_rate": 3.7831456662952783e-06,
+      "loss": 0.1772,
+      "step": 14201
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4083831164497156,
+      "learning_rate": 3.7818557344126807e-06,
+      "loss": 0.1642,
+      "step": 14202
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8579470681672271,
+      "learning_rate": 3.7805659711976007e-06,
+      "loss": 0.1697,
+      "step": 14203
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0276154408565137,
+      "learning_rate": 3.779276376685017e-06,
+      "loss": 0.1614,
+      "step": 14204
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1292194703813447,
+      "learning_rate": 3.7779869509099166e-06,
+      "loss": 0.1716,
+      "step": 14205
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9358439759067537,
+      "learning_rate": 3.7766976939072673e-06,
+      "loss": 0.173,
+      "step": 14206
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7796355039284816,
+      "learning_rate": 3.7754086057120486e-06,
+      "loss": 0.1449,
+      "step": 14207
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.056734691141821,
+      "learning_rate": 3.77411968635922e-06,
+      "loss": 0.1628,
+      "step": 14208
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9221052651830727,
+      "learning_rate": 3.772830935883749e-06,
+      "loss": 0.1563,
+      "step": 14209
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0631227692867489,
+      "learning_rate": 3.7715423543205875e-06,
+      "loss": 0.1683,
+      "step": 14210
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7410763969252387,
+      "learning_rate": 3.7702539417046923e-06,
+      "loss": 0.1603,
+      "step": 14211
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9452639562381826,
+      "learning_rate": 3.7689656980710132e-06,
+      "loss": 0.189,
+      "step": 14212
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6908879201876998,
+      "learning_rate": 3.7676776234544876e-06,
+      "loss": 0.1638,
+      "step": 14213
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0159422655316006,
+      "learning_rate": 3.7663897178900634e-06,
+      "loss": 0.1612,
+      "step": 14214
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.870214529701722,
+      "learning_rate": 3.7651019814126656e-06,
+      "loss": 0.1632,
+      "step": 14215
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3077317836042965,
+      "learning_rate": 3.763814414057233e-06,
+      "loss": 0.1788,
+      "step": 14216
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0002530570640173,
+      "learning_rate": 3.7625270158586824e-06,
+      "loss": 0.1536,
+      "step": 14217
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3738436004166406,
+      "learning_rate": 3.761239786851939e-06,
+      "loss": 0.1561,
+      "step": 14218
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.354949019454525,
+      "learning_rate": 3.7599527270719183e-06,
+      "loss": 0.1437,
+      "step": 14219
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5472545692675839,
+      "learning_rate": 3.7586658365535367e-06,
+      "loss": 0.1568,
+      "step": 14220
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9656835326636025,
+      "learning_rate": 3.757379115331693e-06,
+      "loss": 0.1802,
+      "step": 14221
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9776349541649698,
+      "learning_rate": 3.756092563441297e-06,
+      "loss": 0.1656,
+      "step": 14222
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8852204987916643,
+      "learning_rate": 3.754806180917239e-06,
+      "loss": 0.1921,
+      "step": 14223
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9167565711720689,
+      "learning_rate": 3.75351996779442e-06,
+      "loss": 0.1775,
+      "step": 14224
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8952438830379588,
+      "learning_rate": 3.752233924107721e-06,
+      "loss": 0.16,
+      "step": 14225
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2955265466290538,
+      "learning_rate": 3.7509480498920325e-06,
+      "loss": 0.1617,
+      "step": 14226
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9378117311047315,
+      "learning_rate": 3.749662345182229e-06,
+      "loss": 0.1589,
+      "step": 14227
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1886155582753781,
+      "learning_rate": 3.7483768100131857e-06,
+      "loss": 0.1915,
+      "step": 14228
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0826107082122094,
+      "learning_rate": 3.7470914444197793e-06,
+      "loss": 0.1709,
+      "step": 14229
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0393811372304311,
+      "learning_rate": 3.745806248436866e-06,
+      "loss": 0.1554,
+      "step": 14230
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.236921987649103,
+      "learning_rate": 3.7445212220993167e-06,
+      "loss": 0.1815,
+      "step": 14231
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2104292988068672,
+      "learning_rate": 3.743236365441978e-06,
+      "loss": 0.2004,
+      "step": 14232
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6439370614882838,
+      "learning_rate": 3.7419516784997102e-06,
+      "loss": 0.1716,
+      "step": 14233
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8458498080848184,
+      "learning_rate": 3.740667161307352e-06,
+      "loss": 0.1576,
+      "step": 14234
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1550376497930346,
+      "learning_rate": 3.7393828138997543e-06,
+      "loss": 0.1706,
+      "step": 14235
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3056922913399747,
+      "learning_rate": 3.7380986363117488e-06,
+      "loss": 0.1802,
+      "step": 14236
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0611836357952285,
+      "learning_rate": 3.7368146285781716e-06,
+      "loss": 0.1553,
+      "step": 14237
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5617880165016602,
+      "learning_rate": 3.73553079073385e-06,
+      "loss": 0.173,
+      "step": 14238
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1187420114648672,
+      "learning_rate": 3.7342471228136148e-06,
+      "loss": 0.1655,
+      "step": 14239
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8656618888307698,
+      "learning_rate": 3.732963624852275e-06,
+      "loss": 0.1645,
+      "step": 14240
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0527586977075543,
+      "learning_rate": 3.7316802968846555e-06,
+      "loss": 0.1905,
+      "step": 14241
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9653891403786222,
+      "learning_rate": 3.7303971389455584e-06,
+      "loss": 0.1785,
+      "step": 14242
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8007572110573549,
+      "learning_rate": 3.7291141510697957e-06,
+      "loss": 0.1631,
+      "step": 14243
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0979067825574946,
+      "learning_rate": 3.7278313332921634e-06,
+      "loss": 0.168,
+      "step": 14244
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0376817166504873,
+      "learning_rate": 3.72654868564746e-06,
+      "loss": 0.2146,
+      "step": 14245
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.870826236426534,
+      "learning_rate": 3.7252662081704806e-06,
+      "loss": 0.1576,
+      "step": 14246
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1024186957206004,
+      "learning_rate": 3.7239839008960066e-06,
+      "loss": 0.1845,
+      "step": 14247
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9461416873285,
+      "learning_rate": 3.722701763858828e-06,
+      "loss": 0.1605,
+      "step": 14248
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8203311355747611,
+      "learning_rate": 3.721419797093715e-06,
+      "loss": 0.1727,
+      "step": 14249
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9325230397827498,
+      "learning_rate": 3.720138000635447e-06,
+      "loss": 0.1751,
+      "step": 14250
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.115454803046617,
+      "learning_rate": 3.718856374518788e-06,
+      "loss": 0.1702,
+      "step": 14251
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8502188295421723,
+      "learning_rate": 3.717574918778507e-06,
+      "loss": 0.1666,
+      "step": 14252
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1228808366873253,
+      "learning_rate": 3.7162936334493594e-06,
+      "loss": 0.1701,
+      "step": 14253
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8689147468875956,
+      "learning_rate": 3.7150125185661e-06,
+      "loss": 0.1704,
+      "step": 14254
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9488898299242475,
+      "learning_rate": 3.7137315741634825e-06,
+      "loss": 0.1882,
+      "step": 14255
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0817092141613454,
+      "learning_rate": 3.7124508002762537e-06,
+      "loss": 0.1599,
+      "step": 14256
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.166908832486912,
+      "learning_rate": 3.711170196939149e-06,
+      "loss": 0.1724,
+      "step": 14257
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8814828284152649,
+      "learning_rate": 3.7098897641869113e-06,
+      "loss": 0.1781,
+      "step": 14258
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9455121000726662,
+      "learning_rate": 3.7086095020542655e-06,
+      "loss": 0.1781,
+      "step": 14259
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1232332362340367,
+      "learning_rate": 3.7073294105759462e-06,
+      "loss": 0.1884,
+      "step": 14260
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8402922663453951,
+      "learning_rate": 3.706049489786667e-06,
+      "loss": 0.1541,
+      "step": 14261
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9859218401322686,
+      "learning_rate": 3.704769739721156e-06,
+      "loss": 0.1641,
+      "step": 14262
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.110810403466431,
+      "learning_rate": 3.703490160414117e-06,
+      "loss": 0.187,
+      "step": 14263
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0070633935527822,
+      "learning_rate": 3.7022107519002635e-06,
+      "loss": 0.1602,
+      "step": 14264
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5144713975719286,
+      "learning_rate": 3.7009315142143033e-06,
+      "loss": 0.1777,
+      "step": 14265
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3375853269180695,
+      "learning_rate": 3.6996524473909268e-06,
+      "loss": 0.1619,
+      "step": 14266
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0919991003667495,
+      "learning_rate": 3.6983735514648376e-06,
+      "loss": 0.174,
+      "step": 14267
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1546100421700132,
+      "learning_rate": 3.697094826470717e-06,
+      "loss": 0.176,
+      "step": 14268
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8533756977482386,
+      "learning_rate": 3.6958162724432612e-06,
+      "loss": 0.1706,
+      "step": 14269
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8114168158613203,
+      "learning_rate": 3.6945378894171392e-06,
+      "loss": 0.174,
+      "step": 14270
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.139833257653544,
+      "learning_rate": 3.6932596774270346e-06,
+      "loss": 0.1579,
+      "step": 14271
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.011143112623799,
+      "learning_rate": 3.6919816365076165e-06,
+      "loss": 0.1912,
+      "step": 14272
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.599161063664042,
+      "learning_rate": 3.6907037666935565e-06,
+      "loss": 0.18,
+      "step": 14273
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.111474925295134,
+      "learning_rate": 3.6894260680195105e-06,
+      "loss": 0.1687,
+      "step": 14274
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.071490665473469,
+      "learning_rate": 3.688148540520141e-06,
+      "loss": 0.1563,
+      "step": 14275
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9001205884583297,
+      "learning_rate": 3.6868711842300964e-06,
+      "loss": 0.1532,
+      "step": 14276
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2590964346673659,
+      "learning_rate": 3.6855939991840305e-06,
+      "loss": 0.1857,
+      "step": 14277
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4732035688485705,
+      "learning_rate": 3.6843169854165807e-06,
+      "loss": 0.176,
+      "step": 14278
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.284662221148433,
+      "learning_rate": 3.683040142962393e-06,
+      "loss": 0.1924,
+      "step": 14279
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2777057044689084,
+      "learning_rate": 3.6817634718560947e-06,
+      "loss": 0.1679,
+      "step": 14280
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9109142054246077,
+      "learning_rate": 3.6804869721323187e-06,
+      "loss": 0.1671,
+      "step": 14281
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0592463944026813,
+      "learning_rate": 3.6792106438256937e-06,
+      "loss": 0.1739,
+      "step": 14282
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.810647659156529,
+      "learning_rate": 3.6779344869708344e-06,
+      "loss": 0.1507,
+      "step": 14283
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9415308598837328,
+      "learning_rate": 3.6766585016023624e-06,
+      "loss": 0.1671,
+      "step": 14284
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1358383503657274,
+      "learning_rate": 3.6753826877548817e-06,
+      "loss": 0.1595,
+      "step": 14285
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9346530761399574,
+      "learning_rate": 3.674107045463007e-06,
+      "loss": 0.191,
+      "step": 14286
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.466131046517711,
+      "learning_rate": 3.672831574761332e-06,
+      "loss": 0.1527,
+      "step": 14287
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2540878334454677,
+      "learning_rate": 3.671556275684458e-06,
+      "loss": 0.1754,
+      "step": 14288
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2123331434905684,
+      "learning_rate": 3.6702811482669776e-06,
+      "loss": 0.1581,
+      "step": 14289
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8083395757328803,
+      "learning_rate": 3.6690061925434817e-06,
+      "loss": 0.1594,
+      "step": 14290
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9007273446819845,
+      "learning_rate": 3.667731408548547e-06,
+      "loss": 0.1777,
+      "step": 14291
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1764483153180103,
+      "learning_rate": 3.6664567963167598e-06,
+      "loss": 0.1855,
+      "step": 14292
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0257463654360448,
+      "learning_rate": 3.6651823558826847e-06,
+      "loss": 0.1775,
+      "step": 14293
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.708378143859542,
+      "learning_rate": 3.6639080872809007e-06,
+      "loss": 0.1595,
+      "step": 14294
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.063857693334117,
+      "learning_rate": 3.662633990545964e-06,
+      "loss": 0.1604,
+      "step": 14295
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.015383341678732,
+      "learning_rate": 3.6613600657124416e-06,
+      "loss": 0.1715,
+      "step": 14296
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2290575016512464,
+      "learning_rate": 3.6600863128148823e-06,
+      "loss": 0.1729,
+      "step": 14297
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.937283228394862,
+      "learning_rate": 3.6588127318878398e-06,
+      "loss": 0.1839,
+      "step": 14298
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0972501980810836,
+      "learning_rate": 3.657539322965863e-06,
+      "loss": 0.151,
+      "step": 14299
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2100849301088383,
+      "learning_rate": 3.6562660860834866e-06,
+      "loss": 0.1771,
+      "step": 14300
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9269535849349061,
+      "learning_rate": 3.654993021275255e-06,
+      "loss": 0.1659,
+      "step": 14301
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.989396120662106,
+      "learning_rate": 3.6537201285756927e-06,
+      "loss": 0.1663,
+      "step": 14302
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9574458380861439,
+      "learning_rate": 3.652447408019334e-06,
+      "loss": 0.1689,
+      "step": 14303
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7094279189074484,
+      "learning_rate": 3.651174859640694e-06,
+      "loss": 0.1878,
+      "step": 14304
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1206309986972018,
+      "learning_rate": 3.6499024834742967e-06,
+      "loss": 0.1781,
+      "step": 14305
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9028885348627,
+      "learning_rate": 3.6486302795546515e-06,
+      "loss": 0.1713,
+      "step": 14306
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9185930338847872,
+      "learning_rate": 3.6473582479162684e-06,
+      "loss": 0.1493,
+      "step": 14307
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8589657424822319,
+      "learning_rate": 3.6460863885936514e-06,
+      "loss": 0.1684,
+      "step": 14308
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9463367078690121,
+      "learning_rate": 3.644814701621303e-06,
+      "loss": 0.1705,
+      "step": 14309
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9896721370691746,
+      "learning_rate": 3.6435431870337123e-06,
+      "loss": 0.1743,
+      "step": 14310
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1614616682775758,
+      "learning_rate": 3.642271844865375e-06,
+      "loss": 0.1781,
+      "step": 14311
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9824090286288907,
+      "learning_rate": 3.641000675150769e-06,
+      "loss": 0.1646,
+      "step": 14312
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7977130453754009,
+      "learning_rate": 3.639729677924382e-06,
+      "loss": 0.1507,
+      "step": 14313
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8735573137115935,
+      "learning_rate": 3.638458853220683e-06,
+      "loss": 0.1643,
+      "step": 14314
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9780952797888448,
+      "learning_rate": 3.637188201074149e-06,
+      "loss": 0.1755,
+      "step": 14315
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9590069219781341,
+      "learning_rate": 3.635917721519245e-06,
+      "loss": 0.1515,
+      "step": 14316
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7895245555191475,
+      "learning_rate": 3.634647414590431e-06,
+      "loss": 0.1827,
+      "step": 14317
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.844476378491929,
+      "learning_rate": 3.6333772803221677e-06,
+      "loss": 0.1677,
+      "step": 14318
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.05269931311854,
+      "learning_rate": 3.632107318748903e-06,
+      "loss": 0.1904,
+      "step": 14319
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.636618686759367,
+      "learning_rate": 3.630837529905089e-06,
+      "loss": 0.1683,
+      "step": 14320
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8617969918168314,
+      "learning_rate": 3.6295679138251637e-06,
+      "loss": 0.1592,
+      "step": 14321
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9486332624925874,
+      "learning_rate": 3.628298470543572e-06,
+      "loss": 0.1892,
+      "step": 14322
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4453375339022392,
+      "learning_rate": 3.6270292000947417e-06,
+      "loss": 0.1761,
+      "step": 14323
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.146975579000541,
+      "learning_rate": 3.625760102513103e-06,
+      "loss": 0.1698,
+      "step": 14324
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6093006977976625,
+      "learning_rate": 3.6244911778330826e-06,
+      "loss": 0.1641,
+      "step": 14325
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1152540194062124,
+      "learning_rate": 3.6232224260891012e-06,
+      "loss": 0.1529,
+      "step": 14326
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8843950514675752,
+      "learning_rate": 3.621953847315569e-06,
+      "loss": 0.1527,
+      "step": 14327
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.096107420351601,
+      "learning_rate": 3.620685441546903e-06,
+      "loss": 0.1987,
+      "step": 14328
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1006170106896094,
+      "learning_rate": 3.6194172088175005e-06,
+      "loss": 0.1886,
+      "step": 14329
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.026145337870011,
+      "learning_rate": 3.6181491491617706e-06,
+      "loss": 0.1496,
+      "step": 14330
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.029899205795693,
+      "learning_rate": 3.6168812626141e-06,
+      "loss": 0.1987,
+      "step": 14331
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0966232569516159,
+      "learning_rate": 3.6156135492088915e-06,
+      "loss": 0.1475,
+      "step": 14332
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1319156136893418,
+      "learning_rate": 3.6143460089805214e-06,
+      "loss": 0.1609,
+      "step": 14333
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3430977218507882,
+      "learning_rate": 3.613078641963377e-06,
+      "loss": 0.1658,
+      "step": 14334
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8104722022322258,
+      "learning_rate": 3.611811448191839e-06,
+      "loss": 0.1578,
+      "step": 14335
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3388835404316106,
+      "learning_rate": 3.610544427700272e-06,
+      "loss": 0.1761,
+      "step": 14336
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8960803836041868,
+      "learning_rate": 3.6092775805230516e-06,
+      "loss": 0.1594,
+      "step": 14337
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8458899384765411,
+      "learning_rate": 3.6080109066945357e-06,
+      "loss": 0.1749,
+      "step": 14338
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8441196825783069,
+      "learning_rate": 3.6067444062490875e-06,
+      "loss": 0.18,
+      "step": 14339
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7732150398842218,
+      "learning_rate": 3.6054780792210542e-06,
+      "loss": 0.1627,
+      "step": 14340
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1086405766696839,
+      "learning_rate": 3.6042119256447904e-06,
+      "loss": 0.1675,
+      "step": 14341
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8720343885082179,
+      "learning_rate": 3.602945945554639e-06,
+      "loss": 0.1747,
+      "step": 14342
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8623553602417626,
+      "learning_rate": 3.6016801389849434e-06,
+      "loss": 0.1829,
+      "step": 14343
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2966082489742936,
+      "learning_rate": 3.6004145059700313e-06,
+      "loss": 0.1676,
+      "step": 14344
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.157181184083306,
+      "learning_rate": 3.5991490465442413e-06,
+      "loss": 0.1957,
+      "step": 14345
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8556719271462281,
+      "learning_rate": 3.5978837607418914e-06,
+      "loss": 0.1431,
+      "step": 14346
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9223430518682103,
+      "learning_rate": 3.5966186485973097e-06,
+      "loss": 0.1711,
+      "step": 14347
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5919130161900463,
+      "learning_rate": 3.5953537101448053e-06,
+      "loss": 0.1598,
+      "step": 14348
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5950698902412257,
+      "learning_rate": 3.5940889454186965e-06,
+      "loss": 0.1786,
+      "step": 14349
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.05794885305146,
+      "learning_rate": 3.5928243544532835e-06,
+      "loss": 0.1447,
+      "step": 14350
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9346209259921906,
+      "learning_rate": 3.5915599372828725e-06,
+      "loss": 0.1607,
+      "step": 14351
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0338066235146515,
+      "learning_rate": 3.590295693941763e-06,
+      "loss": 0.1827,
+      "step": 14352
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0959471715990214,
+      "learning_rate": 3.5890316244642408e-06,
+      "loss": 0.1626,
+      "step": 14353
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.165534101280814,
+      "learning_rate": 3.5877677288846023e-06,
+      "loss": 0.158,
+      "step": 14354
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.947660632419828,
+      "learning_rate": 3.5865040072371228e-06,
+      "loss": 0.1776,
+      "step": 14355
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8461107495185389,
+      "learning_rate": 3.5852404595560876e-06,
+      "loss": 0.1931,
+      "step": 14356
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1027775274177167,
+      "learning_rate": 3.5839770858757627e-06,
+      "loss": 0.1729,
+      "step": 14357
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8846669049565868,
+      "learning_rate": 3.5827138862304266e-06,
+      "loss": 0.1689,
+      "step": 14358
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2012854026999105,
+      "learning_rate": 3.581450860654335e-06,
+      "loss": 0.1838,
+      "step": 14359
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.24401535072047,
+      "learning_rate": 3.580188009181751e-06,
+      "loss": 0.1878,
+      "step": 14360
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4983760604381224,
+      "learning_rate": 3.57892533184693e-06,
+      "loss": 0.1877,
+      "step": 14361
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0436754053513289,
+      "learning_rate": 3.577662828684125e-06,
+      "loss": 0.1759,
+      "step": 14362
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.258726118552222,
+      "learning_rate": 3.576400499727576e-06,
+      "loss": 0.19,
+      "step": 14363
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9488113239079438,
+      "learning_rate": 3.5751383450115298e-06,
+      "loss": 0.1769,
+      "step": 14364
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7623195980440788,
+      "learning_rate": 3.5738763645702145e-06,
+      "loss": 0.1558,
+      "step": 14365
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.855507636654625,
+      "learning_rate": 3.572614558437869e-06,
+      "loss": 0.1645,
+      "step": 14366
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3798623658098415,
+      "learning_rate": 3.5713529266487145e-06,
+      "loss": 0.1922,
+      "step": 14367
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9842413518768752,
+      "learning_rate": 3.5700914692369738e-06,
+      "loss": 0.1842,
+      "step": 14368
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7150008532521748,
+      "learning_rate": 3.568830186236869e-06,
+      "loss": 0.1782,
+      "step": 14369
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4899325092697677,
+      "learning_rate": 3.5675690776826055e-06,
+      "loss": 0.187,
+      "step": 14370
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0130933708959058,
+      "learning_rate": 3.5663081436083967e-06,
+      "loss": 0.1641,
+      "step": 14371
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.913539667921112,
+      "learning_rate": 3.5650473840484402e-06,
+      "loss": 0.1573,
+      "step": 14372
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0861222144561535,
+      "learning_rate": 3.56378679903694e-06,
+      "loss": 0.1931,
+      "step": 14373
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0768559467476906,
+      "learning_rate": 3.562526388608083e-06,
+      "loss": 0.1692,
+      "step": 14374
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1798169488505947,
+      "learning_rate": 3.5612661527960646e-06,
+      "loss": 0.165,
+      "step": 14375
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9430744626611828,
+      "learning_rate": 3.560006091635062e-06,
+      "loss": 0.1657,
+      "step": 14376
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0014914667902954,
+      "learning_rate": 3.558746205159258e-06,
+      "loss": 0.1796,
+      "step": 14377
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9003099615219436,
+      "learning_rate": 3.5574864934028275e-06,
+      "loss": 0.1637,
+      "step": 14378
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1470960761940354,
+      "learning_rate": 3.556226956399943e-06,
+      "loss": 0.1825,
+      "step": 14379
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7413078096043346,
+      "learning_rate": 3.554967594184762e-06,
+      "loss": 0.1853,
+      "step": 14380
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7265083010133702,
+      "learning_rate": 3.553708406791453e-06,
+      "loss": 0.1709,
+      "step": 14381
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1558964925446717,
+      "learning_rate": 3.552449394254165e-06,
+      "loss": 0.1857,
+      "step": 14382
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.263887901976451,
+      "learning_rate": 3.5511905566070537e-06,
+      "loss": 0.16,
+      "step": 14383
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9311776746938797,
+      "learning_rate": 3.549931893884259e-06,
+      "loss": 0.1575,
+      "step": 14384
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8371777695890396,
+      "learning_rate": 3.5486734061199266e-06,
+      "loss": 0.1647,
+      "step": 14385
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9259804564620615,
+      "learning_rate": 3.5474150933481955e-06,
+      "loss": 0.1845,
+      "step": 14386
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4184095276211304,
+      "learning_rate": 3.5461569556031915e-06,
+      "loss": 0.1696,
+      "step": 14387
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7718611178275522,
+      "learning_rate": 3.544898992919048e-06,
+      "loss": 0.1823,
+      "step": 14388
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9339948773745854,
+      "learning_rate": 3.543641205329881e-06,
+      "loss": 0.165,
+      "step": 14389
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7963601667531222,
+      "learning_rate": 3.5423835928698126e-06,
+      "loss": 0.152,
+      "step": 14390
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9641537768577647,
+      "learning_rate": 3.5411261555729513e-06,
+      "loss": 0.1606,
+      "step": 14391
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0070060877029965,
+      "learning_rate": 3.5398688934734125e-06,
+      "loss": 0.175,
+      "step": 14392
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0048912561155683,
+      "learning_rate": 3.53861180660529e-06,
+      "loss": 0.1551,
+      "step": 14393
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0342423921380977,
+      "learning_rate": 3.5373548950026882e-06,
+      "loss": 0.1561,
+      "step": 14394
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1324394346700175,
+      "learning_rate": 3.536098158699699e-06,
+      "loss": 0.1861,
+      "step": 14395
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8373933752901664,
+      "learning_rate": 3.5348415977304165e-06,
+      "loss": 0.148,
+      "step": 14396
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4163575087438194,
+      "learning_rate": 3.5335852121289172e-06,
+      "loss": 0.1624,
+      "step": 14397
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0223524984740415,
+      "learning_rate": 3.5323290019292867e-06,
+      "loss": 0.1646,
+      "step": 14398
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0621961867309728,
+      "learning_rate": 3.531072967165595e-06,
+      "loss": 0.1641,
+      "step": 14399
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0916386681798416,
+      "learning_rate": 3.529817107871918e-06,
+      "loss": 0.2014,
+      "step": 14400
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.125081982430693,
+      "learning_rate": 3.5285614240823128e-06,
+      "loss": 0.1665,
+      "step": 14401
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0002037186679038,
+      "learning_rate": 3.5273059158308487e-06,
+      "loss": 0.1696,
+      "step": 14402
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.114430281565667,
+      "learning_rate": 3.5260505831515736e-06,
+      "loss": 0.1829,
+      "step": 14403
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9266988557304616,
+      "learning_rate": 3.5247954260785422e-06,
+      "loss": 0.15,
+      "step": 14404
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9081102708912858,
+      "learning_rate": 3.523540444645804e-06,
+      "loss": 0.169,
+      "step": 14405
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8531204158047896,
+      "learning_rate": 3.522285638887394e-06,
+      "loss": 0.1861,
+      "step": 14406
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8542913650553401,
+      "learning_rate": 3.5210310088373544e-06,
+      "loss": 0.1799,
+      "step": 14407
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2123191910112574,
+      "learning_rate": 3.5197765545297124e-06,
+      "loss": 0.1628,
+      "step": 14408
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0883468660358517,
+      "learning_rate": 3.5185222759984993e-06,
+      "loss": 0.1844,
+      "step": 14409
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9928291607993566,
+      "learning_rate": 3.5172681732777335e-06,
+      "loss": 0.1583,
+      "step": 14410
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8216944156970188,
+      "learning_rate": 3.5160142464014336e-06,
+      "loss": 0.155,
+      "step": 14411
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0854388217429343,
+      "learning_rate": 3.514760495403614e-06,
+      "loss": 0.1634,
+      "step": 14412
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8440810900533375,
+      "learning_rate": 3.5135069203182858e-06,
+      "loss": 0.1536,
+      "step": 14413
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8591150792858231,
+      "learning_rate": 3.512253521179445e-06,
+      "loss": 0.1564,
+      "step": 14414
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.701393035668696,
+      "learning_rate": 3.511000298021098e-06,
+      "loss": 0.1776,
+      "step": 14415
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0878208543064773,
+      "learning_rate": 3.5097472508772302e-06,
+      "loss": 0.1722,
+      "step": 14416
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9959521739940412,
+      "learning_rate": 3.508494379781838e-06,
+      "loss": 0.1483,
+      "step": 14417
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2945797851190817,
+      "learning_rate": 3.5072416847688993e-06,
+      "loss": 0.1735,
+      "step": 14418
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8609757094301027,
+      "learning_rate": 3.505989165872401e-06,
+      "loss": 0.1569,
+      "step": 14419
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9284580215713893,
+      "learning_rate": 3.504736823126309e-06,
+      "loss": 0.1708,
+      "step": 14420
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8923394895343267,
+      "learning_rate": 3.5034846565645973e-06,
+      "loss": 0.1718,
+      "step": 14421
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4204732071601351,
+      "learning_rate": 3.5022326662212347e-06,
+      "loss": 0.1712,
+      "step": 14422
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4474184116828497,
+      "learning_rate": 3.5009808521301746e-06,
+      "loss": 0.1592,
+      "step": 14423
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8142217738799143,
+      "learning_rate": 3.49972921432538e-06,
+      "loss": 0.1545,
+      "step": 14424
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9384311497554553,
+      "learning_rate": 3.4984777528407944e-06,
+      "loss": 0.1645,
+      "step": 14425
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.947750884547131,
+      "learning_rate": 3.4972264677103694e-06,
+      "loss": 0.1805,
+      "step": 14426
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8440441374202798,
+      "learning_rate": 3.495975358968041e-06,
+      "loss": 0.1713,
+      "step": 14427
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4807376614383012,
+      "learning_rate": 3.4947244266477507e-06,
+      "loss": 0.1677,
+      "step": 14428
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9859634373605151,
+      "learning_rate": 3.493473670783426e-06,
+      "loss": 0.1568,
+      "step": 14429
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8734431242446054,
+      "learning_rate": 3.492223091408994e-06,
+      "loss": 0.1804,
+      "step": 14430
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2104883056005962,
+      "learning_rate": 3.4909726885583782e-06,
+      "loss": 0.1562,
+      "step": 14431
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1914531731251647,
+      "learning_rate": 3.4897224622655e-06,
+      "loss": 0.1766,
+      "step": 14432
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.219156978564906,
+      "learning_rate": 3.4884724125642646e-06,
+      "loss": 0.1753,
+      "step": 14433
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.880056487254269,
+      "learning_rate": 3.487222539488586e-06,
+      "loss": 0.1731,
+      "step": 14434
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9074694159953327,
+      "learning_rate": 3.4859728430723595e-06,
+      "loss": 0.161,
+      "step": 14435
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8519788084340726,
+      "learning_rate": 3.4847233233494916e-06,
+      "loss": 0.1958,
+      "step": 14436
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.034003312618856,
+      "learning_rate": 3.4834739803538686e-06,
+      "loss": 0.1773,
+      "step": 14437
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9177306598389084,
+      "learning_rate": 3.4822248141193816e-06,
+      "loss": 0.1585,
+      "step": 14438
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8999315788165344,
+      "learning_rate": 3.4809758246799173e-06,
+      "loss": 0.166,
+      "step": 14439
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.800926136705932,
+      "learning_rate": 3.479727012069349e-06,
+      "loss": 0.1541,
+      "step": 14440
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5135594002739012,
+      "learning_rate": 3.478478376321558e-06,
+      "loss": 0.1798,
+      "step": 14441
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6799241455933929,
+      "learning_rate": 3.4772299174704048e-06,
+      "loss": 0.1728,
+      "step": 14442
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9344063548247316,
+      "learning_rate": 3.475981635549763e-06,
+      "loss": 0.1697,
+      "step": 14443
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9087585500193363,
+      "learning_rate": 3.4747335305934836e-06,
+      "loss": 0.1703,
+      "step": 14444
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.874986045657428,
+      "learning_rate": 3.47348560263543e-06,
+      "loss": 0.1694,
+      "step": 14445
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8998223181947868,
+      "learning_rate": 3.4722378517094436e-06,
+      "loss": 0.1631,
+      "step": 14446
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.950896245133745,
+      "learning_rate": 3.4709902778493742e-06,
+      "loss": 0.185,
+      "step": 14447
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5040341211529928,
+      "learning_rate": 3.4697428810890634e-06,
+      "loss": 0.1536,
+      "step": 14448
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.430844340350886,
+      "learning_rate": 3.4684956614623476e-06,
+      "loss": 0.1675,
+      "step": 14449
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3461237917573672,
+      "learning_rate": 3.4672486190030543e-06,
+      "loss": 0.1732,
+      "step": 14450
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9382187760505567,
+      "learning_rate": 3.466001753745013e-06,
+      "loss": 0.1739,
+      "step": 14451
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.571179375982628,
+      "learning_rate": 3.4647550657220407e-06,
+      "loss": 0.1646,
+      "step": 14452
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1049373626773231,
+      "learning_rate": 3.463508554967959e-06,
+      "loss": 0.1782,
+      "step": 14453
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.371196296814422,
+      "learning_rate": 3.462262221516575e-06,
+      "loss": 0.1718,
+      "step": 14454
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4900025372462213,
+      "learning_rate": 3.4610160654016987e-06,
+      "loss": 0.1915,
+      "step": 14455
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7389687154100937,
+      "learning_rate": 3.4597700866571294e-06,
+      "loss": 0.1759,
+      "step": 14456
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9413972328791125,
+      "learning_rate": 3.4585242853166657e-06,
+      "loss": 0.1771,
+      "step": 14457
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9685053769263471,
+      "learning_rate": 3.457278661414103e-06,
+      "loss": 0.1851,
+      "step": 14458
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1409849072794902,
+      "learning_rate": 3.456033214983222e-06,
+      "loss": 0.1857,
+      "step": 14459
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1307581819553465,
+      "learning_rate": 3.454787946057814e-06,
+      "loss": 0.1539,
+      "step": 14460
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.105159104219784,
+      "learning_rate": 3.45354285467165e-06,
+      "loss": 0.1663,
+      "step": 14461
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0820023214794368,
+      "learning_rate": 3.452297940858508e-06,
+      "loss": 0.1862,
+      "step": 14462
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7615077258540803,
+      "learning_rate": 3.4510532046521505e-06,
+      "loss": 0.1792,
+      "step": 14463
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0554686371860205,
+      "learning_rate": 3.4498086460863455e-06,
+      "loss": 0.1387,
+      "step": 14464
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9781012241290394,
+      "learning_rate": 3.4485642651948516e-06,
+      "loss": 0.1652,
+      "step": 14465
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.85615413732514,
+      "learning_rate": 3.4473200620114245e-06,
+      "loss": 0.1726,
+      "step": 14466
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0215857868340161,
+      "learning_rate": 3.4460760365698078e-06,
+      "loss": 0.1717,
+      "step": 14467
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1543118477931984,
+      "learning_rate": 3.444832188903752e-06,
+      "loss": 0.1546,
+      "step": 14468
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7340101051610519,
+      "learning_rate": 3.4435885190469886e-06,
+      "loss": 0.1799,
+      "step": 14469
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8684807031919314,
+      "learning_rate": 3.4423450270332626e-06,
+      "loss": 0.1616,
+      "step": 14470
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0967064609498811,
+      "learning_rate": 3.4411017128962932e-06,
+      "loss": 0.1609,
+      "step": 14471
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.902430777464027,
+      "learning_rate": 3.4398585766698146e-06,
+      "loss": 0.1776,
+      "step": 14472
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9113227208746828,
+      "learning_rate": 3.4386156183875384e-06,
+      "loss": 0.1683,
+      "step": 14473
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8535988903463739,
+      "learning_rate": 3.437372838083184e-06,
+      "loss": 0.1612,
+      "step": 14474
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1787012390630862,
+      "learning_rate": 3.4361302357904657e-06,
+      "loss": 0.1583,
+      "step": 14475
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3044864076379246,
+      "learning_rate": 3.4348878115430827e-06,
+      "loss": 0.1637,
+      "step": 14476
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1927482457591698,
+      "learning_rate": 3.4336455653747414e-06,
+      "loss": 0.1669,
+      "step": 14477
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8094448998728605,
+      "learning_rate": 3.432403497319132e-06,
+      "loss": 0.1597,
+      "step": 14478
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0774702938327079,
+      "learning_rate": 3.4311616074099517e-06,
+      "loss": 0.1621,
+      "step": 14479
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9542580107319009,
+      "learning_rate": 3.429919895680881e-06,
+      "loss": 0.1632,
+      "step": 14480
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9611707086909844,
+      "learning_rate": 3.428678362165607e-06,
+      "loss": 0.1443,
+      "step": 14481
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.116401520493127,
+      "learning_rate": 3.4274370068978013e-06,
+      "loss": 0.1587,
+      "step": 14482
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.304449259136624,
+      "learning_rate": 3.4261958299111363e-06,
+      "loss": 0.1763,
+      "step": 14483
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1314091816500065,
+      "learning_rate": 3.424954831239282e-06,
+      "loss": 0.1945,
+      "step": 14484
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0671220035516777,
+      "learning_rate": 3.4237140109159015e-06,
+      "loss": 0.2072,
+      "step": 14485
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1149443924818006,
+      "learning_rate": 3.422473368974648e-06,
+      "loss": 0.1663,
+      "step": 14486
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7938422789239769,
+      "learning_rate": 3.4212329054491775e-06,
+      "loss": 0.1876,
+      "step": 14487
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8671460207650743,
+      "learning_rate": 3.419992620373134e-06,
+      "loss": 0.1636,
+      "step": 14488
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2237197498575294,
+      "learning_rate": 3.418752513780166e-06,
+      "loss": 0.2031,
+      "step": 14489
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7849908273815244,
+      "learning_rate": 3.4175125857039027e-06,
+      "loss": 0.1407,
+      "step": 14490
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1118225123332441,
+      "learning_rate": 3.416272836177984e-06,
+      "loss": 0.1758,
+      "step": 14491
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4168430278920894,
+      "learning_rate": 3.4150332652360386e-06,
+      "loss": 0.169,
+      "step": 14492
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8217197474909118,
+      "learning_rate": 3.413793872911685e-06,
+      "loss": 0.1589,
+      "step": 14493
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0719618045102857,
+      "learning_rate": 3.4125546592385483e-06,
+      "loss": 0.1777,
+      "step": 14494
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0332384870154756,
+      "learning_rate": 3.4113156242502345e-06,
+      "loss": 0.1622,
+      "step": 14495
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4227475903179927,
+      "learning_rate": 3.4100767679803605e-06,
+      "loss": 0.1686,
+      "step": 14496
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2006207316704358,
+      "learning_rate": 3.4088380904625217e-06,
+      "loss": 0.1633,
+      "step": 14497
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0980208062719765,
+      "learning_rate": 3.4075995917303263e-06,
+      "loss": 0.175,
+      "step": 14498
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3096078988390378,
+      "learning_rate": 3.4063612718173613e-06,
+      "loss": 0.1759,
+      "step": 14499
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1581398109088064,
+      "learning_rate": 3.4051231307572187e-06,
+      "loss": 0.1498,
+      "step": 14500
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8573471344318443,
+      "learning_rate": 3.403885168583484e-06,
+      "loss": 0.1721,
+      "step": 14501
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8507531056569723,
+      "learning_rate": 3.4026473853297394e-06,
+      "loss": 0.15,
+      "step": 14502
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1802770854199611,
+      "learning_rate": 3.4014097810295542e-06,
+      "loss": 0.1805,
+      "step": 14503
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.322535646862338,
+      "learning_rate": 3.4001723557165046e-06,
+      "loss": 0.1681,
+      "step": 14504
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8631282651600433,
+      "learning_rate": 3.3989351094241496e-06,
+      "loss": 0.1683,
+      "step": 14505
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.157037757388056,
+      "learning_rate": 3.3976980421860563e-06,
+      "loss": 0.1839,
+      "step": 14506
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4696535664183503,
+      "learning_rate": 3.396461154035772e-06,
+      "loss": 0.1608,
+      "step": 14507
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9476656230693181,
+      "learning_rate": 3.3952244450068527e-06,
+      "loss": 0.1616,
+      "step": 14508
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1449168374763314,
+      "learning_rate": 3.393987915132846e-06,
+      "loss": 0.1748,
+      "step": 14509
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9197974944526394,
+      "learning_rate": 3.3927515644472876e-06,
+      "loss": 0.1768,
+      "step": 14510
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.86307723423347,
+      "learning_rate": 3.3915153929837186e-06,
+      "loss": 0.174,
+      "step": 14511
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9776784552407588,
+      "learning_rate": 3.3902794007756655e-06,
+      "loss": 0.1774,
+      "step": 14512
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0365769962900846,
+      "learning_rate": 3.38904358785666e-06,
+      "loss": 0.1589,
+      "step": 14513
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.917243783214414,
+      "learning_rate": 3.3878079542602172e-06,
+      "loss": 0.1795,
+      "step": 14514
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9496799831594078,
+      "learning_rate": 3.38657250001986e-06,
+      "loss": 0.1648,
+      "step": 14515
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.645315205313346,
+      "learning_rate": 3.3853372251690943e-06,
+      "loss": 0.1571,
+      "step": 14516
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0114640248745363,
+      "learning_rate": 3.38410212974143e-06,
+      "loss": 0.1705,
+      "step": 14517
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1624848313084417,
+      "learning_rate": 3.382867213770369e-06,
+      "loss": 0.1802,
+      "step": 14518
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9069656129089444,
+      "learning_rate": 3.3816324772894116e-06,
+      "loss": 0.1599,
+      "step": 14519
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1188970379680807,
+      "learning_rate": 3.380397920332045e-06,
+      "loss": 0.1778,
+      "step": 14520
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8696674808427313,
+      "learning_rate": 3.3791635429317602e-06,
+      "loss": 0.1655,
+      "step": 14521
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2652170450369873,
+      "learning_rate": 3.377929345122036e-06,
+      "loss": 0.152,
+      "step": 14522
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8047940293282426,
+      "learning_rate": 3.3766953269363555e-06,
+      "loss": 0.1647,
+      "step": 14523
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0025267726557285,
+      "learning_rate": 3.375461488408185e-06,
+      "loss": 0.1743,
+      "step": 14524
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9846755541578239,
+      "learning_rate": 3.3742278295709996e-06,
+      "loss": 0.1684,
+      "step": 14525
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4390956632534884,
+      "learning_rate": 3.372994350458254e-06,
+      "loss": 0.1666,
+      "step": 14526
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.097163868094141,
+      "learning_rate": 3.3717610511034116e-06,
+      "loss": 0.1888,
+      "step": 14527
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4382002789137311,
+      "learning_rate": 3.370527931539929e-06,
+      "loss": 0.156,
+      "step": 14528
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8815583376865935,
+      "learning_rate": 3.3692949918012464e-06,
+      "loss": 0.1724,
+      "step": 14529
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.213918545715467,
+      "learning_rate": 3.3680622319208158e-06,
+      "loss": 0.1833,
+      "step": 14530
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6967342650065715,
+      "learning_rate": 3.3668296519320676e-06,
+      "loss": 0.1699,
+      "step": 14531
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0196662112296344,
+      "learning_rate": 3.3655972518684433e-06,
+      "loss": 0.1757,
+      "step": 14532
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9265938286137334,
+      "learning_rate": 3.3643650317633645e-06,
+      "loss": 0.1703,
+      "step": 14533
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0321780931122735,
+      "learning_rate": 3.36313299165026e-06,
+      "loss": 0.1564,
+      "step": 14534
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3583160837462,
+      "learning_rate": 3.361901131562547e-06,
+      "loss": 0.1786,
+      "step": 14535
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0028582835706312,
+      "learning_rate": 3.3606694515336457e-06,
+      "loss": 0.1837,
+      "step": 14536
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2658375346336457,
+      "learning_rate": 3.3594379515969555e-06,
+      "loss": 0.1654,
+      "step": 14537
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9410506696252773,
+      "learning_rate": 3.3582066317858898e-06,
+      "loss": 0.1474,
+      "step": 14538
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.256438474519618,
+      "learning_rate": 3.3569754921338416e-06,
+      "loss": 0.1898,
+      "step": 14539
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0700580104442847,
+      "learning_rate": 3.355744532674211e-06,
+      "loss": 0.1682,
+      "step": 14540
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9123242811741862,
+      "learning_rate": 3.3545137534403814e-06,
+      "loss": 0.1786,
+      "step": 14541
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9746035844674514,
+      "learning_rate": 3.3532831544657464e-06,
+      "loss": 0.1773,
+      "step": 14542
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.647138127338118,
+      "learning_rate": 3.3520527357836764e-06,
+      "loss": 0.1675,
+      "step": 14543
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9351064187096528,
+      "learning_rate": 3.3508224974275517e-06,
+      "loss": 0.1534,
+      "step": 14544
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.317826242356967,
+      "learning_rate": 3.3495924394307466e-06,
+      "loss": 0.1794,
+      "step": 14545
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8461449815181312,
+      "learning_rate": 3.348362561826618e-06,
+      "loss": 0.1895,
+      "step": 14546
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9737778528397362,
+      "learning_rate": 3.3471328646485345e-06,
+      "loss": 0.1576,
+      "step": 14547
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.890239126549413,
+      "learning_rate": 3.3459033479298444e-06,
+      "loss": 0.1635,
+      "step": 14548
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1067314555766055,
+      "learning_rate": 3.3446740117039045e-06,
+      "loss": 0.1694,
+      "step": 14549
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7909680510638762,
+      "learning_rate": 3.3434448560040544e-06,
+      "loss": 0.1624,
+      "step": 14550
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8687447740217409,
+      "learning_rate": 3.342215880863643e-06,
+      "loss": 0.1763,
+      "step": 14551
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7685163826835121,
+      "learning_rate": 3.3409870863159977e-06,
+      "loss": 0.1599,
+      "step": 14552
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4860196312297997,
+      "learning_rate": 3.3397584723944542e-06,
+      "loss": 0.1447,
+      "step": 14553
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0828676969745032,
+      "learning_rate": 3.3385300391323384e-06,
+      "loss": 0.1508,
+      "step": 14554
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9196225158622006,
+      "learning_rate": 3.3373017865629742e-06,
+      "loss": 0.1647,
+      "step": 14555
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1366321195647922,
+      "learning_rate": 3.336073714719673e-06,
+      "loss": 0.1741,
+      "step": 14556
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9609351328524763,
+      "learning_rate": 3.3348458236357517e-06,
+      "loss": 0.1374,
+      "step": 14557
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5918165884053737,
+      "learning_rate": 3.33361811334451e-06,
+      "loss": 0.1787,
+      "step": 14558
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6087642687901853,
+      "learning_rate": 3.332390583879257e-06,
+      "loss": 0.1831,
+      "step": 14559
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.954159681884898,
+      "learning_rate": 3.3311632352732826e-06,
+      "loss": 0.1633,
+      "step": 14560
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2349214257106662,
+      "learning_rate": 3.3299360675598826e-06,
+      "loss": 0.1772,
+      "step": 14561
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3786250169409255,
+      "learning_rate": 3.3287090807723466e-06,
+      "loss": 0.1524,
+      "step": 14562
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9314633192128212,
+      "learning_rate": 3.3274822749439506e-06,
+      "loss": 0.1697,
+      "step": 14563
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3280040060596827,
+      "learning_rate": 3.3262556501079777e-06,
+      "loss": 0.1545,
+      "step": 14564
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7767872809005165,
+      "learning_rate": 3.325029206297694e-06,
+      "loss": 0.1468,
+      "step": 14565
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.264531729338008,
+      "learning_rate": 3.3238029435463727e-06,
+      "loss": 0.1636,
+      "step": 14566
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2727697010218713,
+      "learning_rate": 3.3225768618872712e-06,
+      "loss": 0.1635,
+      "step": 14567
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0529185194920503,
+      "learning_rate": 3.321350961353652e-06,
+      "loss": 0.1579,
+      "step": 14568
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.050110099196728,
+      "learning_rate": 3.320125241978762e-06,
+      "loss": 0.1659,
+      "step": 14569
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.080844483099251,
+      "learning_rate": 3.3188997037958535e-06,
+      "loss": 0.1584,
+      "step": 14570
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0033168561510555,
+      "learning_rate": 3.3176743468381665e-06,
+      "loss": 0.1654,
+      "step": 14571
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4874031465916626,
+      "learning_rate": 3.3164491711389434e-06,
+      "loss": 0.1753,
+      "step": 14572
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5101430635489643,
+      "learning_rate": 3.3152241767314117e-06,
+      "loss": 0.1632,
+      "step": 14573
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7342798075701323,
+      "learning_rate": 3.3139993636488042e-06,
+      "loss": 0.1567,
+      "step": 14574
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.040576258566937,
+      "learning_rate": 3.3127747319243385e-06,
+      "loss": 0.1759,
+      "step": 14575
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0946480073485154,
+      "learning_rate": 3.31155028159124e-06,
+      "loss": 0.1689,
+      "step": 14576
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0050106610475777,
+      "learning_rate": 3.3103260126827143e-06,
+      "loss": 0.185,
+      "step": 14577
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.086747451695257,
+      "learning_rate": 3.3091019252319755e-06,
+      "loss": 0.1706,
+      "step": 14578
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.041621202794767,
+      "learning_rate": 3.3078780192722225e-06,
+      "loss": 0.1531,
+      "step": 14579
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0667084003881273,
+      "learning_rate": 3.3066542948366564e-06,
+      "loss": 0.1624,
+      "step": 14580
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0626094125015544,
+      "learning_rate": 3.3054307519584737e-06,
+      "loss": 0.1639,
+      "step": 14581
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9810072837276894,
+      "learning_rate": 3.304207390670856e-06,
+      "loss": 0.1596,
+      "step": 14582
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2386279164336178,
+      "learning_rate": 3.302984211006995e-06,
+      "loss": 0.175,
+      "step": 14583
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2024887775229773,
+      "learning_rate": 3.301761213000062e-06,
+      "loss": 0.1549,
+      "step": 14584
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.162501635769063,
+      "learning_rate": 3.3005383966832383e-06,
+      "loss": 0.1699,
+      "step": 14585
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0021653115090114,
+      "learning_rate": 3.2993157620896844e-06,
+      "loss": 0.1644,
+      "step": 14586
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9026003490311508,
+      "learning_rate": 3.2980933092525704e-06,
+      "loss": 0.1762,
+      "step": 14587
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0016934769153616,
+      "learning_rate": 3.296871038205053e-06,
+      "loss": 0.1795,
+      "step": 14588
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1099327248410022,
+      "learning_rate": 3.2956489489802902e-06,
+      "loss": 0.176,
+      "step": 14589
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9360943590709028,
+      "learning_rate": 3.2944270416114256e-06,
+      "loss": 0.1513,
+      "step": 14590
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0377900767546822,
+      "learning_rate": 3.29320531613161e-06,
+      "loss": 0.1851,
+      "step": 14591
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1138509631808082,
+      "learning_rate": 3.2919837725739745e-06,
+      "loss": 0.1656,
+      "step": 14592
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8037614164539251,
+      "learning_rate": 3.290762410971663e-06,
+      "loss": 0.1843,
+      "step": 14593
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1417182624558486,
+      "learning_rate": 3.289541231357796e-06,
+      "loss": 0.1875,
+      "step": 14594
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1424511213698094,
+      "learning_rate": 3.288320233765504e-06,
+      "loss": 0.1674,
+      "step": 14595
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8267506888422055,
+      "learning_rate": 3.2870994182279036e-06,
+      "loss": 0.1669,
+      "step": 14596
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.109347917510044,
+      "learning_rate": 3.2858787847781093e-06,
+      "loss": 0.1737,
+      "step": 14597
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4723720577426307,
+      "learning_rate": 3.2846583334492354e-06,
+      "loss": 0.1934,
+      "step": 14598
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4229642799860578,
+      "learning_rate": 3.2834380642743813e-06,
+      "loss": 0.183,
+      "step": 14599
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8882285331677954,
+      "learning_rate": 3.282217977286651e-06,
+      "loss": 0.1626,
+      "step": 14600
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.040414192870245,
+      "learning_rate": 3.280998072519135e-06,
+      "loss": 0.1632,
+      "step": 14601
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5818449754137551,
+      "learning_rate": 3.2797783500049297e-06,
+      "loss": 0.1594,
+      "step": 14602
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0808796494097017,
+      "learning_rate": 3.2785588097771115e-06,
+      "loss": 0.1792,
+      "step": 14603
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8474299748024428,
+      "learning_rate": 3.277339451868766e-06,
+      "loss": 0.1518,
+      "step": 14604
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9284468258094691,
+      "learning_rate": 3.2761202763129674e-06,
+      "loss": 0.1717,
+      "step": 14605
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0873153366497863,
+      "learning_rate": 3.2749012831427897e-06,
+      "loss": 0.189,
+      "step": 14606
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.30278093395129,
+      "learning_rate": 3.2736824723912907e-06,
+      "loss": 0.1631,
+      "step": 14607
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0590908505037824,
+      "learning_rate": 3.272463844091538e-06,
+      "loss": 0.1829,
+      "step": 14608
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9810362215460501,
+      "learning_rate": 3.2712453982765802e-06,
+      "loss": 0.1513,
+      "step": 14609
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8938383648760326,
+      "learning_rate": 3.270027134979474e-06,
+      "loss": 0.1808,
+      "step": 14610
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5913105125941047,
+      "learning_rate": 3.2688090542332597e-06,
+      "loss": 0.1599,
+      "step": 14611
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8290832642919869,
+      "learning_rate": 3.2675911560709826e-06,
+      "loss": 0.1719,
+      "step": 14612
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0200858042748173,
+      "learning_rate": 3.266373440525672e-06,
+      "loss": 0.1687,
+      "step": 14613
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9597410293776677,
+      "learning_rate": 3.265155907630363e-06,
+      "loss": 0.1932,
+      "step": 14614
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9905982909574356,
+      "learning_rate": 3.2639385574180825e-06,
+      "loss": 0.1726,
+      "step": 14615
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0516067557843038,
+      "learning_rate": 3.262721389921847e-06,
+      "loss": 0.1368,
+      "step": 14616
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.488569177243947,
+      "learning_rate": 3.261504405174677e-06,
+      "loss": 0.1905,
+      "step": 14617
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0496740397502935,
+      "learning_rate": 3.2602876032095767e-06,
+      "loss": 0.2094,
+      "step": 14618
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2266278443699472,
+      "learning_rate": 3.2590709840595604e-06,
+      "loss": 0.1694,
+      "step": 14619
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.8910979222441915,
+      "learning_rate": 3.2578545477576195e-06,
+      "loss": 0.1935,
+      "step": 14620
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1269281838780711,
+      "learning_rate": 3.256638294336759e-06,
+      "loss": 0.1769,
+      "step": 14621
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.469746087556023,
+      "learning_rate": 3.2554222238299627e-06,
+      "loss": 0.1743,
+      "step": 14622
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5128575182766357,
+      "learning_rate": 3.2542063362702194e-06,
+      "loss": 0.1768,
+      "step": 14623
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9485174659917036,
+      "learning_rate": 3.25299063169051e-06,
+      "loss": 0.184,
+      "step": 14624
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.819632029538112,
+      "learning_rate": 3.251775110123814e-06,
+      "loss": 0.1994,
+      "step": 14625
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1297795644354593,
+      "learning_rate": 3.250559771603097e-06,
+      "loss": 0.1556,
+      "step": 14626
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5246558499353544,
+      "learning_rate": 3.2493446161613297e-06,
+      "loss": 0.1624,
+      "step": 14627
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1439968605434216,
+      "learning_rate": 3.248129643831467e-06,
+      "loss": 0.1548,
+      "step": 14628
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9945889104991872,
+      "learning_rate": 3.2469148546464734e-06,
+      "loss": 0.1439,
+      "step": 14629
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3193732383010004,
+      "learning_rate": 3.245700248639292e-06,
+      "loss": 0.1741,
+      "step": 14630
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8404263760961086,
+      "learning_rate": 3.2444858258428733e-06,
+      "loss": 0.1655,
+      "step": 14631
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9678957974759184,
+      "learning_rate": 3.243271586290161e-06,
+      "loss": 0.1731,
+      "step": 14632
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.043645278047243,
+      "learning_rate": 3.2420575300140855e-06,
+      "loss": 0.1608,
+      "step": 14633
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1111114593266385,
+      "learning_rate": 3.2408436570475844e-06,
+      "loss": 0.1684,
+      "step": 14634
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.237932636905411,
+      "learning_rate": 3.2396299674235777e-06,
+      "loss": 0.1937,
+      "step": 14635
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9429836293705041,
+      "learning_rate": 3.2384164611749924e-06,
+      "loss": 0.176,
+      "step": 14636
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7866774192846304,
+      "learning_rate": 3.23720313833474e-06,
+      "loss": 0.1658,
+      "step": 14637
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1895483659518495,
+      "learning_rate": 3.235989998935738e-06,
+      "loss": 0.1748,
+      "step": 14638
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0754191928378574,
+      "learning_rate": 3.234777043010886e-06,
+      "loss": 0.175,
+      "step": 14639
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2658732859500361,
+      "learning_rate": 3.2335642705930883e-06,
+      "loss": 0.1662,
+      "step": 14640
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0116684018540003,
+      "learning_rate": 3.2323516817152424e-06,
+      "loss": 0.1697,
+      "step": 14641
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2918276885913262,
+      "learning_rate": 3.2311392764102422e-06,
+      "loss": 0.1726,
+      "step": 14642
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8766387636913678,
+      "learning_rate": 3.2299270547109684e-06,
+      "loss": 0.1728,
+      "step": 14643
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.032418759945567,
+      "learning_rate": 3.228715016650308e-06,
+      "loss": 0.1813,
+      "step": 14644
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.966008159927298,
+      "learning_rate": 3.227503162261133e-06,
+      "loss": 0.157,
+      "step": 14645
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2640327901478297,
+      "learning_rate": 3.22629149157632e-06,
+      "loss": 0.1604,
+      "step": 14646
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5646018269615116,
+      "learning_rate": 3.2250800046287303e-06,
+      "loss": 0.1951,
+      "step": 14647
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1532003300998375,
+      "learning_rate": 3.2238687014512306e-06,
+      "loss": 0.1482,
+      "step": 14648
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0329973344003411,
+      "learning_rate": 3.2226575820766727e-06,
+      "loss": 0.1594,
+      "step": 14649
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8869975421624843,
+      "learning_rate": 3.22144664653791e-06,
+      "loss": 0.1643,
+      "step": 14650
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0117567487631582,
+      "learning_rate": 3.220235894867794e-06,
+      "loss": 0.1664,
+      "step": 14651
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8690142384583647,
+      "learning_rate": 3.219025327099158e-06,
+      "loss": 0.1813,
+      "step": 14652
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1503442008963496,
+      "learning_rate": 3.2178149432648465e-06,
+      "loss": 0.1901,
+      "step": 14653
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.072701341953258,
+      "learning_rate": 3.2166047433976843e-06,
+      "loss": 0.1462,
+      "step": 14654
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8609032986106758,
+      "learning_rate": 3.215394727530504e-06,
+      "loss": 0.1581,
+      "step": 14655
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8603901676502527,
+      "learning_rate": 3.214184895696123e-06,
+      "loss": 0.176,
+      "step": 14656
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8059552025787837,
+      "learning_rate": 3.21297524792736e-06,
+      "loss": 0.162,
+      "step": 14657
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9996015945561221,
+      "learning_rate": 3.211765784257026e-06,
+      "loss": 0.1677,
+      "step": 14658
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2233505941974974,
+      "learning_rate": 3.210556504717932e-06,
+      "loss": 0.1809,
+      "step": 14659
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0029193160617993,
+      "learning_rate": 3.2093474093428733e-06,
+      "loss": 0.1721,
+      "step": 14660
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1138558257840987,
+      "learning_rate": 3.2081384981646534e-06,
+      "loss": 0.1602,
+      "step": 14661
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8493704927249281,
+      "learning_rate": 3.2069297712160563e-06,
+      "loss": 0.1696,
+      "step": 14662
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9300782059342856,
+      "learning_rate": 3.2057212285298767e-06,
+      "loss": 0.1746,
+      "step": 14663
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6489511279917344,
+      "learning_rate": 3.2045128701388883e-06,
+      "loss": 0.1831,
+      "step": 14664
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6715377433329404,
+      "learning_rate": 3.2033046960758763e-06,
+      "loss": 0.148,
+      "step": 14665
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0328666976068182,
+      "learning_rate": 3.202096706373604e-06,
+      "loss": 0.1697,
+      "step": 14666
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.553987915627782,
+      "learning_rate": 3.2008889010648438e-06,
+      "loss": 0.1515,
+      "step": 14667
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7387038180708394,
+      "learning_rate": 3.1996812801823585e-06,
+      "loss": 0.1754,
+      "step": 14668
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0705040882814145,
+      "learning_rate": 3.1984738437588992e-06,
+      "loss": 0.1509,
+      "step": 14669
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6639713788036837,
+      "learning_rate": 3.197266591827225e-06,
+      "loss": 0.1562,
+      "step": 14670
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0838213880179155,
+      "learning_rate": 3.1960595244200745e-06,
+      "loss": 0.1564,
+      "step": 14671
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1254132382331623,
+      "learning_rate": 3.1948526415701973e-06,
+      "loss": 0.1587,
+      "step": 14672
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1114188821514774,
+      "learning_rate": 3.1936459433103238e-06,
+      "loss": 0.1504,
+      "step": 14673
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8106975038011655,
+      "learning_rate": 3.192439429673192e-06,
+      "loss": 0.1733,
+      "step": 14674
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.287735468603681,
+      "learning_rate": 3.191233100691521e-06,
+      "loss": 0.1759,
+      "step": 14675
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9826916736715666,
+      "learning_rate": 3.1900269563980368e-06,
+      "loss": 0.1738,
+      "step": 14676
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9326441042038448,
+      "learning_rate": 3.1888209968254567e-06,
+      "loss": 0.1624,
+      "step": 14677
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.230860934384176,
+      "learning_rate": 3.1876152220064948e-06,
+      "loss": 0.1465,
+      "step": 14678
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.252938672283592,
+      "learning_rate": 3.186409631973851e-06,
+      "loss": 0.1899,
+      "step": 14679
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9513008216253451,
+      "learning_rate": 3.1852042267602344e-06,
+      "loss": 0.169,
+      "step": 14680
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3737216365826161,
+      "learning_rate": 3.183999006398335e-06,
+      "loss": 0.1629,
+      "step": 14681
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0258943551734003,
+      "learning_rate": 3.1827939709208512e-06,
+      "loss": 0.1769,
+      "step": 14682
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1094208877179543,
+      "learning_rate": 3.181589120360462e-06,
+      "loss": 0.1736,
+      "step": 14683
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8428413399334154,
+      "learning_rate": 3.180384454749853e-06,
+      "loss": 0.1562,
+      "step": 14684
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0872615303852258,
+      "learning_rate": 3.1791799741217046e-06,
+      "loss": 0.1958,
+      "step": 14685
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4317740619625547,
+      "learning_rate": 3.17797567850868e-06,
+      "loss": 0.1848,
+      "step": 14686
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2333218938900938,
+      "learning_rate": 3.176771567943455e-06,
+      "loss": 0.1678,
+      "step": 14687
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.200898823296642,
+      "learning_rate": 3.1755676424586835e-06,
+      "loss": 0.1623,
+      "step": 14688
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0968842840505326,
+      "learning_rate": 3.1743639020870277e-06,
+      "loss": 0.164,
+      "step": 14689
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1702189181308977,
+      "learning_rate": 3.173160346861134e-06,
+      "loss": 0.1793,
+      "step": 14690
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7023827975981258,
+      "learning_rate": 3.1719569768136536e-06,
+      "loss": 0.1598,
+      "step": 14691
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1005884174638634,
+      "learning_rate": 3.1707537919772236e-06,
+      "loss": 0.1612,
+      "step": 14692
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8893579964151115,
+      "learning_rate": 3.1695507923844813e-06,
+      "loss": 0.1508,
+      "step": 14693
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.226655951675619,
+      "learning_rate": 3.1683479780680616e-06,
+      "loss": 0.1635,
+      "step": 14694
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.947454141417125,
+      "learning_rate": 3.1671453490605906e-06,
+      "loss": 0.1847,
+      "step": 14695
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0113457917496576,
+      "learning_rate": 3.1659429053946853e-06,
+      "loss": 0.1553,
+      "step": 14696
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.088003262151018,
+      "learning_rate": 3.1647406471029684e-06,
+      "loss": 0.1665,
+      "step": 14697
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1241668611924707,
+      "learning_rate": 3.1635385742180435e-06,
+      "loss": 0.1697,
+      "step": 14698
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8074782348500469,
+      "learning_rate": 3.1623366867725238e-06,
+      "loss": 0.1667,
+      "step": 14699
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9719281670469927,
+      "learning_rate": 3.1611349847990083e-06,
+      "loss": 0.1702,
+      "step": 14700
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.115648697125075,
+      "learning_rate": 3.15993346833009e-06,
+      "loss": 0.1717,
+      "step": 14701
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1447856245133403,
+      "learning_rate": 3.1587321373983616e-06,
+      "loss": 0.1681,
+      "step": 14702
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1970897169384256,
+      "learning_rate": 3.1575309920364106e-06,
+      "loss": 0.1498,
+      "step": 14703
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4211130522484572,
+      "learning_rate": 3.156330032276821e-06,
+      "loss": 0.1797,
+      "step": 14704
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.123241466425086,
+      "learning_rate": 3.1551292581521632e-06,
+      "loss": 0.1764,
+      "step": 14705
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1792557421404262,
+      "learning_rate": 3.1539286696950135e-06,
+      "loss": 0.1903,
+      "step": 14706
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0185275875438167,
+      "learning_rate": 3.152728266937932e-06,
+      "loss": 0.1541,
+      "step": 14707
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1295736663003781,
+      "learning_rate": 3.151528049913487e-06,
+      "loss": 0.183,
+      "step": 14708
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0700529459342265,
+      "learning_rate": 3.150328018654226e-06,
+      "loss": 0.1689,
+      "step": 14709
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.331060876225935,
+      "learning_rate": 3.149128173192706e-06,
+      "loss": 0.1866,
+      "step": 14710
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.011138547228524,
+      "learning_rate": 3.14792851356147e-06,
+      "loss": 0.1602,
+      "step": 14711
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8101396967808473,
+      "learning_rate": 3.1467290397930637e-06,
+      "loss": 0.1807,
+      "step": 14712
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8430158304766958,
+      "learning_rate": 3.1455297519200157e-06,
+      "loss": 0.1553,
+      "step": 14713
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0986957333952236,
+      "learning_rate": 3.144330649974864e-06,
+      "loss": 0.172,
+      "step": 14714
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9840712424089405,
+      "learning_rate": 3.1431317339901267e-06,
+      "loss": 0.1609,
+      "step": 14715
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1991225020841343,
+      "learning_rate": 3.1419330039983333e-06,
+      "loss": 0.1853,
+      "step": 14716
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1482875122879654,
+      "learning_rate": 3.14073446003199e-06,
+      "loss": 0.1649,
+      "step": 14717
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1271257346502312,
+      "learning_rate": 3.1395361021236148e-06,
+      "loss": 0.1825,
+      "step": 14718
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0659969335638813,
+      "learning_rate": 3.1383379303057084e-06,
+      "loss": 0.1443,
+      "step": 14719
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0191410707488946,
+      "learning_rate": 3.137139944610772e-06,
+      "loss": 0.1733,
+      "step": 14720
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.76103460182858,
+      "learning_rate": 3.1359421450713056e-06,
+      "loss": 0.1548,
+      "step": 14721
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8902470007480779,
+      "learning_rate": 3.1347445317197935e-06,
+      "loss": 0.1549,
+      "step": 14722
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2089112463254226,
+      "learning_rate": 3.1335471045887255e-06,
+      "loss": 0.175,
+      "step": 14723
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.973603976260908,
+      "learning_rate": 3.1323498637105787e-06,
+      "loss": 0.1607,
+      "step": 14724
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9537143962114913,
+      "learning_rate": 3.1311528091178324e-06,
+      "loss": 0.149,
+      "step": 14725
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9696065408943665,
+      "learning_rate": 3.12995594084295e-06,
+      "loss": 0.1678,
+      "step": 14726
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9559168981759395,
+      "learning_rate": 3.1287592589184025e-06,
+      "loss": 0.139,
+      "step": 14727
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9479267688834492,
+      "learning_rate": 3.127562763376647e-06,
+      "loss": 0.194,
+      "step": 14728
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9450990399734578,
+      "learning_rate": 3.1263664542501427e-06,
+      "loss": 0.2034,
+      "step": 14729
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.861102551017935,
+      "learning_rate": 3.1251703315713333e-06,
+      "loss": 0.1589,
+      "step": 14730
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0664891017098699,
+      "learning_rate": 3.123974395372671e-06,
+      "loss": 0.1531,
+      "step": 14731
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0335841585823917,
+      "learning_rate": 3.1227786456865883e-06,
+      "loss": 0.1716,
+      "step": 14732
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.00491646111731,
+      "learning_rate": 3.121583082545526e-06,
+      "loss": 0.1683,
+      "step": 14733
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.206580417310946,
+      "learning_rate": 3.1203877059819077e-06,
+      "loss": 0.1858,
+      "step": 14734
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1377213446369543,
+      "learning_rate": 3.1191925160281644e-06,
+      "loss": 0.1572,
+      "step": 14735
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.539961066898607,
+      "learning_rate": 3.1179975127167105e-06,
+      "loss": 0.1629,
+      "step": 14736
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9933739792302864,
+      "learning_rate": 3.1168026960799624e-06,
+      "loss": 0.1745,
+      "step": 14737
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.090709860866361,
+      "learning_rate": 3.115608066150333e-06,
+      "loss": 0.1733,
+      "step": 14738
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1758233601106887,
+      "learning_rate": 3.1144136229602205e-06,
+      "loss": 0.1798,
+      "step": 14739
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.906908514722467,
+      "learning_rate": 3.1132193665420306e-06,
+      "loss": 0.1616,
+      "step": 14740
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0543997428468896,
+      "learning_rate": 3.112025296928152e-06,
+      "loss": 0.178,
+      "step": 14741
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1950618467976908,
+      "learning_rate": 3.110831414150978e-06,
+      "loss": 0.184,
+      "step": 14742
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2784627442075167,
+      "learning_rate": 3.1096377182428885e-06,
+      "loss": 0.1458,
+      "step": 14743
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.032175497273516,
+      "learning_rate": 3.1084442092362675e-06,
+      "loss": 0.1568,
+      "step": 14744
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1464911325739189,
+      "learning_rate": 3.1072508871634843e-06,
+      "loss": 0.1475,
+      "step": 14745
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1112318755193693,
+      "learning_rate": 3.1060577520569103e-06,
+      "loss": 0.1764,
+      "step": 14746
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.084391657633532,
+      "learning_rate": 3.104864803948908e-06,
+      "loss": 0.1764,
+      "step": 14747
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0758458176930714,
+      "learning_rate": 3.1036720428718413e-06,
+      "loss": 0.1891,
+      "step": 14748
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1294542286974087,
+      "learning_rate": 3.1024794688580562e-06,
+      "loss": 0.176,
+      "step": 14749
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9721057798346413,
+      "learning_rate": 3.1012870819399087e-06,
+      "loss": 0.1683,
+      "step": 14750
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0807184805474042,
+      "learning_rate": 3.1000948821497356e-06,
+      "loss": 0.1756,
+      "step": 14751
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8688973454793225,
+      "learning_rate": 3.098902869519882e-06,
+      "loss": 0.1598,
+      "step": 14752
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6859639098812923,
+      "learning_rate": 3.097711044082674e-06,
+      "loss": 0.1586,
+      "step": 14753
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9853578787351401,
+      "learning_rate": 3.096519405870444e-06,
+      "loss": 0.1573,
+      "step": 14754
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9852364539415878,
+      "learning_rate": 3.095327954915519e-06,
+      "loss": 0.1824,
+      "step": 14755
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9889064845105907,
+      "learning_rate": 3.0941366912502092e-06,
+      "loss": 0.1829,
+      "step": 14756
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2036898306677362,
+      "learning_rate": 3.092945614906835e-06,
+      "loss": 0.1591,
+      "step": 14757
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0620697656612537,
+      "learning_rate": 3.0917547259176973e-06,
+      "loss": 0.1856,
+      "step": 14758
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2864163935068007,
+      "learning_rate": 3.090564024315107e-06,
+      "loss": 0.1736,
+      "step": 14759
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.189074879260085,
+      "learning_rate": 3.089373510131354e-06,
+      "loss": 0.1811,
+      "step": 14760
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.345609459317833,
+      "learning_rate": 3.0881831833987387e-06,
+      "loss": 0.1497,
+      "step": 14761
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8882311019034608,
+      "learning_rate": 3.086993044149541e-06,
+      "loss": 0.1618,
+      "step": 14762
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0149783463440294,
+      "learning_rate": 3.0858030924160477e-06,
+      "loss": 0.1901,
+      "step": 14763
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8790084933384287,
+      "learning_rate": 3.084613328230536e-06,
+      "loss": 0.1813,
+      "step": 14764
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8785959266232123,
+      "learning_rate": 3.0834237516252817e-06,
+      "loss": 0.1542,
+      "step": 14765
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1549721382274187,
+      "learning_rate": 3.082234362632546e-06,
+      "loss": 0.1645,
+      "step": 14766
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9054447399940857,
+      "learning_rate": 3.081045161284596e-06,
+      "loss": 0.1561,
+      "step": 14767
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0804171216263059,
+      "learning_rate": 3.0798561476136845e-06,
+      "loss": 0.1687,
+      "step": 14768
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0504575861996284,
+      "learning_rate": 3.078667321652069e-06,
+      "loss": 0.1494,
+      "step": 14769
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6349417730106812,
+      "learning_rate": 3.0774786834319923e-06,
+      "loss": 0.1834,
+      "step": 14770
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4004228489340422,
+      "learning_rate": 3.076290232985696e-06,
+      "loss": 0.1997,
+      "step": 14771
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1630110956367583,
+      "learning_rate": 3.0751019703454164e-06,
+      "loss": 0.1626,
+      "step": 14772
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9623326057353061,
+      "learning_rate": 3.0739138955433878e-06,
+      "loss": 0.1676,
+      "step": 14773
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.006629466980831,
+      "learning_rate": 3.072726008611839e-06,
+      "loss": 0.1617,
+      "step": 14774
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2838886926972721,
+      "learning_rate": 3.0715383095829853e-06,
+      "loss": 0.1548,
+      "step": 14775
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0454608588604146,
+      "learning_rate": 3.070350798489049e-06,
+      "loss": 0.1815,
+      "step": 14776
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.821807555146335,
+      "learning_rate": 3.069163475362237e-06,
+      "loss": 0.1492,
+      "step": 14777
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8491864566705534,
+      "learning_rate": 3.0679763402347584e-06,
+      "loss": 0.1792,
+      "step": 14778
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3916599840610384,
+      "learning_rate": 3.0667893931388117e-06,
+      "loss": 0.1748,
+      "step": 14779
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0579779934680873,
+      "learning_rate": 3.0656026341065927e-06,
+      "loss": 0.141,
+      "step": 14780
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0207426330020186,
+      "learning_rate": 3.0644160631702945e-06,
+      "loss": 0.1737,
+      "step": 14781
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8869670421806797,
+      "learning_rate": 3.0632296803621064e-06,
+      "loss": 0.1637,
+      "step": 14782
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9987869626382742,
+      "learning_rate": 3.0620434857142e-06,
+      "loss": 0.1618,
+      "step": 14783
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9076548216511201,
+      "learning_rate": 3.0608574792587607e-06,
+      "loss": 0.168,
+      "step": 14784
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8639606682030742,
+      "learning_rate": 3.059671661027951e-06,
+      "loss": 0.1503,
+      "step": 14785
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.902688952256046,
+      "learning_rate": 3.0584860310539423e-06,
+      "loss": 0.15,
+      "step": 14786
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.830731620498615,
+      "learning_rate": 3.057300589368891e-06,
+      "loss": 0.1611,
+      "step": 14787
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3496131252289139,
+      "learning_rate": 3.0561153360049513e-06,
+      "loss": 0.1781,
+      "step": 14788
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1472205477231077,
+      "learning_rate": 3.0549302709942753e-06,
+      "loss": 0.1688,
+      "step": 14789
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2001411056709832,
+      "learning_rate": 3.0537453943690076e-06,
+      "loss": 0.1766,
+      "step": 14790
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8679364218872742,
+      "learning_rate": 3.0525607061612918e-06,
+      "loss": 0.1603,
+      "step": 14791
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8910635946636004,
+      "learning_rate": 3.051376206403256e-06,
+      "loss": 0.1595,
+      "step": 14792
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9196162615395724,
+      "learning_rate": 3.050191895127036e-06,
+      "loss": 0.1735,
+      "step": 14793
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8051345823549313,
+      "learning_rate": 3.0490077723647504e-06,
+      "loss": 0.1763,
+      "step": 14794
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.781799246737779,
+      "learning_rate": 3.047823838148525e-06,
+      "loss": 0.1682,
+      "step": 14795
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2158586699703775,
+      "learning_rate": 3.0466400925104665e-06,
+      "loss": 0.1896,
+      "step": 14796
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6484925242012631,
+      "learning_rate": 3.0454565354826926e-06,
+      "loss": 0.1647,
+      "step": 14797
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1769066884903636,
+      "learning_rate": 3.044273167097299e-06,
+      "loss": 0.1789,
+      "step": 14798
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9788467179153428,
+      "learning_rate": 3.0430899873863897e-06,
+      "loss": 0.1666,
+      "step": 14799
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.161712363023313,
+      "learning_rate": 3.041906996382056e-06,
+      "loss": 0.1538,
+      "step": 14800
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5549606547832509,
+      "learning_rate": 3.0407241941163923e-06,
+      "loss": 0.159,
+      "step": 14801
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.505694443799723,
+      "learning_rate": 3.0395415806214735e-06,
+      "loss": 0.193,
+      "step": 14802
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8924218114947182,
+      "learning_rate": 3.0383591559293867e-06,
+      "loss": 0.1531,
+      "step": 14803
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9852185215373485,
+      "learning_rate": 3.0371769200721977e-06,
+      "loss": 0.1806,
+      "step": 14804
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.2306235996053707,
+      "learning_rate": 3.035994873081981e-06,
+      "loss": 0.1845,
+      "step": 14805
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.906085883851588,
+      "learning_rate": 3.0348130149907928e-06,
+      "loss": 0.1791,
+      "step": 14806
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6349173864537299,
+      "learning_rate": 3.0336313458306964e-06,
+      "loss": 0.1894,
+      "step": 14807
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8587567982498606,
+      "learning_rate": 3.0324498656337453e-06,
+      "loss": 0.1749,
+      "step": 14808
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9227183523843092,
+      "learning_rate": 3.0312685744319824e-06,
+      "loss": 0.162,
+      "step": 14809
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.015572029239059,
+      "learning_rate": 3.030087472257456e-06,
+      "loss": 0.1698,
+      "step": 14810
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9089076200824454,
+      "learning_rate": 3.0289065591421974e-06,
+      "loss": 0.1732,
+      "step": 14811
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9013864412772916,
+      "learning_rate": 3.027725835118245e-06,
+      "loss": 0.1719,
+      "step": 14812
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9075117656813344,
+      "learning_rate": 3.0265453002176203e-06,
+      "loss": 0.189,
+      "step": 14813
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9636418655739848,
+      "learning_rate": 3.0253649544723517e-06,
+      "loss": 0.1516,
+      "step": 14814
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9890103632519298,
+      "learning_rate": 3.024184797914449e-06,
+      "loss": 0.1742,
+      "step": 14815
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9107581267039456,
+      "learning_rate": 3.0230048305759274e-06,
+      "loss": 0.1547,
+      "step": 14816
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.351786077343159,
+      "learning_rate": 3.021825052488795e-06,
+      "loss": 0.1514,
+      "step": 14817
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.066387189414217,
+      "learning_rate": 3.0206454636850546e-06,
+      "loss": 0.1591,
+      "step": 14818
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0678313553067573,
+      "learning_rate": 3.0194660641966965e-06,
+      "loss": 0.1619,
+      "step": 14819
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8015416936308438,
+      "learning_rate": 3.01828685405572e-06,
+      "loss": 0.1536,
+      "step": 14820
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1124837008765287,
+      "learning_rate": 3.0171078332941028e-06,
+      "loss": 0.1707,
+      "step": 14821
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9763316659811963,
+      "learning_rate": 3.015929001943834e-06,
+      "loss": 0.1748,
+      "step": 14822
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9133392214693982,
+      "learning_rate": 3.014750360036881e-06,
+      "loss": 0.138,
+      "step": 14823
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1592682615722274,
+      "learning_rate": 3.0135719076052195e-06,
+      "loss": 0.187,
+      "step": 14824
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6634091590839288,
+      "learning_rate": 3.012393644680819e-06,
+      "loss": 0.1678,
+      "step": 14825
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8682168822734787,
+      "learning_rate": 3.0112155712956305e-06,
+      "loss": 0.1535,
+      "step": 14826
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9829329473504432,
+      "learning_rate": 3.0100376874816183e-06,
+      "loss": 0.1736,
+      "step": 14827
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9083554632551712,
+      "learning_rate": 3.0088599932707263e-06,
+      "loss": 0.1657,
+      "step": 14828
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9103203674379348,
+      "learning_rate": 3.007682488694904e-06,
+      "loss": 0.1599,
+      "step": 14829
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8617723421597876,
+      "learning_rate": 3.0065051737860863e-06,
+      "loss": 0.1631,
+      "step": 14830
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5295785801478599,
+      "learning_rate": 3.0053280485762136e-06,
+      "loss": 0.168,
+      "step": 14831
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.010684464650659,
+      "learning_rate": 3.00415111309721e-06,
+      "loss": 0.1756,
+      "step": 14832
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1799490977267986,
+      "learning_rate": 3.0029743673810032e-06,
+      "loss": 0.2012,
+      "step": 14833
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.353078753744399,
+      "learning_rate": 3.0017978114595103e-06,
+      "loss": 0.1659,
+      "step": 14834
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.3560744486295873,
+      "learning_rate": 3.000621445364651e-06,
+      "loss": 0.167,
+      "step": 14835
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3890085822404705,
+      "learning_rate": 2.999445269128327e-06,
+      "loss": 0.1639,
+      "step": 14836
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0976117458451753,
+      "learning_rate": 2.9982692827824487e-06,
+      "loss": 0.1728,
+      "step": 14837
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.079496472102237,
+      "learning_rate": 2.9970934863589086e-06,
+      "loss": 0.1678,
+      "step": 14838
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.842838928379633,
+      "learning_rate": 2.9959178798896062e-06,
+      "loss": 0.1422,
+      "step": 14839
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9306792323418289,
+      "learning_rate": 2.994742463406427e-06,
+      "loss": 0.1736,
+      "step": 14840
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8812207056004026,
+      "learning_rate": 2.9935672369412507e-06,
+      "loss": 0.1755,
+      "step": 14841
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0940408765288119,
+      "learning_rate": 2.9923922005259588e-06,
+      "loss": 0.1793,
+      "step": 14842
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8373831947473562,
+      "learning_rate": 2.991217354192425e-06,
+      "loss": 0.1765,
+      "step": 14843
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3787908438981957,
+      "learning_rate": 2.9900426979725183e-06,
+      "loss": 0.1743,
+      "step": 14844
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1502864105435349,
+      "learning_rate": 2.9888682318980975e-06,
+      "loss": 0.1636,
+      "step": 14845
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.044664524211976,
+      "learning_rate": 2.9876939560010244e-06,
+      "loss": 0.1583,
+      "step": 14846
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9738137676351397,
+      "learning_rate": 2.986519870313146e-06,
+      "loss": 0.172,
+      "step": 14847
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8642222069852814,
+      "learning_rate": 2.985345974866315e-06,
+      "loss": 0.1423,
+      "step": 14848
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.810328271150909,
+      "learning_rate": 2.984172269692367e-06,
+      "loss": 0.1637,
+      "step": 14849
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1340679026784688,
+      "learning_rate": 2.9829987548231432e-06,
+      "loss": 0.1611,
+      "step": 14850
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.070376228632435,
+      "learning_rate": 2.9818254302904735e-06,
+      "loss": 0.1681,
+      "step": 14851
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0497269808800511,
+      "learning_rate": 2.9806522961261896e-06,
+      "loss": 0.1623,
+      "step": 14852
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4840784587493108,
+      "learning_rate": 2.979479352362106e-06,
+      "loss": 0.1548,
+      "step": 14853
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9688261626228124,
+      "learning_rate": 2.9783065990300432e-06,
+      "loss": 0.1758,
+      "step": 14854
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9197866418393205,
+      "learning_rate": 2.9771340361618075e-06,
+      "loss": 0.1787,
+      "step": 14855
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6718879704798661,
+      "learning_rate": 2.975961663789212e-06,
+      "loss": 0.1573,
+      "step": 14856
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9578131340707566,
+      "learning_rate": 2.9747894819440514e-06,
+      "loss": 0.172,
+      "step": 14857
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.930379079874052,
+      "learning_rate": 2.9736174906581216e-06,
+      "loss": 0.1655,
+      "step": 14858
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7535419036472633,
+      "learning_rate": 2.9724456899632126e-06,
+      "loss": 0.1555,
+      "step": 14859
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2210499593610467,
+      "learning_rate": 2.971274079891112e-06,
+      "loss": 0.1867,
+      "step": 14860
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1231733960892714,
+      "learning_rate": 2.970102660473603e-06,
+      "loss": 0.1519,
+      "step": 14861
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.111326106758271,
+      "learning_rate": 2.9689314317424513e-06,
+      "loss": 0.1765,
+      "step": 14862
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.061444520286897,
+      "learning_rate": 2.9677603937294364e-06,
+      "loss": 0.1544,
+      "step": 14863
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0349430214778528,
+      "learning_rate": 2.966589546466314e-06,
+      "loss": 0.1518,
+      "step": 14864
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0634051726805134,
+      "learning_rate": 2.96541888998485e-06,
+      "loss": 0.1696,
+      "step": 14865
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1166888090548637,
+      "learning_rate": 2.964248424316795e-06,
+      "loss": 0.1612,
+      "step": 14866
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3298552807639807,
+      "learning_rate": 2.9630781494938997e-06,
+      "loss": 0.1833,
+      "step": 14867
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.236808824422805,
+      "learning_rate": 2.961908065547905e-06,
+      "loss": 0.1529,
+      "step": 14868
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4039050352062317,
+      "learning_rate": 2.960738172510551e-06,
+      "loss": 0.1812,
+      "step": 14869
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1263685791612776,
+      "learning_rate": 2.959568470413572e-06,
+      "loss": 0.1675,
+      "step": 14870
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2335040738408105,
+      "learning_rate": 2.9583989592886985e-06,
+      "loss": 0.1848,
+      "step": 14871
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0305781290114908,
+      "learning_rate": 2.957229639167648e-06,
+      "loss": 0.1641,
+      "step": 14872
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8071061008183066,
+      "learning_rate": 2.956060510082145e-06,
+      "loss": 0.1619,
+      "step": 14873
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0075024836831434,
+      "learning_rate": 2.954891572063895e-06,
+      "loss": 0.1607,
+      "step": 14874
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1847496166015081,
+      "learning_rate": 2.9537228251446125e-06,
+      "loss": 0.1713,
+      "step": 14875
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.177493790578264,
+      "learning_rate": 2.9525542693559926e-06,
+      "loss": 0.1785,
+      "step": 14876
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0309264603516655,
+      "learning_rate": 2.9513859047297366e-06,
+      "loss": 0.178,
+      "step": 14877
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.493847712149435,
+      "learning_rate": 2.9502177312975387e-06,
+      "loss": 0.1551,
+      "step": 14878
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1904513588353383,
+      "learning_rate": 2.9490497490910806e-06,
+      "loss": 0.1753,
+      "step": 14879
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8749388948121452,
+      "learning_rate": 2.9478819581420493e-06,
+      "loss": 0.1617,
+      "step": 14880
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2930194430877306,
+      "learning_rate": 2.9467143584821145e-06,
+      "loss": 0.1655,
+      "step": 14881
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9587201065916398,
+      "learning_rate": 2.9455469501429557e-06,
+      "loss": 0.1638,
+      "step": 14882
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.031977326446882,
+      "learning_rate": 2.9443797331562295e-06,
+      "loss": 0.1904,
+      "step": 14883
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1442108789360643,
+      "learning_rate": 2.9432127075536056e-06,
+      "loss": 0.1624,
+      "step": 14884
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5933491740422148,
+      "learning_rate": 2.942045873366731e-06,
+      "loss": 0.169,
+      "step": 14885
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.243282759250556,
+      "learning_rate": 2.9408792306272625e-06,
+      "loss": 0.17,
+      "step": 14886
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8418987826724236,
+      "learning_rate": 2.9397127793668435e-06,
+      "loss": 0.1578,
+      "step": 14887
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9680157927451623,
+      "learning_rate": 2.938546519617116e-06,
+      "loss": 0.1646,
+      "step": 14888
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.483610722538024,
+      "learning_rate": 2.93738045140971e-06,
+      "loss": 0.1769,
+      "step": 14889
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7884359484807628,
+      "learning_rate": 2.9362145747762626e-06,
+      "loss": 0.1601,
+      "step": 14890
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.121494941591859,
+      "learning_rate": 2.9350488897483897e-06,
+      "loss": 0.1894,
+      "step": 14891
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2684480232810904,
+      "learning_rate": 2.9338833963577184e-06,
+      "loss": 0.1585,
+      "step": 14892
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3305637146146216,
+      "learning_rate": 2.932718094635858e-06,
+      "loss": 0.2072,
+      "step": 14893
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3770407432912448,
+      "learning_rate": 2.9315529846144162e-06,
+      "loss": 0.169,
+      "step": 14894
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9470699738751776,
+      "learning_rate": 2.9303880663249985e-06,
+      "loss": 0.1744,
+      "step": 14895
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0590649066073299,
+      "learning_rate": 2.9292233397992043e-06,
+      "loss": 0.1538,
+      "step": 14896
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0300673069287118,
+      "learning_rate": 2.9280588050686287e-06,
+      "loss": 0.1912,
+      "step": 14897
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0014920304678387,
+      "learning_rate": 2.9268944621648554e-06,
+      "loss": 0.1973,
+      "step": 14898
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6558249390820992,
+      "learning_rate": 2.925730311119471e-06,
+      "loss": 0.1602,
+      "step": 14899
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1327455208047599,
+      "learning_rate": 2.92456635196405e-06,
+      "loss": 0.1705,
+      "step": 14900
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0468497436593192,
+      "learning_rate": 2.9234025847301685e-06,
+      "loss": 0.1679,
+      "step": 14901
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8848502370869115,
+      "learning_rate": 2.922239009449388e-06,
+      "loss": 0.1557,
+      "step": 14902
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.23931656859147,
+      "learning_rate": 2.9210756261532746e-06,
+      "loss": 0.1499,
+      "step": 14903
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5678878010715138,
+      "learning_rate": 2.919912434873385e-06,
+      "loss": 0.1665,
+      "step": 14904
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8378693122157262,
+      "learning_rate": 2.918749435641274e-06,
+      "loss": 0.1657,
+      "step": 14905
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.856513967247774,
+      "learning_rate": 2.91758662848848e-06,
+      "loss": 0.1677,
+      "step": 14906
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5243921986143543,
+      "learning_rate": 2.9164240134465527e-06,
+      "loss": 0.1661,
+      "step": 14907
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8660583832399535,
+      "learning_rate": 2.9152615905470216e-06,
+      "loss": 0.1686,
+      "step": 14908
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1635873514860704,
+      "learning_rate": 2.9140993598214217e-06,
+      "loss": 0.1904,
+      "step": 14909
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3164720376059806,
+      "learning_rate": 2.912937321301278e-06,
+      "loss": 0.1663,
+      "step": 14910
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0820820698019624,
+      "learning_rate": 2.911775475018106e-06,
+      "loss": 0.1674,
+      "step": 14911
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9985268905374507,
+      "learning_rate": 2.910613821003425e-06,
+      "loss": 0.1577,
+      "step": 14912
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1918499534603288,
+      "learning_rate": 2.9094523592887446e-06,
+      "loss": 0.182,
+      "step": 14913
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2000467515295787,
+      "learning_rate": 2.9082910899055717e-06,
+      "loss": 0.1647,
+      "step": 14914
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.425609422227406,
+      "learning_rate": 2.9071300128854007e-06,
+      "loss": 0.1677,
+      "step": 14915
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4507515531337645,
+      "learning_rate": 2.9059691282597325e-06,
+      "loss": 0.1733,
+      "step": 14916
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3031539416406344,
+      "learning_rate": 2.9048084360600494e-06,
+      "loss": 0.1608,
+      "step": 14917
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8005481370728588,
+      "learning_rate": 2.9036479363178405e-06,
+      "loss": 0.1748,
+      "step": 14918
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4637440704208484,
+      "learning_rate": 2.9024876290645787e-06,
+      "loss": 0.1587,
+      "step": 14919
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.022621538622618,
+      "learning_rate": 2.9013275143317453e-06,
+      "loss": 0.1641,
+      "step": 14920
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5938994367208572,
+      "learning_rate": 2.9001675921507998e-06,
+      "loss": 0.1825,
+      "step": 14921
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5829965650281022,
+      "learning_rate": 2.8990078625532104e-06,
+      "loss": 0.147,
+      "step": 14922
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.609257440006581,
+      "learning_rate": 2.8978483255704325e-06,
+      "loss": 0.1568,
+      "step": 14923
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4008987644802413,
+      "learning_rate": 2.8966889812339237e-06,
+      "loss": 0.1795,
+      "step": 14924
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3581868558224195,
+      "learning_rate": 2.8955298295751245e-06,
+      "loss": 0.1589,
+      "step": 14925
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9135404012632948,
+      "learning_rate": 2.8943708706254824e-06,
+      "loss": 0.1602,
+      "step": 14926
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0471223644901613,
+      "learning_rate": 2.893212104416432e-06,
+      "loss": 0.1847,
+      "step": 14927
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.853204413537538,
+      "learning_rate": 2.8920535309794018e-06,
+      "loss": 0.1418,
+      "step": 14928
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1207432193111928,
+      "learning_rate": 2.8908951503458217e-06,
+      "loss": 0.1736,
+      "step": 14929
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4981305908868932,
+      "learning_rate": 2.8897369625471105e-06,
+      "loss": 0.1729,
+      "step": 14930
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1182424351020628,
+      "learning_rate": 2.8885789676146903e-06,
+      "loss": 0.1718,
+      "step": 14931
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.354916786961543,
+      "learning_rate": 2.887421165579963e-06,
+      "loss": 0.1484,
+      "step": 14932
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5863329120631235,
+      "learning_rate": 2.8862635564743424e-06,
+      "loss": 0.1692,
+      "step": 14933
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9166221221801297,
+      "learning_rate": 2.8851061403292213e-06,
+      "loss": 0.1578,
+      "step": 14934
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.035905778944065,
+      "learning_rate": 2.8839489171760015e-06,
+      "loss": 0.1755,
+      "step": 14935
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8688123045023516,
+      "learning_rate": 2.882791887046066e-06,
+      "loss": 0.1514,
+      "step": 14936
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9109323250386723,
+      "learning_rate": 2.8816350499708044e-06,
+      "loss": 0.1676,
+      "step": 14937
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.116979008834176,
+      "learning_rate": 2.8804784059815914e-06,
+      "loss": 0.1597,
+      "step": 14938
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9080606464625794,
+      "learning_rate": 2.879321955109805e-06,
+      "loss": 0.1694,
+      "step": 14939
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4528067891010092,
+      "learning_rate": 2.878165697386812e-06,
+      "loss": 0.1516,
+      "step": 14940
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0721172933704726,
+      "learning_rate": 2.87700963284398e-06,
+      "loss": 0.1821,
+      "step": 14941
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0226958352661308,
+      "learning_rate": 2.87585376151266e-06,
+      "loss": 0.1745,
+      "step": 14942
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.973993706553113,
+      "learning_rate": 2.8746980834242133e-06,
+      "loss": 0.1899,
+      "step": 14943
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0773078735796286,
+      "learning_rate": 2.8735425986099796e-06,
+      "loss": 0.1737,
+      "step": 14944
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9840464557781372,
+      "learning_rate": 2.87238730710131e-06,
+      "loss": 0.172,
+      "step": 14945
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9383012071885413,
+      "learning_rate": 2.871232208929533e-06,
+      "loss": 0.1812,
+      "step": 14946
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2462761175413661,
+      "learning_rate": 2.8700773041259844e-06,
+      "loss": 0.1845,
+      "step": 14947
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.32041091829035,
+      "learning_rate": 2.8689225927219956e-06,
+      "loss": 0.1949,
+      "step": 14948
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1985684244279526,
+      "learning_rate": 2.8677680747488812e-06,
+      "loss": 0.1804,
+      "step": 14949
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1156791419115983,
+      "learning_rate": 2.8666137502379632e-06,
+      "loss": 0.199,
+      "step": 14950
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.149339106701358,
+      "learning_rate": 2.8654596192205476e-06,
+      "loss": 0.1742,
+      "step": 14951
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9928395353613638,
+      "learning_rate": 2.8643056817279448e-06,
+      "loss": 0.1598,
+      "step": 14952
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0961845244112949,
+      "learning_rate": 2.863151937791452e-06,
+      "loss": 0.1732,
+      "step": 14953
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8936819931072454,
+      "learning_rate": 2.8619983874423672e-06,
+      "loss": 0.1709,
+      "step": 14954
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.083003245432732,
+      "learning_rate": 2.8608450307119772e-06,
+      "loss": 0.1599,
+      "step": 14955
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0890360019663619,
+      "learning_rate": 2.8596918676315687e-06,
+      "loss": 0.1804,
+      "step": 14956
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.921592315801085,
+      "learning_rate": 2.8585388982324226e-06,
+      "loss": 0.1718,
+      "step": 14957
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0191836834087407,
+      "learning_rate": 2.8573861225458143e-06,
+      "loss": 0.1588,
+      "step": 14958
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.049360441030582,
+      "learning_rate": 2.8562335406030074e-06,
+      "loss": 0.1789,
+      "step": 14959
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.021851023037416,
+      "learning_rate": 2.8550811524352727e-06,
+      "loss": 0.1371,
+      "step": 14960
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0470103475321244,
+      "learning_rate": 2.8539289580738627e-06,
+      "loss": 0.1542,
+      "step": 14961
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.033767752087917,
+      "learning_rate": 2.8527769575500363e-06,
+      "loss": 0.1691,
+      "step": 14962
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7569499057857303,
+      "learning_rate": 2.851625150895039e-06,
+      "loss": 0.1773,
+      "step": 14963
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9601401947005221,
+      "learning_rate": 2.850473538140108e-06,
+      "loss": 0.1709,
+      "step": 14964
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9617429520069206,
+      "learning_rate": 2.8493221193164886e-06,
+      "loss": 0.1505,
+      "step": 14965
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6305919387900003,
+      "learning_rate": 2.84817089445541e-06,
+      "loss": 0.1857,
+      "step": 14966
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.992601166343874,
+      "learning_rate": 2.847019863588102e-06,
+      "loss": 0.1686,
+      "step": 14967
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.213736998481106,
+      "learning_rate": 2.845869026745781e-06,
+      "loss": 0.196,
+      "step": 14968
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9976238049114304,
+      "learning_rate": 2.8447183839596705e-06,
+      "loss": 0.1558,
+      "step": 14969
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8547754154366524,
+      "learning_rate": 2.8435679352609747e-06,
+      "loss": 0.1787,
+      "step": 14970
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2194540168258508,
+      "learning_rate": 2.8424176806809068e-06,
+      "loss": 0.1417,
+      "step": 14971
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.14402890444547,
+      "learning_rate": 2.8412676202506596e-06,
+      "loss": 0.1685,
+      "step": 14972
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9557704528886504,
+      "learning_rate": 2.8401177540014323e-06,
+      "loss": 0.173,
+      "step": 14973
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.920417194706065,
+      "learning_rate": 2.838968081964416e-06,
+      "loss": 0.1653,
+      "step": 14974
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1526702838319869,
+      "learning_rate": 2.8378186041707977e-06,
+      "loss": 0.1799,
+      "step": 14975
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0949852077610864,
+      "learning_rate": 2.8366693206517503e-06,
+      "loss": 0.1584,
+      "step": 14976
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2539237468301876,
+      "learning_rate": 2.835520231438457e-06,
+      "loss": 0.1617,
+      "step": 14977
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.015585176023755,
+      "learning_rate": 2.834371336562077e-06,
+      "loss": 0.1508,
+      "step": 14978
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0288380937596868,
+      "learning_rate": 2.833222636053784e-06,
+      "loss": 0.2067,
+      "step": 14979
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4883629291388756,
+      "learning_rate": 2.8320741299447306e-06,
+      "loss": 0.1725,
+      "step": 14980
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8305302577491667,
+      "learning_rate": 2.8309258182660693e-06,
+      "loss": 0.1578,
+      "step": 14981
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9041171256480882,
+      "learning_rate": 2.829777701048949e-06,
+      "loss": 0.1483,
+      "step": 14982
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1448825644487437,
+      "learning_rate": 2.828629778324514e-06,
+      "loss": 0.1614,
+      "step": 14983
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3660210089224283,
+      "learning_rate": 2.827482050123905e-06,
+      "loss": 0.1702,
+      "step": 14984
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3638108245767226,
+      "learning_rate": 2.8263345164782473e-06,
+      "loss": 0.1741,
+      "step": 14985
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2411375513006744,
+      "learning_rate": 2.8251871774186736e-06,
+      "loss": 0.1566,
+      "step": 14986
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9342897147302336,
+      "learning_rate": 2.8240400329762994e-06,
+      "loss": 0.1805,
+      "step": 14987
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0304143141046265,
+      "learning_rate": 2.8228930831822486e-06,
+      "loss": 0.1634,
+      "step": 14988
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.775467905274122,
+      "learning_rate": 2.821746328067625e-06,
+      "loss": 0.1691,
+      "step": 14989
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9910104080265708,
+      "learning_rate": 2.82059976766354e-06,
+      "loss": 0.1616,
+      "step": 14990
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6828123750215387,
+      "learning_rate": 2.81945340200109e-06,
+      "loss": 0.1768,
+      "step": 14991
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0531142469858463,
+      "learning_rate": 2.818307231111371e-06,
+      "loss": 0.197,
+      "step": 14992
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2279946254505785,
+      "learning_rate": 2.8171612550254746e-06,
+      "loss": 0.1581,
+      "step": 14993
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2522759008118522,
+      "learning_rate": 2.816015473774487e-06,
+      "loss": 0.1314,
+      "step": 14994
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.995273571793144,
+      "learning_rate": 2.814869887389483e-06,
+      "loss": 0.172,
+      "step": 14995
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0064004224764271,
+      "learning_rate": 2.813724495901543e-06,
+      "loss": 0.1722,
+      "step": 14996
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0028392290053534,
+      "learning_rate": 2.812579299341731e-06,
+      "loss": 0.1667,
+      "step": 14997
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9677404251213896,
+      "learning_rate": 2.811434297741108e-06,
+      "loss": 0.1836,
+      "step": 14998
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9845281773459438,
+      "learning_rate": 2.8102894911307367e-06,
+      "loss": 0.1618,
+      "step": 14999
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8616214642095336,
+      "learning_rate": 2.809144879541669e-06,
+      "loss": 0.162,
+      "step": 15000
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6859524501238814,
+      "learning_rate": 2.808000463004954e-06,
+      "loss": 0.1528,
+      "step": 15001
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.160218092080421,
+      "learning_rate": 2.8068562415516308e-06,
+      "loss": 0.1746,
+      "step": 15002
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1401600807234633,
+      "learning_rate": 2.8057122152127413e-06,
+      "loss": 0.1789,
+      "step": 15003
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.032889345595397,
+      "learning_rate": 2.804568384019312e-06,
+      "loss": 0.1571,
+      "step": 15004
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8592553578925399,
+      "learning_rate": 2.8034247480023735e-06,
+      "loss": 0.1717,
+      "step": 15005
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1253400311781885,
+      "learning_rate": 2.8022813071929434e-06,
+      "loss": 0.1502,
+      "step": 15006
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9738577653061913,
+      "learning_rate": 2.8011380616220407e-06,
+      "loss": 0.1847,
+      "step": 15007
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3238903583814505,
+      "learning_rate": 2.7999950113206732e-06,
+      "loss": 0.1883,
+      "step": 15008
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8885027709396675,
+      "learning_rate": 2.798852156319847e-06,
+      "loss": 0.1562,
+      "step": 15009
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0721210227832059,
+      "learning_rate": 2.7977094966505624e-06,
+      "loss": 0.1961,
+      "step": 15010
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1721695304850344,
+      "learning_rate": 2.7965670323438178e-06,
+      "loss": 0.179,
+      "step": 15011
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3159953959620643,
+      "learning_rate": 2.7954247634305965e-06,
+      "loss": 0.1562,
+      "step": 15012
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9882891878751142,
+      "learning_rate": 2.7942826899418886e-06,
+      "loss": 0.1814,
+      "step": 15013
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.071625820697916,
+      "learning_rate": 2.7931408119086668e-06,
+      "loss": 0.1998,
+      "step": 15014
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1167208938383686,
+      "learning_rate": 2.79199912936191e-06,
+      "loss": 0.1751,
+      "step": 15015
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0858374521812366,
+      "learning_rate": 2.790857642332584e-06,
+      "loss": 0.1723,
+      "step": 15016
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9124234370105371,
+      "learning_rate": 2.789716350851649e-06,
+      "loss": 0.1581,
+      "step": 15017
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3363800719743475,
+      "learning_rate": 2.7885752549500644e-06,
+      "loss": 0.1549,
+      "step": 15018
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0255205464470902,
+      "learning_rate": 2.7874343546587846e-06,
+      "loss": 0.1846,
+      "step": 15019
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9317828134518765,
+      "learning_rate": 2.7862936500087566e-06,
+      "loss": 0.164,
+      "step": 15020
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8632663766608207,
+      "learning_rate": 2.7851531410309194e-06,
+      "loss": 0.1741,
+      "step": 15021
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1484226086902631,
+      "learning_rate": 2.784012827756213e-06,
+      "loss": 0.1755,
+      "step": 15022
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8400458929925192,
+      "learning_rate": 2.7828727102155627e-06,
+      "loss": 0.1696,
+      "step": 15023
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9106478377375347,
+      "learning_rate": 2.7817327884399014e-06,
+      "loss": 0.1882,
+      "step": 15024
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3584652549796459,
+      "learning_rate": 2.7805930624601427e-06,
+      "loss": 0.147,
+      "step": 15025
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9858682454425882,
+      "learning_rate": 2.779453532307206e-06,
+      "loss": 0.1864,
+      "step": 15026
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9690793197636596,
+      "learning_rate": 2.778314198011999e-06,
+      "loss": 0.1641,
+      "step": 15027
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9040955497405057,
+      "learning_rate": 2.7771750596054305e-06,
+      "loss": 0.1608,
+      "step": 15028
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2710265204702358,
+      "learning_rate": 2.7760361171183934e-06,
+      "loss": 0.1689,
+      "step": 15029
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1442824752872833,
+      "learning_rate": 2.774897370581787e-06,
+      "loss": 0.1542,
+      "step": 15030
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0493120502774447,
+      "learning_rate": 2.7737588200264953e-06,
+      "loss": 0.1577,
+      "step": 15031
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0022591482167709,
+      "learning_rate": 2.7726204654834067e-06,
+      "loss": 0.1754,
+      "step": 15032
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.844231737981002,
+      "learning_rate": 2.7714823069833964e-06,
+      "loss": 0.1736,
+      "step": 15033
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4792458478526416,
+      "learning_rate": 2.770344344557333e-06,
+      "loss": 0.1683,
+      "step": 15034
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2239895492597026,
+      "learning_rate": 2.7692065782360876e-06,
+      "loss": 0.1632,
+      "step": 15035
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.991586953334937,
+      "learning_rate": 2.7680690080505234e-06,
+      "loss": 0.1479,
+      "step": 15036
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2104851538907433,
+      "learning_rate": 2.7669316340314977e-06,
+      "loss": 0.1594,
+      "step": 15037
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9131732170375813,
+      "learning_rate": 2.765794456209857e-06,
+      "loss": 0.1563,
+      "step": 15038
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.228509649935169,
+      "learning_rate": 2.7646574746164533e-06,
+      "loss": 0.1604,
+      "step": 15039
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9544290582199312,
+      "learning_rate": 2.763520689282122e-06,
+      "loss": 0.1653,
+      "step": 15040
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9012892594587696,
+      "learning_rate": 2.762384100237703e-06,
+      "loss": 0.1657,
+      "step": 15041
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0570949666178724,
+      "learning_rate": 2.761247707514021e-06,
+      "loss": 0.1753,
+      "step": 15042
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0871063071718943,
+      "learning_rate": 2.7601115111419043e-06,
+      "loss": 0.1647,
+      "step": 15043
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8618285709761413,
+      "learning_rate": 2.758975511152171e-06,
+      "loss": 0.1569,
+      "step": 15044
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.547419540473778,
+      "learning_rate": 2.7578397075756404e-06,
+      "loss": 0.1648,
+      "step": 15045
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0384641122566953,
+      "learning_rate": 2.756704100443113e-06,
+      "loss": 0.1722,
+      "step": 15046
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8121363605603122,
+      "learning_rate": 2.7555686897853983e-06,
+      "loss": 0.1751,
+      "step": 15047
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7748124221997982,
+      "learning_rate": 2.754433475633289e-06,
+      "loss": 0.1717,
+      "step": 15048
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2459613623991952,
+      "learning_rate": 2.753298458017585e-06,
+      "loss": 0.1689,
+      "step": 15049
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7775055960696998,
+      "learning_rate": 2.7521636369690687e-06,
+      "loss": 0.1628,
+      "step": 15050
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0486466173882525,
+      "learning_rate": 2.7510290125185203e-06,
+      "loss": 0.172,
+      "step": 15051
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0057226924940363,
+      "learning_rate": 2.7498945846967197e-06,
+      "loss": 0.1626,
+      "step": 15052
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7624894425040479,
+      "learning_rate": 2.7487603535344375e-06,
+      "loss": 0.1775,
+      "step": 15053
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1466543189425418,
+      "learning_rate": 2.747626319062444e-06,
+      "loss": 0.1886,
+      "step": 15054
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8899801858769728,
+      "learning_rate": 2.7464924813114926e-06,
+      "loss": 0.1682,
+      "step": 15055
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.07558876667753,
+      "learning_rate": 2.7453588403123453e-06,
+      "loss": 0.1724,
+      "step": 15056
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0008619053830783,
+      "learning_rate": 2.7442253960957466e-06,
+      "loss": 0.1681,
+      "step": 15057
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0007558626762345,
+      "learning_rate": 2.743092148692447e-06,
+      "loss": 0.1731,
+      "step": 15058
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.289803694977765,
+      "learning_rate": 2.741959098133179e-06,
+      "loss": 0.1628,
+      "step": 15059
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4436849600973654,
+      "learning_rate": 2.7408262444486844e-06,
+      "loss": 0.1721,
+      "step": 15060
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9500321666703038,
+      "learning_rate": 2.739693587669684e-06,
+      "loss": 0.1584,
+      "step": 15061
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.902479170382834,
+      "learning_rate": 2.7385611278269054e-06,
+      "loss": 0.1546,
+      "step": 15062
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9355144421315881,
+      "learning_rate": 2.737428864951066e-06,
+      "loss": 0.1703,
+      "step": 15063
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8297513558457941,
+      "learning_rate": 2.736296799072883e-06,
+      "loss": 0.1633,
+      "step": 15064
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9871101636958638,
+      "learning_rate": 2.7351649302230553e-06,
+      "loss": 0.157,
+      "step": 15065
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.167113858545876,
+      "learning_rate": 2.7340332584322927e-06,
+      "loss": 0.1538,
+      "step": 15066
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8959136261904793,
+      "learning_rate": 2.7329017837312875e-06,
+      "loss": 0.1492,
+      "step": 15067
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0774540888808917,
+      "learning_rate": 2.7317705061507306e-06,
+      "loss": 0.166,
+      "step": 15068
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1236019697314357,
+      "learning_rate": 2.7306394257213078e-06,
+      "loss": 0.1596,
+      "step": 15069
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.474374951856886,
+      "learning_rate": 2.729508542473702e-06,
+      "loss": 0.1653,
+      "step": 15070
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7946612833079612,
+      "learning_rate": 2.72837785643859e-06,
+      "loss": 0.18,
+      "step": 15071
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8050299194968318,
+      "learning_rate": 2.727247367646637e-06,
+      "loss": 0.1701,
+      "step": 15072
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9506576876624673,
+      "learning_rate": 2.726117076128513e-06,
+      "loss": 0.1817,
+      "step": 15073
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2612952414807954,
+      "learning_rate": 2.72498698191487e-06,
+      "loss": 0.1674,
+      "step": 15074
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3330606571640877,
+      "learning_rate": 2.72385708503637e-06,
+      "loss": 0.1822,
+      "step": 15075
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2579774240785415,
+      "learning_rate": 2.7227273855236535e-06,
+      "loss": 0.1622,
+      "step": 15076
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8701657202529226,
+      "learning_rate": 2.721597883407372e-06,
+      "loss": 0.1991,
+      "step": 15077
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8020324879684333,
+      "learning_rate": 2.720468578718155e-06,
+      "loss": 0.1824,
+      "step": 15078
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0579987620792803,
+      "learning_rate": 2.7193394714866396e-06,
+      "loss": 0.1741,
+      "step": 15079
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2534880239050814,
+      "learning_rate": 2.7182105617434516e-06,
+      "loss": 0.1839,
+      "step": 15080
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.070027748514126,
+      "learning_rate": 2.7170818495192163e-06,
+      "loss": 0.1591,
+      "step": 15081
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.145745299916052,
+      "learning_rate": 2.7159533348445455e-06,
+      "loss": 0.1667,
+      "step": 15082
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0809253146050162,
+      "learning_rate": 2.7148250177500534e-06,
+      "loss": 0.1813,
+      "step": 15083
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0653540590236674,
+      "learning_rate": 2.7136968982663427e-06,
+      "loss": 0.1828,
+      "step": 15084
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5754998020627489,
+      "learning_rate": 2.7125689764240173e-06,
+      "loss": 0.1576,
+      "step": 15085
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.740473150238454,
+      "learning_rate": 2.711441252253669e-06,
+      "loss": 0.1617,
+      "step": 15086
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5122081253911168,
+      "learning_rate": 2.7103137257858867e-06,
+      "loss": 0.1749,
+      "step": 15087
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.131401733962826,
+      "learning_rate": 2.7091863970512564e-06,
+      "loss": 0.1797,
+      "step": 15088
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0242405027137653,
+      "learning_rate": 2.708059266080356e-06,
+      "loss": 0.1844,
+      "step": 15089
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8234663848030147,
+      "learning_rate": 2.7069323329037632e-06,
+      "loss": 0.1545,
+      "step": 15090
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.363966608963242,
+      "learning_rate": 2.7058055975520405e-06,
+      "loss": 0.1817,
+      "step": 15091
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4655481719724168,
+      "learning_rate": 2.704679060055755e-06,
+      "loss": 0.1859,
+      "step": 15092
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3100547717657192,
+      "learning_rate": 2.703552720445459e-06,
+      "loss": 0.1922,
+      "step": 15093
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.968730998319128,
+      "learning_rate": 2.702426578751711e-06,
+      "loss": 0.1686,
+      "step": 15094
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.464987708079324,
+      "learning_rate": 2.701300635005052e-06,
+      "loss": 0.1654,
+      "step": 15095
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1518904234958036,
+      "learning_rate": 2.7001748892360247e-06,
+      "loss": 0.1543,
+      "step": 15096
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1980341783397113,
+      "learning_rate": 2.6990493414751652e-06,
+      "loss": 0.152,
+      "step": 15097
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0117771613701103,
+      "learning_rate": 2.697923991753009e-06,
+      "loss": 0.1548,
+      "step": 15098
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3844556881797818,
+      "learning_rate": 2.696798840100072e-06,
+      "loss": 0.1779,
+      "step": 15099
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2282852794525476,
+      "learning_rate": 2.6956738865468832e-06,
+      "loss": 0.1638,
+      "step": 15100
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9476410124228332,
+      "learning_rate": 2.6945491311239504e-06,
+      "loss": 0.1522,
+      "step": 15101
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8403668096547272,
+      "learning_rate": 2.693424573861787e-06,
+      "loss": 0.1584,
+      "step": 15102
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0560591291079178,
+      "learning_rate": 2.692300214790895e-06,
+      "loss": 0.1808,
+      "step": 15103
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9424701654290392,
+      "learning_rate": 2.6911760539417698e-06,
+      "loss": 0.1649,
+      "step": 15104
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.54019391223232,
+      "learning_rate": 2.690052091344907e-06,
+      "loss": 0.1624,
+      "step": 15105
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1614143182985948,
+      "learning_rate": 2.688928327030793e-06,
+      "loss": 0.175,
+      "step": 15106
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.184879417210331,
+      "learning_rate": 2.6878047610299152e-06,
+      "loss": 0.1578,
+      "step": 15107
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.531118791624129,
+      "learning_rate": 2.686681393372743e-06,
+      "loss": 0.1774,
+      "step": 15108
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9591536192176581,
+      "learning_rate": 2.6855582240897536e-06,
+      "loss": 0.1653,
+      "step": 15109
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0411427116028142,
+      "learning_rate": 2.6844352532114084e-06,
+      "loss": 0.1697,
+      "step": 15110
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.956752950405443,
+      "learning_rate": 2.6833124807681722e-06,
+      "loss": 0.1704,
+      "step": 15111
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.111203752316407,
+      "learning_rate": 2.6821899067904956e-06,
+      "loss": 0.1614,
+      "step": 15112
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9663017267242348,
+      "learning_rate": 2.6810675313088343e-06,
+      "loss": 0.1689,
+      "step": 15113
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0200424958227927,
+      "learning_rate": 2.6799453543536256e-06,
+      "loss": 0.1824,
+      "step": 15114
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0230622487319652,
+      "learning_rate": 2.678823375955314e-06,
+      "loss": 0.1733,
+      "step": 15115
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9057815915974087,
+      "learning_rate": 2.677701596144331e-06,
+      "loss": 0.1597,
+      "step": 15116
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3771708626699875,
+      "learning_rate": 2.6765800149511088e-06,
+      "loss": 0.1747,
+      "step": 15117
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.141733838833967,
+      "learning_rate": 2.6754586324060637e-06,
+      "loss": 0.1738,
+      "step": 15118
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0691457001073312,
+      "learning_rate": 2.6743374485396212e-06,
+      "loss": 0.1564,
+      "step": 15119
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.082285544415921,
+      "learning_rate": 2.673216463382189e-06,
+      "loss": 0.1542,
+      "step": 15120
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0461481068809784,
+      "learning_rate": 2.67209567696417e-06,
+      "loss": 0.1884,
+      "step": 15121
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5995436617467997,
+      "learning_rate": 2.6709750893159705e-06,
+      "loss": 0.1822,
+      "step": 15122
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1283157762236709,
+      "learning_rate": 2.6698547004679853e-06,
+      "loss": 0.1556,
+      "step": 15123
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2046144453330978,
+      "learning_rate": 2.6687345104506092e-06,
+      "loss": 0.152,
+      "step": 15124
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1382689858156538,
+      "learning_rate": 2.6676145192942194e-06,
+      "loss": 0.1657,
+      "step": 15125
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9200882739660786,
+      "learning_rate": 2.666494727029203e-06,
+      "loss": 0.1579,
+      "step": 15126
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.029441048435042,
+      "learning_rate": 2.6653751336859292e-06,
+      "loss": 0.1602,
+      "step": 15127
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9313604753550834,
+      "learning_rate": 2.6642557392947722e-06,
+      "loss": 0.1747,
+      "step": 15128
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6679795449211656,
+      "learning_rate": 2.663136543886089e-06,
+      "loss": 0.1757,
+      "step": 15129
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0950242043797351,
+      "learning_rate": 2.6620175474902444e-06,
+      "loss": 0.1506,
+      "step": 15130
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1122224424154379,
+      "learning_rate": 2.660898750137585e-06,
+      "loss": 0.1647,
+      "step": 15131
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0499158991950026,
+      "learning_rate": 2.659780151858462e-06,
+      "loss": 0.1696,
+      "step": 15132
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8062873238846927,
+      "learning_rate": 2.658661752683217e-06,
+      "loss": 0.1594,
+      "step": 15133
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2888923453349863,
+      "learning_rate": 2.657543552642189e-06,
+      "loss": 0.1772,
+      "step": 15134
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6795161267833896,
+      "learning_rate": 2.6564255517657044e-06,
+      "loss": 0.1526,
+      "step": 15135
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9699837684539419,
+      "learning_rate": 2.655307750084094e-06,
+      "loss": 0.1669,
+      "step": 15136
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8597368366303423,
+      "learning_rate": 2.6541901476276767e-06,
+      "loss": 0.1964,
+      "step": 15137
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2499292435659783,
+      "learning_rate": 2.653072744426762e-06,
+      "loss": 0.1752,
+      "step": 15138
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.031377290048498,
+      "learning_rate": 2.6519555405116683e-06,
+      "loss": 0.1867,
+      "step": 15139
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8232779349067384,
+      "learning_rate": 2.650838535912692e-06,
+      "loss": 0.1527,
+      "step": 15140
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0424697846561075,
+      "learning_rate": 2.6497217306601365e-06,
+      "loss": 0.1901,
+      "step": 15141
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1272349519057292,
+      "learning_rate": 2.6486051247842935e-06,
+      "loss": 0.2006,
+      "step": 15142
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.801356139954553,
+      "learning_rate": 2.647488718315454e-06,
+      "loss": 0.1599,
+      "step": 15143
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3660574893840298,
+      "learning_rate": 2.6463725112838968e-06,
+      "loss": 0.1848,
+      "step": 15144
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8537934209396962,
+      "learning_rate": 2.645256503719902e-06,
+      "loss": 0.1759,
+      "step": 15145
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1754110127965345,
+      "learning_rate": 2.6441406956537376e-06,
+      "loss": 0.1752,
+      "step": 15146
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7957917440483896,
+      "learning_rate": 2.643025087115676e-06,
+      "loss": 0.1595,
+      "step": 15147
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0393249746050452,
+      "learning_rate": 2.6419096781359698e-06,
+      "loss": 0.193,
+      "step": 15148
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.287822483747699,
+      "learning_rate": 2.6407944687448804e-06,
+      "loss": 0.1492,
+      "step": 15149
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7834088365545043,
+      "learning_rate": 2.639679458972657e-06,
+      "loss": 0.1561,
+      "step": 15150
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.253358728554841,
+      "learning_rate": 2.6385646488495466e-06,
+      "loss": 0.1622,
+      "step": 15151
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4703873122591984,
+      "learning_rate": 2.637450038405782e-06,
+      "loss": 0.1869,
+      "step": 15152
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8843757015526899,
+      "learning_rate": 2.6363356276716046e-06,
+      "loss": 0.1897,
+      "step": 15153
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.082280409237539,
+      "learning_rate": 2.6352214166772363e-06,
+      "loss": 0.1589,
+      "step": 15154
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9841338075804467,
+      "learning_rate": 2.634107405452906e-06,
+      "loss": 0.1781,
+      "step": 15155
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9614771366458215,
+      "learning_rate": 2.6329935940288286e-06,
+      "loss": 0.1715,
+      "step": 15156
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2490440362855904,
+      "learning_rate": 2.6318799824352125e-06,
+      "loss": 0.1555,
+      "step": 15157
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.14649656482353,
+      "learning_rate": 2.6307665707022678e-06,
+      "loss": 0.1588,
+      "step": 15158
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8545450488293231,
+      "learning_rate": 2.629653358860197e-06,
+      "loss": 0.1686,
+      "step": 15159
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7601598800170505,
+      "learning_rate": 2.628540346939198e-06,
+      "loss": 0.1584,
+      "step": 15160
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9209688037861604,
+      "learning_rate": 2.6274275349694544e-06,
+      "loss": 0.1715,
+      "step": 15161
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9961783908841748,
+      "learning_rate": 2.6263149229811592e-06,
+      "loss": 0.17,
+      "step": 15162
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0209861385594814,
+      "learning_rate": 2.6252025110044852e-06,
+      "loss": 0.1623,
+      "step": 15163
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4066217365859481,
+      "learning_rate": 2.6240902990696126e-06,
+      "loss": 0.1661,
+      "step": 15164
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7784038130987568,
+      "learning_rate": 2.6229782872067042e-06,
+      "loss": 0.183,
+      "step": 15165
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2715081433842295,
+      "learning_rate": 2.6218664754459267e-06,
+      "loss": 0.1943,
+      "step": 15166
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.08772074648383,
+      "learning_rate": 2.6207548638174374e-06,
+      "loss": 0.1765,
+      "step": 15167
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6357270332436424,
+      "learning_rate": 2.6196434523513916e-06,
+      "loss": 0.1594,
+      "step": 15168
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4580920469900076,
+      "learning_rate": 2.6185322410779312e-06,
+      "loss": 0.1907,
+      "step": 15169
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5075212882245206,
+      "learning_rate": 2.617421230027205e-06,
+      "loss": 0.1922,
+      "step": 15170
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.527004105493696,
+      "learning_rate": 2.616310419229341e-06,
+      "loss": 0.1636,
+      "step": 15171
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4319575105625417,
+      "learning_rate": 2.615199808714476e-06,
+      "loss": 0.1791,
+      "step": 15172
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.545697916364387,
+      "learning_rate": 2.614089398512735e-06,
+      "loss": 0.1862,
+      "step": 15173
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0144438909510543,
+      "learning_rate": 2.6129791886542323e-06,
+      "loss": 0.1704,
+      "step": 15174
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1318400246614295,
+      "learning_rate": 2.6118691791690865e-06,
+      "loss": 0.1844,
+      "step": 15175
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9289726534512144,
+      "learning_rate": 2.6107593700874056e-06,
+      "loss": 0.1372,
+      "step": 15176
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1153456205612922,
+      "learning_rate": 2.609649761439298e-06,
+      "loss": 0.1693,
+      "step": 15177
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9627905346417706,
+      "learning_rate": 2.6085403532548547e-06,
+      "loss": 0.1608,
+      "step": 15178
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9862059209427578,
+      "learning_rate": 2.6074311455641756e-06,
+      "loss": 0.1863,
+      "step": 15179
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9294597573359676,
+      "learning_rate": 2.60632213839734e-06,
+      "loss": 0.1724,
+      "step": 15180
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.899107516158811,
+      "learning_rate": 2.6052133317844387e-06,
+      "loss": 0.164,
+      "step": 15181
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3657017149652035,
+      "learning_rate": 2.6041047257555384e-06,
+      "loss": 0.1706,
+      "step": 15182
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.480391779689135,
+      "learning_rate": 2.6029963203407195e-06,
+      "loss": 0.1561,
+      "step": 15183
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.596005320977434,
+      "learning_rate": 2.6018881155700403e-06,
+      "loss": 0.1809,
+      "step": 15184
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.873634574087102,
+      "learning_rate": 2.600780111473563e-06,
+      "loss": 0.1833,
+      "step": 15185
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.740690382810113,
+      "learning_rate": 2.5996723080813433e-06,
+      "loss": 0.1672,
+      "step": 15186
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.025542092337058,
+      "learning_rate": 2.5985647054234332e-06,
+      "loss": 0.1653,
+      "step": 15187
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.02212783368464,
+      "learning_rate": 2.597457303529871e-06,
+      "loss": 0.1616,
+      "step": 15188
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.548952516964932,
+      "learning_rate": 2.5963501024307005e-06,
+      "loss": 0.189,
+      "step": 15189
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3022275673627424,
+      "learning_rate": 2.595243102155951e-06,
+      "loss": 0.1571,
+      "step": 15190
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0470981966654658,
+      "learning_rate": 2.594136302735648e-06,
+      "loss": 0.1625,
+      "step": 15191
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5030610793104051,
+      "learning_rate": 2.5930297041998152e-06,
+      "loss": 0.1643,
+      "step": 15192
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8238779173041473,
+      "learning_rate": 2.591923306578471e-06,
+      "loss": 0.1498,
+      "step": 15193
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0571831014548954,
+      "learning_rate": 2.590817109901629e-06,
+      "loss": 0.1685,
+      "step": 15194
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8950212936038725,
+      "learning_rate": 2.589711114199287e-06,
+      "loss": 0.1648,
+      "step": 15195
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0896129472676792,
+      "learning_rate": 2.5886053195014537e-06,
+      "loss": 0.1841,
+      "step": 15196
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1069006521388893,
+      "learning_rate": 2.587499725838116e-06,
+      "loss": 0.1755,
+      "step": 15197
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.305433030656861,
+      "learning_rate": 2.5863943332392703e-06,
+      "loss": 0.1801,
+      "step": 15198
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0925570729945107,
+      "learning_rate": 2.5852891417348933e-06,
+      "loss": 0.148,
+      "step": 15199
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1185001788765023,
+      "learning_rate": 2.5841841513549703e-06,
+      "loss": 0.1791,
+      "step": 15200
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0426008164602087,
+      "learning_rate": 2.583079362129469e-06,
+      "loss": 0.1624,
+      "step": 15201
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.985214885876121,
+      "learning_rate": 2.5819747740883584e-06,
+      "loss": 0.176,
+      "step": 15202
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5640615652580436,
+      "learning_rate": 2.5808703872616014e-06,
+      "loss": 0.1742,
+      "step": 15203
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2960441256428383,
+      "learning_rate": 2.5797662016791556e-06,
+      "loss": 0.153,
+      "step": 15204
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1077933440093228,
+      "learning_rate": 2.5786622173709695e-06,
+      "loss": 0.1634,
+      "step": 15205
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9797945234196949,
+      "learning_rate": 2.5775584343669926e-06,
+      "loss": 0.1636,
+      "step": 15206
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9108251801362562,
+      "learning_rate": 2.576454852697161e-06,
+      "loss": 0.1462,
+      "step": 15207
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8377769081012556,
+      "learning_rate": 2.5753514723914098e-06,
+      "loss": 0.1479,
+      "step": 15208
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0428186371239592,
+      "learning_rate": 2.574248293479671e-06,
+      "loss": 0.1864,
+      "step": 15209
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4175156519849,
+      "learning_rate": 2.573145315991864e-06,
+      "loss": 0.1769,
+      "step": 15210
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1093744001581627,
+      "learning_rate": 2.5720425399579095e-06,
+      "loss": 0.1659,
+      "step": 15211
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.180261198591606,
+      "learning_rate": 2.5709399654077204e-06,
+      "loss": 0.1641,
+      "step": 15212
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.423271440155294,
+      "learning_rate": 2.5698375923712083e-06,
+      "loss": 0.1562,
+      "step": 15213
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1382874794726652,
+      "learning_rate": 2.568735420878268e-06,
+      "loss": 0.1581,
+      "step": 15214
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5057172478871992,
+      "learning_rate": 2.567633450958801e-06,
+      "loss": 0.1454,
+      "step": 15215
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.190316189265075,
+      "learning_rate": 2.5665316826426946e-06,
+      "loss": 0.1836,
+      "step": 15216
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8338288755160992,
+      "learning_rate": 2.5654301159598384e-06,
+      "loss": 0.1512,
+      "step": 15217
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0685886262029858,
+      "learning_rate": 2.564328750940107e-06,
+      "loss": 0.1578,
+      "step": 15218
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0227346756920659,
+      "learning_rate": 2.5632275876133794e-06,
+      "loss": 0.1693,
+      "step": 15219
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0084711151912433,
+      "learning_rate": 2.562126626009522e-06,
+      "loss": 0.1963,
+      "step": 15220
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0826122914850413,
+      "learning_rate": 2.561025866158404e-06,
+      "loss": 0.1828,
+      "step": 15221
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2022230055717948,
+      "learning_rate": 2.5599253080898767e-06,
+      "loss": 0.1574,
+      "step": 15222
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5532502806186688,
+      "learning_rate": 2.558824951833798e-06,
+      "loss": 0.1625,
+      "step": 15223
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2549004826501027,
+      "learning_rate": 2.5577247974200103e-06,
+      "loss": 0.1733,
+      "step": 15224
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0977056913219776,
+      "learning_rate": 2.55662484487836e-06,
+      "loss": 0.1581,
+      "step": 15225
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0576616231310711,
+      "learning_rate": 2.555525094238682e-06,
+      "loss": 0.1915,
+      "step": 15226
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8492059396337894,
+      "learning_rate": 2.5544255455308032e-06,
+      "loss": 0.1763,
+      "step": 15227
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9742120885664564,
+      "learning_rate": 2.5533261987845525e-06,
+      "loss": 0.1606,
+      "step": 15228
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0710840678114584,
+      "learning_rate": 2.552227054029749e-06,
+      "loss": 0.1594,
+      "step": 15229
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1422554132137888,
+      "learning_rate": 2.5511281112962096e-06,
+      "loss": 0.1949,
+      "step": 15230
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9458474201377018,
+      "learning_rate": 2.550029370613738e-06,
+      "loss": 0.1586,
+      "step": 15231
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.362543961001248,
+      "learning_rate": 2.548930832012143e-06,
+      "loss": 0.1677,
+      "step": 15232
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4492155243197786,
+      "learning_rate": 2.5478324955212186e-06,
+      "loss": 0.1613,
+      "step": 15233
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5213788798995165,
+      "learning_rate": 2.5467343611707607e-06,
+      "loss": 0.1697,
+      "step": 15234
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8861019921569067,
+      "learning_rate": 2.545636428990551e-06,
+      "loss": 0.1524,
+      "step": 15235
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0214376716215812,
+      "learning_rate": 2.5445386990103773e-06,
+      "loss": 0.1632,
+      "step": 15236
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4462903199411787,
+      "learning_rate": 2.5434411712600095e-06,
+      "loss": 0.1953,
+      "step": 15237
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.088113114275449,
+      "learning_rate": 2.542343845769222e-06,
+      "loss": 0.1591,
+      "step": 15238
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9099559658873935,
+      "learning_rate": 2.5412467225677774e-06,
+      "loss": 0.1473,
+      "step": 15239
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2840902837839696,
+      "learning_rate": 2.540149801685441e-06,
+      "loss": 0.1561,
+      "step": 15240
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2896921273460544,
+      "learning_rate": 2.5390530831519587e-06,
+      "loss": 0.1589,
+      "step": 15241
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.233985557375335,
+      "learning_rate": 2.5379565669970864e-06,
+      "loss": 0.1873,
+      "step": 15242
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9574844819931793,
+      "learning_rate": 2.5368602532505637e-06,
+      "loss": 0.1647,
+      "step": 15243
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1470495999146328,
+      "learning_rate": 2.535764141942124e-06,
+      "loss": 0.1885,
+      "step": 15244
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7780938539703632,
+      "learning_rate": 2.534668233101505e-06,
+      "loss": 0.1521,
+      "step": 15245
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4385410709228255,
+      "learning_rate": 2.533572526758431e-06,
+      "loss": 0.1688,
+      "step": 15246
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0239198319044815,
+      "learning_rate": 2.5324770229426276e-06,
+      "loss": 0.1568,
+      "step": 15247
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.158525030798928,
+      "learning_rate": 2.5313817216838034e-06,
+      "loss": 0.1787,
+      "step": 15248
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4164698569917662,
+      "learning_rate": 2.530286623011675e-06,
+      "loss": 0.1805,
+      "step": 15249
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.924334069260275,
+      "learning_rate": 2.5291917269559408e-06,
+      "loss": 0.175,
+      "step": 15250
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.894267709513381,
+      "learning_rate": 2.528097033546305e-06,
+      "loss": 0.1582,
+      "step": 15251
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8781609591285007,
+      "learning_rate": 2.527002542812457e-06,
+      "loss": 0.1582,
+      "step": 15252
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.542183566342674,
+      "learning_rate": 2.5259082547840907e-06,
+      "loss": 0.1541,
+      "step": 15253
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.65101598504308,
+      "learning_rate": 2.524814169490881e-06,
+      "loss": 0.1781,
+      "step": 15254
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8947350917702382,
+      "learning_rate": 2.52372028696251e-06,
+      "loss": 0.1769,
+      "step": 15255
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9696609784115476,
+      "learning_rate": 2.5226266072286475e-06,
+      "loss": 0.1721,
+      "step": 15256
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4945032623101626,
+      "learning_rate": 2.521533130318965e-06,
+      "loss": 0.1764,
+      "step": 15257
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0840645808247387,
+      "learning_rate": 2.520439856263115e-06,
+      "loss": 0.1657,
+      "step": 15258
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.211028713539488,
+      "learning_rate": 2.5193467850907583e-06,
+      "loss": 0.1807,
+      "step": 15259
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8345107762432861,
+      "learning_rate": 2.5182539168315435e-06,
+      "loss": 0.1776,
+      "step": 15260
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1090353856444384,
+      "learning_rate": 2.517161251515111e-06,
+      "loss": 0.1765,
+      "step": 15261
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9632084661239485,
+      "learning_rate": 2.516068789171102e-06,
+      "loss": 0.1592,
+      "step": 15262
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.075643005645192,
+      "learning_rate": 2.5149765298291508e-06,
+      "loss": 0.1472,
+      "step": 15263
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9688207506122695,
+      "learning_rate": 2.513884473518885e-06,
+      "loss": 0.1789,
+      "step": 15264
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9284611694669284,
+      "learning_rate": 2.512792620269924e-06,
+      "loss": 0.153,
+      "step": 15265
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2757413064160035,
+      "learning_rate": 2.5117009701118888e-06,
+      "loss": 0.1467,
+      "step": 15266
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9680738544022398,
+      "learning_rate": 2.5106095230743844e-06,
+      "loss": 0.1595,
+      "step": 15267
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0877239157747474,
+      "learning_rate": 2.5095182791870234e-06,
+      "loss": 0.1698,
+      "step": 15268
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2268524665114693,
+      "learning_rate": 2.5084272384793985e-06,
+      "loss": 0.1561,
+      "step": 15269
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.36790180905005,
+      "learning_rate": 2.5073364009811107e-06,
+      "loss": 0.1757,
+      "step": 15270
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0109270739079395,
+      "learning_rate": 2.5062457667217433e-06,
+      "loss": 0.161,
+      "step": 15271
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7892760497472147,
+      "learning_rate": 2.505155335730883e-06,
+      "loss": 0.1766,
+      "step": 15272
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6048470029419701,
+      "learning_rate": 2.5040651080381084e-06,
+      "loss": 0.1753,
+      "step": 15273
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0007696327568976,
+      "learning_rate": 2.5029750836729926e-06,
+      "loss": 0.1457,
+      "step": 15274
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0214733625354664,
+      "learning_rate": 2.501885262665099e-06,
+      "loss": 0.1636,
+      "step": 15275
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5284937432774757,
+      "learning_rate": 2.500795645043994e-06,
+      "loss": 0.1556,
+      "step": 15276
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1343197681393309,
+      "learning_rate": 2.4997062308392304e-06,
+      "loss": 0.1648,
+      "step": 15277
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.261112057405962,
+      "learning_rate": 2.498617020080356e-06,
+      "loss": 0.174,
+      "step": 15278
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3837587883415603,
+      "learning_rate": 2.4975280127969214e-06,
+      "loss": 0.1876,
+      "step": 15279
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9992224350374982,
+      "learning_rate": 2.496439209018461e-06,
+      "loss": 0.1656,
+      "step": 15280
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1779324775338382,
+      "learning_rate": 2.4953506087745107e-06,
+      "loss": 0.1844,
+      "step": 15281
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9112312928556008,
+      "learning_rate": 2.494262212094598e-06,
+      "loss": 0.1557,
+      "step": 15282
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9344392435604768,
+      "learning_rate": 2.4931740190082497e-06,
+      "loss": 0.1636,
+      "step": 15283
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8794078901771586,
+      "learning_rate": 2.4920860295449787e-06,
+      "loss": 0.1581,
+      "step": 15284
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.134697289302735,
+      "learning_rate": 2.4909982437342993e-06,
+      "loss": 0.181,
+      "step": 15285
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0548278226712828,
+      "learning_rate": 2.4899106616057155e-06,
+      "loss": 0.1757,
+      "step": 15286
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.676475877009159,
+      "learning_rate": 2.4888232831887304e-06,
+      "loss": 0.1624,
+      "step": 15287
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7828993584768532,
+      "learning_rate": 2.487736108512836e-06,
+      "loss": 0.1836,
+      "step": 15288
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.020287114138541,
+      "learning_rate": 2.486649137607524e-06,
+      "loss": 0.1586,
+      "step": 15289
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2777417930524249,
+      "learning_rate": 2.485562370502279e-06,
+      "loss": 0.1821,
+      "step": 15290
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0818007196529444,
+      "learning_rate": 2.4844758072265806e-06,
+      "loss": 0.1737,
+      "step": 15291
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4445018677484833,
+      "learning_rate": 2.4833894478098983e-06,
+      "loss": 0.1685,
+      "step": 15292
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.980705724826226,
+      "learning_rate": 2.4823032922817045e-06,
+      "loss": 0.1708,
+      "step": 15293
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9967433356885528,
+      "learning_rate": 2.481217340671457e-06,
+      "loss": 0.1614,
+      "step": 15294
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9265582135084066,
+      "learning_rate": 2.4801315930086147e-06,
+      "loss": 0.1866,
+      "step": 15295
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4209191915362267,
+      "learning_rate": 2.479046049322629e-06,
+      "loss": 0.1861,
+      "step": 15296
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8959157659091018,
+      "learning_rate": 2.4779607096429403e-06,
+      "loss": 0.1587,
+      "step": 15297
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0865709667746095,
+      "learning_rate": 2.4768755739989925e-06,
+      "loss": 0.1525,
+      "step": 15298
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.992508724360173,
+      "learning_rate": 2.475790642420219e-06,
+      "loss": 0.134,
+      "step": 15299
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3318596943250995,
+      "learning_rate": 2.474705914936053e-06,
+      "loss": 0.1858,
+      "step": 15300
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9430394381293802,
+      "learning_rate": 2.473621391575911e-06,
+      "loss": 0.1697,
+      "step": 15301
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8667635085832349,
+      "learning_rate": 2.4725370723692164e-06,
+      "loss": 0.1514,
+      "step": 15302
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5783046632381894,
+      "learning_rate": 2.471452957345376e-06,
+      "loss": 0.166,
+      "step": 15303
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.949053055901867,
+      "learning_rate": 2.4703690465338025e-06,
+      "loss": 0.1693,
+      "step": 15304
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.099586983786691,
+      "learning_rate": 2.469285339963892e-06,
+      "loss": 0.1668,
+      "step": 15305
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.146907604484705,
+      "learning_rate": 2.468201837665043e-06,
+      "loss": 0.1638,
+      "step": 15306
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6703895998487162,
+      "learning_rate": 2.467118539666643e-06,
+      "loss": 0.1559,
+      "step": 15307
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0693170607150535,
+      "learning_rate": 2.4660354459980775e-06,
+      "loss": 0.1422,
+      "step": 15308
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.870622931041626,
+      "learning_rate": 2.4649525566887267e-06,
+      "loss": 0.1573,
+      "step": 15309
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.496630248263848,
+      "learning_rate": 2.4638698717679653e-06,
+      "loss": 0.1695,
+      "step": 15310
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.133075124349076,
+      "learning_rate": 2.462787391265157e-06,
+      "loss": 0.1695,
+      "step": 15311
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0358289180935836,
+      "learning_rate": 2.4617051152096696e-06,
+      "loss": 0.1568,
+      "step": 15312
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1225526395435053,
+      "learning_rate": 2.4606230436308554e-06,
+      "loss": 0.1659,
+      "step": 15313
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0468975380099528,
+      "learning_rate": 2.4595411765580645e-06,
+      "loss": 0.1858,
+      "step": 15314
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9816839310890371,
+      "learning_rate": 2.4584595140206457e-06,
+      "loss": 0.1598,
+      "step": 15315
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.085821885409933,
+      "learning_rate": 2.4573780560479387e-06,
+      "loss": 0.1631,
+      "step": 15316
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0087642673161445,
+      "learning_rate": 2.4562968026692803e-06,
+      "loss": 0.1661,
+      "step": 15317
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0371578322321766,
+      "learning_rate": 2.4552157539139944e-06,
+      "loss": 0.1793,
+      "step": 15318
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9130220149814539,
+      "learning_rate": 2.45413490981141e-06,
+      "loss": 0.171,
+      "step": 15319
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.290014807266559,
+      "learning_rate": 2.45305427039084e-06,
+      "loss": 0.1791,
+      "step": 15320
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9635165312055891,
+      "learning_rate": 2.4519738356816015e-06,
+      "loss": 0.1666,
+      "step": 15321
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2303516824006075,
+      "learning_rate": 2.450893605712996e-06,
+      "loss": 0.1817,
+      "step": 15322
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3058539158515092,
+      "learning_rate": 2.449813580514332e-06,
+      "loss": 0.1724,
+      "step": 15323
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9524281611944755,
+      "learning_rate": 2.4487337601148975e-06,
+      "loss": 0.1874,
+      "step": 15324
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.073158181893844,
+      "learning_rate": 2.447654144543986e-06,
+      "loss": 0.1632,
+      "step": 15325
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9513488743035756,
+      "learning_rate": 2.446574733830882e-06,
+      "loss": 0.1662,
+      "step": 15326
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8947551127078637,
+      "learning_rate": 2.4454955280048688e-06,
+      "loss": 0.1641,
+      "step": 15327
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8500710551120902,
+      "learning_rate": 2.4444165270952126e-06,
+      "loss": 0.1786,
+      "step": 15328
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6713358847496242,
+      "learning_rate": 2.4433377311311878e-06,
+      "loss": 0.1814,
+      "step": 15329
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3139186587490201,
+      "learning_rate": 2.4422591401420537e-06,
+      "loss": 0.1675,
+      "step": 15330
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3539006655141135,
+      "learning_rate": 2.4411807541570643e-06,
+      "loss": 0.198,
+      "step": 15331
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8974215267445775,
+      "learning_rate": 2.440102573205477e-06,
+      "loss": 0.1586,
+      "step": 15332
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1506955324100798,
+      "learning_rate": 2.4390245973165316e-06,
+      "loss": 0.1703,
+      "step": 15333
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2501650244300122,
+      "learning_rate": 2.4379468265194707e-06,
+      "loss": 0.1611,
+      "step": 15334
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.033876510720954,
+      "learning_rate": 2.4368692608435294e-06,
+      "loss": 0.1792,
+      "step": 15335
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7951661193570492,
+      "learning_rate": 2.4357919003179396e-06,
+      "loss": 0.1471,
+      "step": 15336
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9520345764506051,
+      "learning_rate": 2.434714744971919e-06,
+      "loss": 0.1776,
+      "step": 15337
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0833822675819067,
+      "learning_rate": 2.4336377948346912e-06,
+      "loss": 0.1885,
+      "step": 15338
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1029213485645695,
+      "learning_rate": 2.432561049935462e-06,
+      "loss": 0.1834,
+      "step": 15339
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4361044589948972,
+      "learning_rate": 2.4314845103034456e-06,
+      "loss": 0.176,
+      "step": 15340
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8442285067210624,
+      "learning_rate": 2.4304081759678357e-06,
+      "loss": 0.1595,
+      "step": 15341
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9579393533005156,
+      "learning_rate": 2.429332046957832e-06,
+      "loss": 0.1709,
+      "step": 15342
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9511031711141813,
+      "learning_rate": 2.4282561233026236e-06,
+      "loss": 0.2082,
+      "step": 15343
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.189125025522632,
+      "learning_rate": 2.4271804050313984e-06,
+      "loss": 0.1759,
+      "step": 15344
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8696093166445392,
+      "learning_rate": 2.42610489217333e-06,
+      "loss": 0.153,
+      "step": 15345
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7661160288872305,
+      "learning_rate": 2.4250295847575967e-06,
+      "loss": 0.1701,
+      "step": 15346
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0094608338108795,
+      "learning_rate": 2.4239544828133632e-06,
+      "loss": 0.1876,
+      "step": 15347
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1947365085635253,
+      "learning_rate": 2.422879586369791e-06,
+      "loss": 0.172,
+      "step": 15348
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.02607706559597,
+      "learning_rate": 2.421804895456039e-06,
+      "loss": 0.1691,
+      "step": 15349
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1105200321284465,
+      "learning_rate": 2.420730410101255e-06,
+      "loss": 0.1751,
+      "step": 15350
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1219152032536965,
+      "learning_rate": 2.419656130334588e-06,
+      "loss": 0.1913,
+      "step": 15351
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4192179540898853,
+      "learning_rate": 2.4185820561851747e-06,
+      "loss": 0.1603,
+      "step": 15352
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6876340282880435,
+      "learning_rate": 2.417508187682156e-06,
+      "loss": 0.1619,
+      "step": 15353
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9298730731346014,
+      "learning_rate": 2.4164345248546517e-06,
+      "loss": 0.1627,
+      "step": 15354
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0761477246483009,
+      "learning_rate": 2.415361067731793e-06,
+      "loss": 0.1488,
+      "step": 15355
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9804762037101074,
+      "learning_rate": 2.41428781634269e-06,
+      "loss": 0.1548,
+      "step": 15356
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0872119575860812,
+      "learning_rate": 2.413214770716462e-06,
+      "loss": 0.1627,
+      "step": 15357
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9356716591209069,
+      "learning_rate": 2.412141930882208e-06,
+      "loss": 0.1662,
+      "step": 15358
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5485828606207372,
+      "learning_rate": 2.4110692968690364e-06,
+      "loss": 0.1539,
+      "step": 15359
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2746909556741446,
+      "learning_rate": 2.409996868706036e-06,
+      "loss": 0.1875,
+      "step": 15360
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.020526020568585,
+      "learning_rate": 2.4089246464222995e-06,
+      "loss": 0.1603,
+      "step": 15361
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1005090547533583,
+      "learning_rate": 2.4078526300469097e-06,
+      "loss": 0.1756,
+      "step": 15362
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0681842201569682,
+      "learning_rate": 2.4067808196089493e-06,
+      "loss": 0.1727,
+      "step": 15363
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3422597996378893,
+      "learning_rate": 2.4057092151374885e-06,
+      "loss": 0.1712,
+      "step": 15364
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9132433065779902,
+      "learning_rate": 2.40463781666159e-06,
+      "loss": 0.1484,
+      "step": 15365
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0703530978743458,
+      "learning_rate": 2.403566624210324e-06,
+      "loss": 0.1645,
+      "step": 15366
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0893874090214328,
+      "learning_rate": 2.4024956378127396e-06,
+      "loss": 0.1551,
+      "step": 15367
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.896536719043762,
+      "learning_rate": 2.401424857497889e-06,
+      "loss": 0.1692,
+      "step": 15368
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1534646592762958,
+      "learning_rate": 2.400354283294819e-06,
+      "loss": 0.1708,
+      "step": 15369
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.20144166772448,
+      "learning_rate": 2.399283915232571e-06,
+      "loss": 0.1728,
+      "step": 15370
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8001018929783066,
+      "learning_rate": 2.398213753340174e-06,
+      "loss": 0.159,
+      "step": 15371
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1613880996230987,
+      "learning_rate": 2.3971437976466604e-06,
+      "loss": 0.1494,
+      "step": 15372
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0741129775209934,
+      "learning_rate": 2.3960740481810475e-06,
+      "loss": 0.1668,
+      "step": 15373
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.998500376985148,
+      "learning_rate": 2.3950045049723593e-06,
+      "loss": 0.1767,
+      "step": 15374
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.446899401008942,
+      "learning_rate": 2.3939351680495994e-06,
+      "loss": 0.1734,
+      "step": 15375
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9632523661077028,
+      "learning_rate": 2.392866037441781e-06,
+      "loss": 0.1463,
+      "step": 15376
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0273977255365927,
+      "learning_rate": 2.3917971131778982e-06,
+      "loss": 0.1737,
+      "step": 15377
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2251126007220232,
+      "learning_rate": 2.3907283952869485e-06,
+      "loss": 0.1656,
+      "step": 15378
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.41623658769232,
+      "learning_rate": 2.389659883797921e-06,
+      "loss": 0.1625,
+      "step": 15379
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2970703730778923,
+      "learning_rate": 2.3885915787398016e-06,
+      "loss": 0.1631,
+      "step": 15380
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.010114637458024,
+      "learning_rate": 2.3875234801415626e-06,
+      "loss": 0.1665,
+      "step": 15381
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2164200120028743,
+      "learning_rate": 2.3864555880321828e-06,
+      "loss": 0.1599,
+      "step": 15382
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4672333643271673,
+      "learning_rate": 2.3853879024406244e-06,
+      "loss": 0.1615,
+      "step": 15383
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.120252041381067,
+      "learning_rate": 2.3843204233958463e-06,
+      "loss": 0.1483,
+      "step": 15384
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0045893183847037,
+      "learning_rate": 2.3832531509268076e-06,
+      "loss": 0.1928,
+      "step": 15385
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1425096056549606,
+      "learning_rate": 2.382186085062457e-06,
+      "loss": 0.1675,
+      "step": 15386
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3341631701663308,
+      "learning_rate": 2.3811192258317416e-06,
+      "loss": 0.1693,
+      "step": 15387
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.257601081796507,
+      "learning_rate": 2.3800525732635946e-06,
+      "loss": 0.1686,
+      "step": 15388
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.000073423100724,
+      "learning_rate": 2.3789861273869553e-06,
+      "loss": 0.1693,
+      "step": 15389
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0257732649799796,
+      "learning_rate": 2.3779198882307443e-06,
+      "loss": 0.1654,
+      "step": 15390
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5535326206303897,
+      "learning_rate": 2.3768538558238895e-06,
+      "loss": 0.1524,
+      "step": 15391
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2975656393098935,
+      "learning_rate": 2.375788030195303e-06,
+      "loss": 0.1834,
+      "step": 15392
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3739833482123904,
+      "learning_rate": 2.3747224113738985e-06,
+      "loss": 0.146,
+      "step": 15393
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9476331047559899,
+      "learning_rate": 2.373656999388576e-06,
+      "loss": 0.1535,
+      "step": 15394
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4355877710988458,
+      "learning_rate": 2.3725917942682397e-06,
+      "loss": 0.1889,
+      "step": 15395
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9183881432091104,
+      "learning_rate": 2.3715267960417798e-06,
+      "loss": 0.1549,
+      "step": 15396
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3000475050786378,
+      "learning_rate": 2.370462004738091e-06,
+      "loss": 0.1591,
+      "step": 15397
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8743533785508302,
+      "learning_rate": 2.3693974203860472e-06,
+      "loss": 0.1637,
+      "step": 15398
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.416431709572786,
+      "learning_rate": 2.3683330430145333e-06,
+      "loss": 0.1899,
+      "step": 15399
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5467261188083323,
+      "learning_rate": 2.367268872652416e-06,
+      "loss": 0.156,
+      "step": 15400
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9294656354998612,
+      "learning_rate": 2.366204909328559e-06,
+      "loss": 0.1539,
+      "step": 15401
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8225038847893258,
+      "learning_rate": 2.3651411530718272e-06,
+      "loss": 0.1698,
+      "step": 15402
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.262942076782041,
+      "learning_rate": 2.36407760391107e-06,
+      "loss": 0.1768,
+      "step": 15403
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0086611381985109,
+      "learning_rate": 2.3630142618751405e-06,
+      "loss": 0.1811,
+      "step": 15404
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9710681600261067,
+      "learning_rate": 2.3619511269928784e-06,
+      "loss": 0.1837,
+      "step": 15405
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3854709074678964,
+      "learning_rate": 2.360888199293128e-06,
+      "loss": 0.1789,
+      "step": 15406
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9299119097883016,
+      "learning_rate": 2.3598254788047136e-06,
+      "loss": 0.1539,
+      "step": 15407
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.011282534568099,
+      "learning_rate": 2.358762965556467e-06,
+      "loss": 0.1656,
+      "step": 15408
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2551793735265222,
+      "learning_rate": 2.3577006595772032e-06,
+      "loss": 0.1673,
+      "step": 15409
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0943491917388624,
+      "learning_rate": 2.3566385608957443e-06,
+      "loss": 0.1731,
+      "step": 15410
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3107629112380317,
+      "learning_rate": 2.355576669540893e-06,
+      "loss": 0.1575,
+      "step": 15411
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8515648899316381,
+      "learning_rate": 2.354514985541456e-06,
+      "loss": 0.1599,
+      "step": 15412
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.293967723582203,
+      "learning_rate": 2.353453508926232e-06,
+      "loss": 0.1566,
+      "step": 15413
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9095448026517428,
+      "learning_rate": 2.3523922397240163e-06,
+      "loss": 0.1534,
+      "step": 15414
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8032411398733617,
+      "learning_rate": 2.3513311779635904e-06,
+      "loss": 0.1514,
+      "step": 15415
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1203727977935276,
+      "learning_rate": 2.3502703236737412e-06,
+      "loss": 0.1597,
+      "step": 15416
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3115090337752493,
+      "learning_rate": 2.3492096768832417e-06,
+      "loss": 0.1593,
+      "step": 15417
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.154189297651346,
+      "learning_rate": 2.348149237620858e-06,
+      "loss": 0.1696,
+      "step": 15418
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.738661265681747,
+      "learning_rate": 2.3470890059153616e-06,
+      "loss": 0.16,
+      "step": 15419
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8918477045125891,
+      "learning_rate": 2.3460289817955063e-06,
+      "loss": 0.154,
+      "step": 15420
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7360933710102129,
+      "learning_rate": 2.3449691652900464e-06,
+      "loss": 0.1658,
+      "step": 15421
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0560603543571885,
+      "learning_rate": 2.3439095564277305e-06,
+      "loss": 0.1629,
+      "step": 15422
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9505114766577443,
+      "learning_rate": 2.342850155237303e-06,
+      "loss": 0.1818,
+      "step": 15423
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0034429408534218,
+      "learning_rate": 2.341790961747494e-06,
+      "loss": 0.1752,
+      "step": 15424
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.088320359735791,
+      "learning_rate": 2.340731975987042e-06,
+      "loss": 0.1705,
+      "step": 15425
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9125680830719255,
+      "learning_rate": 2.3396731979846634e-06,
+      "loss": 0.1506,
+      "step": 15426
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.830417739612693,
+      "learning_rate": 2.3386146277690858e-06,
+      "loss": 0.1519,
+      "step": 15427
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.849114943252066,
+      "learning_rate": 2.3375562653690166e-06,
+      "loss": 0.1603,
+      "step": 15428
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3629818306238608,
+      "learning_rate": 2.336498110813168e-06,
+      "loss": 0.1476,
+      "step": 15429
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.823546883376828,
+      "learning_rate": 2.3354401641302395e-06,
+      "loss": 0.1539,
+      "step": 15430
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.978023459235194,
+      "learning_rate": 2.3343824253489277e-06,
+      "loss": 0.154,
+      "step": 15431
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0018849611636187,
+      "learning_rate": 2.333324894497927e-06,
+      "loss": 0.1513,
+      "step": 15432
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9875086947887675,
+      "learning_rate": 2.332267571605924e-06,
+      "loss": 0.1709,
+      "step": 15433
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5938673758138624,
+      "learning_rate": 2.331210456701597e-06,
+      "loss": 0.1562,
+      "step": 15434
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4053685858854916,
+      "learning_rate": 2.330153549813615e-06,
+      "loss": 0.1774,
+      "step": 15435
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.934543246171554,
+      "learning_rate": 2.329096850970656e-06,
+      "loss": 0.1549,
+      "step": 15436
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3750873777918644,
+      "learning_rate": 2.3280403602013735e-06,
+      "loss": 0.1707,
+      "step": 15437
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.974035012887203,
+      "learning_rate": 2.326984077534431e-06,
+      "loss": 0.1733,
+      "step": 15438
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.003644320415384,
+      "learning_rate": 2.3259280029984775e-06,
+      "loss": 0.1602,
+      "step": 15439
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3928202333175244,
+      "learning_rate": 2.324872136622164e-06,
+      "loss": 0.1667,
+      "step": 15440
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.158804952368038,
+      "learning_rate": 2.3238164784341242e-06,
+      "loss": 0.1555,
+      "step": 15441
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9716905368598698,
+      "learning_rate": 2.3227610284629985e-06,
+      "loss": 0.1649,
+      "step": 15442
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4020372115262854,
+      "learning_rate": 2.3217057867374114e-06,
+      "loss": 0.1764,
+      "step": 15443
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0717520983846267,
+      "learning_rate": 2.32065075328599e-06,
+      "loss": 0.1576,
+      "step": 15444
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1663936158767128,
+      "learning_rate": 2.319595928137349e-06,
+      "loss": 0.1636,
+      "step": 15445
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1732575843595243,
+      "learning_rate": 2.318541311320105e-06,
+      "loss": 0.1934,
+      "step": 15446
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9486028250086229,
+      "learning_rate": 2.317486902862859e-06,
+      "loss": 0.1571,
+      "step": 15447
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1498421923735032,
+      "learning_rate": 2.3164327027942147e-06,
+      "loss": 0.1772,
+      "step": 15448
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9350401407713013,
+      "learning_rate": 2.3153787111427673e-06,
+      "loss": 0.1614,
+      "step": 15449
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5593140843812545,
+      "learning_rate": 2.3143249279371085e-06,
+      "loss": 0.1494,
+      "step": 15450
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3508993582556552,
+      "learning_rate": 2.313271353205818e-06,
+      "loss": 0.1521,
+      "step": 15451
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.787726912758565,
+      "learning_rate": 2.3122179869774784e-06,
+      "loss": 0.1542,
+      "step": 15452
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8663480465259822,
+      "learning_rate": 2.311164829280661e-06,
+      "loss": 0.1509,
+      "step": 15453
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2976235796846125,
+      "learning_rate": 2.3101118801439283e-06,
+      "loss": 0.1611,
+      "step": 15454
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6973010981091752,
+      "learning_rate": 2.3090591395958485e-06,
+      "loss": 0.1373,
+      "step": 15455
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1117049975822124,
+      "learning_rate": 2.3080066076649697e-06,
+      "loss": 0.1694,
+      "step": 15456
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1801660374439589,
+      "learning_rate": 2.3069542843798476e-06,
+      "loss": 0.1734,
+      "step": 15457
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9972892773542359,
+      "learning_rate": 2.3059021697690254e-06,
+      "loss": 0.1578,
+      "step": 15458
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9699548310098502,
+      "learning_rate": 2.3048502638610427e-06,
+      "loss": 0.164,
+      "step": 15459
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9651201995600788,
+      "learning_rate": 2.3037985666844297e-06,
+      "loss": 0.1616,
+      "step": 15460
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9133523025137816,
+      "learning_rate": 2.3027470782677173e-06,
+      "loss": 0.1678,
+      "step": 15461
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0927839593337292,
+      "learning_rate": 2.3016957986394228e-06,
+      "loss": 0.17,
+      "step": 15462
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8773806402542051,
+      "learning_rate": 2.3006447278280676e-06,
+      "loss": 0.1647,
+      "step": 15463
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3811845661656832,
+      "learning_rate": 2.299593865862155e-06,
+      "loss": 0.1737,
+      "step": 15464
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9285844656456148,
+      "learning_rate": 2.2985432127701945e-06,
+      "loss": 0.1498,
+      "step": 15465
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8726490367071329,
+      "learning_rate": 2.2974927685806848e-06,
+      "loss": 0.1588,
+      "step": 15466
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.927809791960326,
+      "learning_rate": 2.296442533322121e-06,
+      "loss": 0.1699,
+      "step": 15467
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9565631359875368,
+      "learning_rate": 2.2953925070229865e-06,
+      "loss": 0.1768,
+      "step": 15468
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.552914352956403,
+      "learning_rate": 2.2943426897117672e-06,
+      "loss": 0.1757,
+      "step": 15469
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2338495932171418,
+      "learning_rate": 2.2932930814169383e-06,
+      "loss": 0.1785,
+      "step": 15470
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3102501076279853,
+      "learning_rate": 2.292243682166967e-06,
+      "loss": 0.16,
+      "step": 15471
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.131636846718695,
+      "learning_rate": 2.291194491990324e-06,
+      "loss": 0.1681,
+      "step": 15472
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0837521584107503,
+      "learning_rate": 2.2901455109154626e-06,
+      "loss": 0.1696,
+      "step": 15473
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1508754181295,
+      "learning_rate": 2.2890967389708396e-06,
+      "loss": 0.1715,
+      "step": 15474
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9192566541574264,
+      "learning_rate": 2.2880481761849037e-06,
+      "loss": 0.1358,
+      "step": 15475
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.054010923229116,
+      "learning_rate": 2.286999822586099e-06,
+      "loss": 0.1643,
+      "step": 15476
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9484038897437859,
+      "learning_rate": 2.285951678202857e-06,
+      "loss": 0.1605,
+      "step": 15477
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8489375942887908,
+      "learning_rate": 2.2849037430636135e-06,
+      "loss": 0.1565,
+      "step": 15478
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4204227390337094,
+      "learning_rate": 2.2838560171967906e-06,
+      "loss": 0.1644,
+      "step": 15479
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9216879840577646,
+      "learning_rate": 2.28280850063081e-06,
+      "loss": 0.1564,
+      "step": 15480
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9092868562804686,
+      "learning_rate": 2.281761193394083e-06,
+      "loss": 0.1472,
+      "step": 15481
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.920977308430054,
+      "learning_rate": 2.2807140955150198e-06,
+      "loss": 0.1597,
+      "step": 15482
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9738559801843604,
+      "learning_rate": 2.2796672070220217e-06,
+      "loss": 0.1674,
+      "step": 15483
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3205100337184514,
+      "learning_rate": 2.27862052794349e-06,
+      "loss": 0.1505,
+      "step": 15484
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0861368779297695,
+      "learning_rate": 2.27757405830781e-06,
+      "loss": 0.1811,
+      "step": 15485
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8969086125838965,
+      "learning_rate": 2.276527798143372e-06,
+      "loss": 0.1669,
+      "step": 15486
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9619151407615854,
+      "learning_rate": 2.275481747478554e-06,
+      "loss": 0.1779,
+      "step": 15487
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2675080890008574,
+      "learning_rate": 2.2744359063417276e-06,
+      "loss": 0.1436,
+      "step": 15488
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.945079506316101,
+      "learning_rate": 2.2733902747612656e-06,
+      "loss": 0.1646,
+      "step": 15489
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6382623359251274,
+      "learning_rate": 2.2723448527655267e-06,
+      "loss": 0.1597,
+      "step": 15490
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.530052373076095,
+      "learning_rate": 2.27129964038287e-06,
+      "loss": 0.1808,
+      "step": 15491
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.0969313221348065,
+      "learning_rate": 2.2702546376416467e-06,
+      "loss": 0.1987,
+      "step": 15492
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.909649025671053,
+      "learning_rate": 2.269209844570206e-06,
+      "loss": 0.1699,
+      "step": 15493
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3532387603007856,
+      "learning_rate": 2.268165261196882e-06,
+      "loss": 0.1581,
+      "step": 15494
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7315958100266694,
+      "learning_rate": 2.267120887550015e-06,
+      "loss": 0.16,
+      "step": 15495
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9825168423437397,
+      "learning_rate": 2.2660767236579275e-06,
+      "loss": 0.1693,
+      "step": 15496
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8804275441419434,
+      "learning_rate": 2.265032769548948e-06,
+      "loss": 0.1865,
+      "step": 15497
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3730261395856402,
+      "learning_rate": 2.26398902525139e-06,
+      "loss": 0.1503,
+      "step": 15498
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0981105723529896,
+      "learning_rate": 2.2629454907935687e-06,
+      "loss": 0.1803,
+      "step": 15499
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1567464441134347,
+      "learning_rate": 2.2619021662037855e-06,
+      "loss": 0.1713,
+      "step": 15500
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.029235331388223,
+      "learning_rate": 2.2608590515103425e-06,
+      "loss": 0.1663,
+      "step": 15501
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2843708418108912,
+      "learning_rate": 2.2598161467415357e-06,
+      "loss": 0.1569,
+      "step": 15502
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1181951782321329,
+      "learning_rate": 2.2587734519256556e-06,
+      "loss": 0.1742,
+      "step": 15503
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0213400485525141,
+      "learning_rate": 2.257730967090982e-06,
+      "loss": 0.1643,
+      "step": 15504
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5573104906756543,
+      "learning_rate": 2.2566886922657917e-06,
+      "loss": 0.1647,
+      "step": 15505
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8940242870876924,
+      "learning_rate": 2.2556466274783596e-06,
+      "loss": 0.1733,
+      "step": 15506
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2743541925464599,
+      "learning_rate": 2.2546047727569475e-06,
+      "loss": 0.1742,
+      "step": 15507
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.939656053009,
+      "learning_rate": 2.253563128129819e-06,
+      "loss": 0.1604,
+      "step": 15508
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5790346633369725,
+      "learning_rate": 2.252521693625228e-06,
+      "loss": 0.1426,
+      "step": 15509
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7542193003523812,
+      "learning_rate": 2.2514804692714264e-06,
+      "loss": 0.1828,
+      "step": 15510
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0661220233266782,
+      "learning_rate": 2.2504394550966513e-06,
+      "loss": 0.1616,
+      "step": 15511
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2734750540730406,
+      "learning_rate": 2.249398651129148e-06,
+      "loss": 0.1979,
+      "step": 15512
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3328836438750213,
+      "learning_rate": 2.2483580573971396e-06,
+      "loss": 0.1572,
+      "step": 15513
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0606100818038617,
+      "learning_rate": 2.2473176739288603e-06,
+      "loss": 0.1738,
+      "step": 15514
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0724834883126513,
+      "learning_rate": 2.246277500752524e-06,
+      "loss": 0.1526,
+      "step": 15515
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0180285929417798,
+      "learning_rate": 2.245237537896351e-06,
+      "loss": 0.1372,
+      "step": 15516
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3299070440761946,
+      "learning_rate": 2.2441977853885454e-06,
+      "loss": 0.1568,
+      "step": 15517
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8112916247301079,
+      "learning_rate": 2.2431582432573127e-06,
+      "loss": 0.1701,
+      "step": 15518
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9341991066548048,
+      "learning_rate": 2.2421189115308506e-06,
+      "loss": 0.1682,
+      "step": 15519
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9881390734593312,
+      "learning_rate": 2.241079790237355e-06,
+      "loss": 0.1502,
+      "step": 15520
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0181825157115165,
+      "learning_rate": 2.2400408794050045e-06,
+      "loss": 0.1626,
+      "step": 15521
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4345164270616522,
+      "learning_rate": 2.2390021790619863e-06,
+      "loss": 0.1739,
+      "step": 15522
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9341993239393256,
+      "learning_rate": 2.237963689236472e-06,
+      "loss": 0.1921,
+      "step": 15523
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9828588397695067,
+      "learning_rate": 2.2369254099566283e-06,
+      "loss": 0.1603,
+      "step": 15524
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2853311671086198,
+      "learning_rate": 2.2358873412506254e-06,
+      "loss": 0.1998,
+      "step": 15525
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9950469959633863,
+      "learning_rate": 2.2348494831466127e-06,
+      "loss": 0.1758,
+      "step": 15526
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3029094611737277,
+      "learning_rate": 2.2338118356727466e-06,
+      "loss": 0.1748,
+      "step": 15527
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.810663812794333,
+      "learning_rate": 2.2327743988571738e-06,
+      "loss": 0.1567,
+      "step": 15528
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9306164030658635,
+      "learning_rate": 2.231737172728037e-06,
+      "loss": 0.1562,
+      "step": 15529
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5012657439900254,
+      "learning_rate": 2.2307001573134646e-06,
+      "loss": 0.1914,
+      "step": 15530
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0339311943769462,
+      "learning_rate": 2.2296633526415924e-06,
+      "loss": 0.1805,
+      "step": 15531
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8261741916565691,
+      "learning_rate": 2.2286267587405376e-06,
+      "loss": 0.182,
+      "step": 15532
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9142063944559767,
+      "learning_rate": 2.227590375638423e-06,
+      "loss": 0.1617,
+      "step": 15533
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3244697895988153,
+      "learning_rate": 2.226554203363357e-06,
+      "loss": 0.154,
+      "step": 15534
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3274544121566074,
+      "learning_rate": 2.225518241943446e-06,
+      "loss": 0.1635,
+      "step": 15535
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9677118097166073,
+      "learning_rate": 2.2244824914067932e-06,
+      "loss": 0.1533,
+      "step": 15536
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3979123867254049,
+      "learning_rate": 2.2234469517814937e-06,
+      "loss": 0.1597,
+      "step": 15537
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.757339040284125,
+      "learning_rate": 2.2224116230956326e-06,
+      "loss": 0.1425,
+      "step": 15538
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0385628676593655,
+      "learning_rate": 2.2213765053772984e-06,
+      "loss": 0.1956,
+      "step": 15539
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1618767087057638,
+      "learning_rate": 2.220341598654565e-06,
+      "loss": 0.1637,
+      "step": 15540
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0568778997025638,
+      "learning_rate": 2.2193069029555035e-06,
+      "loss": 0.1813,
+      "step": 15541
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8628483075554302,
+      "learning_rate": 2.2182724183081837e-06,
+      "loss": 0.1805,
+      "step": 15542
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6281039310208947,
+      "learning_rate": 2.217238144740662e-06,
+      "loss": 0.1737,
+      "step": 15543
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0431107402561968,
+      "learning_rate": 2.216204082280995e-06,
+      "loss": 0.1733,
+      "step": 15544
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8442010781512328,
+      "learning_rate": 2.215170230957231e-06,
+      "loss": 0.1601,
+      "step": 15545
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8923269661973011,
+      "learning_rate": 2.2141365907974176e-06,
+      "loss": 0.1776,
+      "step": 15546
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2454688797031142,
+      "learning_rate": 2.213103161829586e-06,
+      "loss": 0.1488,
+      "step": 15547
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0118447767791443,
+      "learning_rate": 2.212069944081774e-06,
+      "loss": 0.1654,
+      "step": 15548
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.570171035100308,
+      "learning_rate": 2.2110369375820016e-06,
+      "loss": 0.1827,
+      "step": 15549
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.126571649081451,
+      "learning_rate": 2.2100041423582954e-06,
+      "loss": 0.1923,
+      "step": 15550
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8934819284786086,
+      "learning_rate": 2.208971558438664e-06,
+      "loss": 0.1508,
+      "step": 15551
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9617945397982444,
+      "learning_rate": 2.2079391858511214e-06,
+      "loss": 0.174,
+      "step": 15552
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.322771889638723,
+      "learning_rate": 2.2069070246236658e-06,
+      "loss": 0.1791,
+      "step": 15553
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9423969982106858,
+      "learning_rate": 2.2058750747842974e-06,
+      "loss": 0.1592,
+      "step": 15554
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1030627892681681,
+      "learning_rate": 2.2048433363610077e-06,
+      "loss": 0.1518,
+      "step": 15555
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1999783035530458,
+      "learning_rate": 2.203811809381785e-06,
+      "loss": 0.1547,
+      "step": 15556
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0678737575178097,
+      "learning_rate": 2.2027804938746087e-06,
+      "loss": 0.154,
+      "step": 15557
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2888427274702388,
+      "learning_rate": 2.201749389867448e-06,
+      "loss": 0.1794,
+      "step": 15558
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8089560589472962,
+      "learning_rate": 2.200718497388279e-06,
+      "loss": 0.1611,
+      "step": 15559
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0591013646114105,
+      "learning_rate": 2.199687816465058e-06,
+      "loss": 0.1696,
+      "step": 15560
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0265503614760638,
+      "learning_rate": 2.1986573471257456e-06,
+      "loss": 0.1726,
+      "step": 15561
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0804107453865288,
+      "learning_rate": 2.1976270893982934e-06,
+      "loss": 0.1647,
+      "step": 15562
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9829531661161499,
+      "learning_rate": 2.19659704331065e-06,
+      "loss": 0.1648,
+      "step": 15563
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1559582416470937,
+      "learning_rate": 2.19556720889075e-06,
+      "loss": 0.1708,
+      "step": 15564
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.511957859694274,
+      "learning_rate": 2.194537586166532e-06,
+      "loss": 0.1703,
+      "step": 15565
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9399930014089914,
+      "learning_rate": 2.1935081751659214e-06,
+      "loss": 0.1495,
+      "step": 15566
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1744683179840691,
+      "learning_rate": 2.192478975916844e-06,
+      "loss": 0.1646,
+      "step": 15567
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.052596780486219,
+      "learning_rate": 2.191449988447213e-06,
+      "loss": 0.1721,
+      "step": 15568
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9301386245281534,
+      "learning_rate": 2.1904212127849455e-06,
+      "loss": 0.1756,
+      "step": 15569
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2471355427079205,
+      "learning_rate": 2.18939264895794e-06,
+      "loss": 0.146,
+      "step": 15570
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8480913113661026,
+      "learning_rate": 2.188364296994101e-06,
+      "loss": 0.1825,
+      "step": 15571
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8866841429424297,
+      "learning_rate": 2.1873361569213204e-06,
+      "loss": 0.1692,
+      "step": 15572
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9416588430775136,
+      "learning_rate": 2.186308228767492e-06,
+      "loss": 0.1639,
+      "step": 15573
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.925607189488996,
+      "learning_rate": 2.1852805125604924e-06,
+      "loss": 0.1766,
+      "step": 15574
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0407514437395915,
+      "learning_rate": 2.184253008328199e-06,
+      "loss": 0.1669,
+      "step": 15575
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9420561629019543,
+      "learning_rate": 2.1832257160984873e-06,
+      "loss": 0.1662,
+      "step": 15576
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.22011044368606,
+      "learning_rate": 2.182198635899215e-06,
+      "loss": 0.166,
+      "step": 15577
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6176134255385397,
+      "learning_rate": 2.18117176775825e-06,
+      "loss": 0.1818,
+      "step": 15578
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8881455447897848,
+      "learning_rate": 2.18014511170344e-06,
+      "loss": 0.1482,
+      "step": 15579
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.196847111229772,
+      "learning_rate": 2.179118667762635e-06,
+      "loss": 0.1732,
+      "step": 15580
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7944044035893635,
+      "learning_rate": 2.178092435963678e-06,
+      "loss": 0.1418,
+      "step": 15581
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.813756883174713,
+      "learning_rate": 2.177066416334409e-06,
+      "loss": 0.163,
+      "step": 15582
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7979800318997707,
+      "learning_rate": 2.176040608902652e-06,
+      "loss": 0.1461,
+      "step": 15583
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1091856405851366,
+      "learning_rate": 2.175015013696238e-06,
+      "loss": 0.155,
+      "step": 15584
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0806486704666542,
+      "learning_rate": 2.173989630742981e-06,
+      "loss": 0.1778,
+      "step": 15585
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2480394070848813,
+      "learning_rate": 2.172964460070699e-06,
+      "loss": 0.1795,
+      "step": 15586
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0284664104629266,
+      "learning_rate": 2.1719395017071966e-06,
+      "loss": 0.1626,
+      "step": 15587
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7285129265356551,
+      "learning_rate": 2.170914755680277e-06,
+      "loss": 0.1747,
+      "step": 15588
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0249807644458409,
+      "learning_rate": 2.1698902220177365e-06,
+      "loss": 0.16,
+      "step": 15589
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4069729783476428,
+      "learning_rate": 2.16886590074737e-06,
+      "loss": 0.1767,
+      "step": 15590
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1186356411264524,
+      "learning_rate": 2.167841791896954e-06,
+      "loss": 0.1698,
+      "step": 15591
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4428332295754163,
+      "learning_rate": 2.1668178954942754e-06,
+      "loss": 0.1612,
+      "step": 15592
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8692424415989399,
+      "learning_rate": 2.1657942115671037e-06,
+      "loss": 0.1542,
+      "step": 15593
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.892577770113712,
+      "learning_rate": 2.164770740143203e-06,
+      "loss": 0.1696,
+      "step": 15594
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1374430428559237,
+      "learning_rate": 2.163747481250342e-06,
+      "loss": 0.1705,
+      "step": 15595
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.096025517966074,
+      "learning_rate": 2.1627244349162702e-06,
+      "loss": 0.1597,
+      "step": 15596
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4958794914876075,
+      "learning_rate": 2.161701601168741e-06,
+      "loss": 0.1549,
+      "step": 15597
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4401233627884298,
+      "learning_rate": 2.1606789800354978e-06,
+      "loss": 0.156,
+      "step": 15598
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.243981167898369,
+      "learning_rate": 2.1596565715442843e-06,
+      "loss": 0.1524,
+      "step": 15599
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.12636656464376,
+      "learning_rate": 2.1586343757228247e-06,
+      "loss": 0.1605,
+      "step": 15600
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5292957878929483,
+      "learning_rate": 2.1576123925988548e-06,
+      "loss": 0.1592,
+      "step": 15601
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8964219413195446,
+      "learning_rate": 2.1565906222000877e-06,
+      "loss": 0.1647,
+      "step": 15602
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2008320785613633,
+      "learning_rate": 2.155569064554246e-06,
+      "loss": 0.1705,
+      "step": 15603
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5424164419925088,
+      "learning_rate": 2.154547719689034e-06,
+      "loss": 0.1898,
+      "step": 15604
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9392973979831717,
+      "learning_rate": 2.1535265876321574e-06,
+      "loss": 0.1622,
+      "step": 15605
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9378226880420547,
+      "learning_rate": 2.1525056684113166e-06,
+      "loss": 0.164,
+      "step": 15606
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1582115722775614,
+      "learning_rate": 2.1514849620542045e-06,
+      "loss": 0.1742,
+      "step": 15607
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9833527156186123,
+      "learning_rate": 2.1504644685885044e-06,
+      "loss": 0.1541,
+      "step": 15608
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1202416352338174,
+      "learning_rate": 2.1494441880419005e-06,
+      "loss": 0.1652,
+      "step": 15609
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2100725249448117,
+      "learning_rate": 2.1484241204420676e-06,
+      "loss": 0.1691,
+      "step": 15610
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2245008690966699,
+      "learning_rate": 2.1474042658166703e-06,
+      "loss": 0.1777,
+      "step": 15611
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8136968098830883,
+      "learning_rate": 2.1463846241933785e-06,
+      "loss": 0.1627,
+      "step": 15612
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0309363959270292,
+      "learning_rate": 2.1453651955998445e-06,
+      "loss": 0.1667,
+      "step": 15613
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2038948066066886,
+      "learning_rate": 2.1443459800637234e-06,
+      "loss": 0.1653,
+      "step": 15614
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8752988126685001,
+      "learning_rate": 2.143326977612662e-06,
+      "loss": 0.1613,
+      "step": 15615
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.81493830010666,
+      "learning_rate": 2.1423081882743026e-06,
+      "loss": 0.1677,
+      "step": 15616
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8819559453162542,
+      "learning_rate": 2.1412896120762738e-06,
+      "loss": 0.1824,
+      "step": 15617
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8360825523913156,
+      "learning_rate": 2.1402712490462106e-06,
+      "loss": 0.1689,
+      "step": 15618
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8747309523646191,
+      "learning_rate": 2.139253099211732e-06,
+      "loss": 0.1606,
+      "step": 15619
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0068570576272733,
+      "learning_rate": 2.1382351626004595e-06,
+      "loss": 0.1643,
+      "step": 15620
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0247923129436445,
+      "learning_rate": 2.1372174392400003e-06,
+      "loss": 0.1668,
+      "step": 15621
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0541891741832206,
+      "learning_rate": 2.1361999291579636e-06,
+      "loss": 0.167,
+      "step": 15622
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.006406001594581,
+      "learning_rate": 2.135182632381946e-06,
+      "loss": 0.1552,
+      "step": 15623
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0631465203649262,
+      "learning_rate": 2.134165548939543e-06,
+      "loss": 0.1555,
+      "step": 15624
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2124724084787541,
+      "learning_rate": 2.1331486788583444e-06,
+      "loss": 0.1926,
+      "step": 15625
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1730304656747461,
+      "learning_rate": 2.132132022165935e-06,
+      "loss": 0.1615,
+      "step": 15626
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9980917211125898,
+      "learning_rate": 2.1311155788898884e-06,
+      "loss": 0.1673,
+      "step": 15627
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3509935622005758,
+      "learning_rate": 2.130099349057774e-06,
+      "loss": 0.1931,
+      "step": 15628
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2673298856578072,
+      "learning_rate": 2.1290833326971617e-06,
+      "loss": 0.1688,
+      "step": 15629
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.95908081207994,
+      "learning_rate": 2.128067529835606e-06,
+      "loss": 0.155,
+      "step": 15630
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0957765186131665,
+      "learning_rate": 2.1270519405006618e-06,
+      "loss": 0.1418,
+      "step": 15631
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.9286845139617024,
+      "learning_rate": 2.12603656471988e-06,
+      "loss": 0.154,
+      "step": 15632
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8510502467796793,
+      "learning_rate": 2.1250214025208028e-06,
+      "loss": 0.1514,
+      "step": 15633
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1681591596259013,
+      "learning_rate": 2.1240064539309637e-06,
+      "loss": 0.1829,
+      "step": 15634
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.076521627852123,
+      "learning_rate": 2.122991718977896e-06,
+      "loss": 0.1688,
+      "step": 15635
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7132103153256815,
+      "learning_rate": 2.121977197689119e-06,
+      "loss": 0.1528,
+      "step": 15636
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9492644200091807,
+      "learning_rate": 2.1209628900921597e-06,
+      "loss": 0.1833,
+      "step": 15637
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8755613060414379,
+      "learning_rate": 2.1199487962145236e-06,
+      "loss": 0.1654,
+      "step": 15638
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0228870344740009,
+      "learning_rate": 2.1189349160837247e-06,
+      "loss": 0.17,
+      "step": 15639
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9319147718614745,
+      "learning_rate": 2.1179212497272582e-06,
+      "loss": 0.1625,
+      "step": 15640
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2706661375506798,
+      "learning_rate": 2.116907797172624e-06,
+      "loss": 0.1656,
+      "step": 15641
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0564370189070695,
+      "learning_rate": 2.11589455844731e-06,
+      "loss": 0.1666,
+      "step": 15642
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8366015802686906,
+      "learning_rate": 2.1148815335788044e-06,
+      "loss": 0.1643,
+      "step": 15643
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4744670244805906,
+      "learning_rate": 2.113868722594582e-06,
+      "loss": 0.1693,
+      "step": 15644
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3168260345926481,
+      "learning_rate": 2.1128561255221138e-06,
+      "loss": 0.1713,
+      "step": 15645
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9985770754055425,
+      "learning_rate": 2.1118437423888715e-06,
+      "loss": 0.1794,
+      "step": 15646
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.281413149778363,
+      "learning_rate": 2.1108315732223104e-06,
+      "loss": 0.1713,
+      "step": 15647
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9125070660623015,
+      "learning_rate": 2.109819618049891e-06,
+      "loss": 0.1653,
+      "step": 15648
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8057245462136853,
+      "learning_rate": 2.108807876899057e-06,
+      "loss": 0.1752,
+      "step": 15649
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8487713219507667,
+      "learning_rate": 2.1077963497972555e-06,
+      "loss": 0.1809,
+      "step": 15650
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9196659372204345,
+      "learning_rate": 2.1067850367719247e-06,
+      "loss": 0.1755,
+      "step": 15651
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.99716980725985,
+      "learning_rate": 2.105773937850497e-06,
+      "loss": 0.175,
+      "step": 15652
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8829025384108097,
+      "learning_rate": 2.1047630530603946e-06,
+      "loss": 0.1684,
+      "step": 15653
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.110652541716581,
+      "learning_rate": 2.103752382429043e-06,
+      "loss": 0.1611,
+      "step": 15654
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9720341752038881,
+      "learning_rate": 2.102741925983851e-06,
+      "loss": 0.1536,
+      "step": 15655
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9920772371908424,
+      "learning_rate": 2.101731683752234e-06,
+      "loss": 0.1868,
+      "step": 15656
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7231021520074439,
+      "learning_rate": 2.100721655761587e-06,
+      "loss": 0.182,
+      "step": 15657
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.021432371471618,
+      "learning_rate": 2.0997118420393125e-06,
+      "loss": 0.171,
+      "step": 15658
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9649623834705489,
+      "learning_rate": 2.0987022426128e-06,
+      "loss": 0.1707,
+      "step": 15659
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.258458441032225,
+      "learning_rate": 2.097692857509439e-06,
+      "loss": 0.1984,
+      "step": 15660
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9717680207982854,
+      "learning_rate": 2.096683686756602e-06,
+      "loss": 0.1566,
+      "step": 15661
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7280588099862542,
+      "learning_rate": 2.0956747303816694e-06,
+      "loss": 0.1463,
+      "step": 15662
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9012205407601845,
+      "learning_rate": 2.0946659884120056e-06,
+      "loss": 0.1622,
+      "step": 15663
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.584582682612466,
+      "learning_rate": 2.0936574608749717e-06,
+      "loss": 0.164,
+      "step": 15664
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8224010780566432,
+      "learning_rate": 2.0926491477979272e-06,
+      "loss": 0.1481,
+      "step": 15665
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2102346578962973,
+      "learning_rate": 2.0916410492082195e-06,
+      "loss": 0.1545,
+      "step": 15666
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4038159209736356,
+      "learning_rate": 2.0906331651331945e-06,
+      "loss": 0.1589,
+      "step": 15667
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8479165441994736,
+      "learning_rate": 2.0896254956001916e-06,
+      "loss": 0.1943,
+      "step": 15668
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9723482193483798,
+      "learning_rate": 2.0886180406365465e-06,
+      "loss": 0.2076,
+      "step": 15669
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3228687636011576,
+      "learning_rate": 2.0876108002695817e-06,
+      "loss": 0.1931,
+      "step": 15670
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0600971579261,
+      "learning_rate": 2.0866037745266232e-06,
+      "loss": 0.1512,
+      "step": 15671
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2101345515538404,
+      "learning_rate": 2.0855969634349816e-06,
+      "loss": 0.1677,
+      "step": 15672
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2727403325036004,
+      "learning_rate": 2.0845903670219714e-06,
+      "loss": 0.1886,
+      "step": 15673
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1157062298942677,
+      "learning_rate": 2.083583985314893e-06,
+      "loss": 0.1569,
+      "step": 15674
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1086577491811576,
+      "learning_rate": 2.0825778183410485e-06,
+      "loss": 0.1888,
+      "step": 15675
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.043001791637282,
+      "learning_rate": 2.0815718661277253e-06,
+      "loss": 0.1628,
+      "step": 15676
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8982340742383359,
+      "learning_rate": 2.0805661287022115e-06,
+      "loss": 0.1587,
+      "step": 15677
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7720902538988256,
+      "learning_rate": 2.0795606060917896e-06,
+      "loss": 0.1491,
+      "step": 15678
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4247042004798163,
+      "learning_rate": 2.0785552983237366e-06,
+      "loss": 0.1671,
+      "step": 15679
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8891588378758972,
+      "learning_rate": 2.077550205425317e-06,
+      "loss": 0.1702,
+      "step": 15680
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2919121709259531,
+      "learning_rate": 2.0765453274237944e-06,
+      "loss": 0.1652,
+      "step": 15681
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.960395602437291,
+      "learning_rate": 2.0755406643464293e-06,
+      "loss": 0.1513,
+      "step": 15682
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.094433177420648,
+      "learning_rate": 2.07453621622047e-06,
+      "loss": 0.1857,
+      "step": 15683
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.192579997836186,
+      "learning_rate": 2.0735319830731614e-06,
+      "loss": 0.1708,
+      "step": 15684
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2940041677953567,
+      "learning_rate": 2.0725279649317463e-06,
+      "loss": 0.1768,
+      "step": 15685
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0753642049005883,
+      "learning_rate": 2.0715241618234603e-06,
+      "loss": 0.1733,
+      "step": 15686
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8004753323947372,
+      "learning_rate": 2.0705205737755276e-06,
+      "loss": 0.1839,
+      "step": 15687
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0962965286334978,
+      "learning_rate": 2.069517200815173e-06,
+      "loss": 0.1646,
+      "step": 15688
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1776350758720255,
+      "learning_rate": 2.0685140429696114e-06,
+      "loss": 0.1776,
+      "step": 15689
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1486790236667523,
+      "learning_rate": 2.0675111002660566e-06,
+      "loss": 0.1659,
+      "step": 15690
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0319557213802002,
+      "learning_rate": 2.066508372731708e-06,
+      "loss": 0.1483,
+      "step": 15691
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9726333520511419,
+      "learning_rate": 2.0655058603937704e-06,
+      "loss": 0.1565,
+      "step": 15692
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0562199934217975,
+      "learning_rate": 2.0645035632794317e-06,
+      "loss": 0.1657,
+      "step": 15693
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9933074570339296,
+      "learning_rate": 2.0635014814158826e-06,
+      "loss": 0.1601,
+      "step": 15694
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1359255460270459,
+      "learning_rate": 2.0624996148303043e-06,
+      "loss": 0.19,
+      "step": 15695
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.909456397171364,
+      "learning_rate": 2.0614979635498743e-06,
+      "loss": 0.1661,
+      "step": 15696
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8667047580297196,
+      "learning_rate": 2.060496527601761e-06,
+      "loss": 0.1555,
+      "step": 15697
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7733186418000007,
+      "learning_rate": 2.059495307013125e-06,
+      "loss": 0.1288,
+      "step": 15698
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2790866004491492,
+      "learning_rate": 2.0584943018111304e-06,
+      "loss": 0.1741,
+      "step": 15699
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0679954503666529,
+      "learning_rate": 2.0574935120229224e-06,
+      "loss": 0.1689,
+      "step": 15700
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9779793058532171,
+      "learning_rate": 2.0564929376756526e-06,
+      "loss": 0.1613,
+      "step": 15701
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9095580126414525,
+      "learning_rate": 2.0554925787964596e-06,
+      "loss": 0.1584,
+      "step": 15702
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7880611948799627,
+      "learning_rate": 2.0544924354124828e-06,
+      "loss": 0.1481,
+      "step": 15703
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.94216252208585,
+      "learning_rate": 2.053492507550845e-06,
+      "loss": 0.1826,
+      "step": 15704
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9309447877817286,
+      "learning_rate": 2.052492795238673e-06,
+      "loss": 0.1599,
+      "step": 15705
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2542500450284721,
+      "learning_rate": 2.051493298503081e-06,
+      "loss": 0.179,
+      "step": 15706
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0676784313529195,
+      "learning_rate": 2.050494017371185e-06,
+      "loss": 0.1487,
+      "step": 15707
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2173517816166688,
+      "learning_rate": 2.0494949518700846e-06,
+      "loss": 0.1743,
+      "step": 15708
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1361457952523206,
+      "learning_rate": 2.0484961020268857e-06,
+      "loss": 0.1756,
+      "step": 15709
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8451480774651444,
+      "learning_rate": 2.0474974678686764e-06,
+      "loss": 0.1673,
+      "step": 15710
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9615371002031146,
+      "learning_rate": 2.046499049422548e-06,
+      "loss": 0.1523,
+      "step": 15711
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8893607417171246,
+      "learning_rate": 2.045500846715581e-06,
+      "loss": 0.1546,
+      "step": 15712
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9308432802784401,
+      "learning_rate": 2.0445028597748564e-06,
+      "loss": 0.1748,
+      "step": 15713
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1209141683077861,
+      "learning_rate": 2.04350508862744e-06,
+      "loss": 0.1768,
+      "step": 15714
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9681851429622521,
+      "learning_rate": 2.042507533300395e-06,
+      "loss": 0.1701,
+      "step": 15715
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0106823838571959,
+      "learning_rate": 2.041510193820786e-06,
+      "loss": 0.1589,
+      "step": 15716
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1698606792797457,
+      "learning_rate": 2.0405130702156596e-06,
+      "loss": 0.163,
+      "step": 15717
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3727999721815387,
+      "learning_rate": 2.0395161625120684e-06,
+      "loss": 0.1619,
+      "step": 15718
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2642484204300222,
+      "learning_rate": 2.038519470737048e-06,
+      "loss": 0.1729,
+      "step": 15719
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.390230059267973,
+      "learning_rate": 2.0375229949176367e-06,
+      "loss": 0.1794,
+      "step": 15720
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9163587083719319,
+      "learning_rate": 2.036526735080865e-06,
+      "loss": 0.1682,
+      "step": 15721
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9447935253499224,
+      "learning_rate": 2.0355306912537565e-06,
+      "loss": 0.1703,
+      "step": 15722
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3922281759361557,
+      "learning_rate": 2.0345348634633254e-06,
+      "loss": 0.1785,
+      "step": 15723
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0678235780247454,
+      "learning_rate": 2.0335392517365895e-06,
+      "loss": 0.1396,
+      "step": 15724
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2132485465777194,
+      "learning_rate": 2.032543856100548e-06,
+      "loss": 0.1861,
+      "step": 15725
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0154765236215866,
+      "learning_rate": 2.0315486765822067e-06,
+      "loss": 0.1645,
+      "step": 15726
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.070807295078891,
+      "learning_rate": 2.0305537132085562e-06,
+      "loss": 0.1616,
+      "step": 15727
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8395267663742878,
+      "learning_rate": 2.0295589660065853e-06,
+      "loss": 0.1606,
+      "step": 15728
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0895675484709622,
+      "learning_rate": 2.0285644350032786e-06,
+      "loss": 0.1621,
+      "step": 15729
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.234470650985866,
+      "learning_rate": 2.027570120225614e-06,
+      "loss": 0.1496,
+      "step": 15730
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9022044363184193,
+      "learning_rate": 2.026576021700557e-06,
+      "loss": 0.158,
+      "step": 15731
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0723980217946327,
+      "learning_rate": 2.025582139455078e-06,
+      "loss": 0.1712,
+      "step": 15732
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9710404594344169,
+      "learning_rate": 2.024588473516135e-06,
+      "loss": 0.1716,
+      "step": 15733
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0359442328774089,
+      "learning_rate": 2.023595023910677e-06,
+      "loss": 0.1672,
+      "step": 15734
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0489456256074758,
+      "learning_rate": 2.0226017906656568e-06,
+      "loss": 0.1577,
+      "step": 15735
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0072651874558733,
+      "learning_rate": 2.0216087738080116e-06,
+      "loss": 0.1693,
+      "step": 15736
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1457513882633208,
+      "learning_rate": 2.020615973364679e-06,
+      "loss": 0.1623,
+      "step": 15737
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.297135541941941,
+      "learning_rate": 2.0196233893625883e-06,
+      "loss": 0.1638,
+      "step": 15738
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.971115124295936,
+      "learning_rate": 2.0186310218286674e-06,
+      "loss": 0.1912,
+      "step": 15739
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2845202111648972,
+      "learning_rate": 2.0176388707898274e-06,
+      "loss": 0.1634,
+      "step": 15740
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8963316976105339,
+      "learning_rate": 2.0166469362729868e-06,
+      "loss": 0.147,
+      "step": 15741
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9617179150805765,
+      "learning_rate": 2.015655218305046e-06,
+      "loss": 0.1708,
+      "step": 15742
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0783035229158529,
+      "learning_rate": 2.0146637169129114e-06,
+      "loss": 0.1705,
+      "step": 15743
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.989808277406126,
+      "learning_rate": 2.0136724321234714e-06,
+      "loss": 0.1549,
+      "step": 15744
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1378458311050972,
+      "learning_rate": 2.012681363963621e-06,
+      "loss": 0.1559,
+      "step": 15745
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1096384684415859,
+      "learning_rate": 2.011690512460237e-06,
+      "loss": 0.1661,
+      "step": 15746
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1269780749629759,
+      "learning_rate": 2.010699877640199e-06,
+      "loss": 0.146,
+      "step": 15747
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1432157339065054,
+      "learning_rate": 2.0097094595303766e-06,
+      "loss": 0.1725,
+      "step": 15748
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6841558682447949,
+      "learning_rate": 2.008719258157641e-06,
+      "loss": 0.1405,
+      "step": 15749
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4822552614993239,
+      "learning_rate": 2.0077292735488465e-06,
+      "loss": 0.1702,
+      "step": 15750
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.872921064671971,
+      "learning_rate": 2.0067395057308436e-06,
+      "loss": 0.1593,
+      "step": 15751
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4210836201174981,
+      "learning_rate": 2.005749954730486e-06,
+      "loss": 0.1776,
+      "step": 15752
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0048219241176233,
+      "learning_rate": 2.0047606205746095e-06,
+      "loss": 0.1663,
+      "step": 15753
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1412716222867774,
+      "learning_rate": 2.0037715032900527e-06,
+      "loss": 0.1612,
+      "step": 15754
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0694424370520152,
+      "learning_rate": 2.002782602903647e-06,
+      "loss": 0.1815,
+      "step": 15755
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0157618690035999,
+      "learning_rate": 2.0017939194422153e-06,
+      "loss": 0.1572,
+      "step": 15756
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0320628363686533,
+      "learning_rate": 2.000805452932574e-06,
+      "loss": 0.16,
+      "step": 15757
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9266500851117392,
+      "learning_rate": 1.9998172034015382e-06,
+      "loss": 0.1642,
+      "step": 15758
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2331866837650636,
+      "learning_rate": 1.9988291708759112e-06,
+      "loss": 0.1876,
+      "step": 15759
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.195498850218461,
+      "learning_rate": 1.997841355382497e-06,
+      "loss": 0.1913,
+      "step": 15760
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.292371384434633,
+      "learning_rate": 1.996853756948085e-06,
+      "loss": 0.1433,
+      "step": 15761
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8263446800583641,
+      "learning_rate": 1.99586637559947e-06,
+      "loss": 0.1749,
+      "step": 15762
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9366195331034552,
+      "learning_rate": 1.9948792113634286e-06,
+      "loss": 0.1765,
+      "step": 15763
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9960600752360786,
+      "learning_rate": 1.9938922642667412e-06,
+      "loss": 0.1725,
+      "step": 15764
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.813794636115855,
+      "learning_rate": 1.992905534336178e-06,
+      "loss": 0.1622,
+      "step": 15765
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0309233878437345,
+      "learning_rate": 1.9919190215985073e-06,
+      "loss": 0.1848,
+      "step": 15766
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.122958722281624,
+      "learning_rate": 1.9909327260804857e-06,
+      "loss": 0.1691,
+      "step": 15767
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.999950954499327,
+      "learning_rate": 1.9899466478088624e-06,
+      "loss": 0.172,
+      "step": 15768
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0435904214356926,
+      "learning_rate": 1.988960786810392e-06,
+      "loss": 0.1712,
+      "step": 15769
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0331564521315622,
+      "learning_rate": 1.98797514311181e-06,
+      "loss": 0.1449,
+      "step": 15770
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1759836053908204,
+      "learning_rate": 1.9869897167398576e-06,
+      "loss": 0.1635,
+      "step": 15771
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.959731234772625,
+      "learning_rate": 1.9860045077212596e-06,
+      "loss": 0.1487,
+      "step": 15772
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0413370355997298,
+      "learning_rate": 1.9850195160827413e-06,
+      "loss": 0.1479,
+      "step": 15773
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0646452471085828,
+      "learning_rate": 1.984034741851022e-06,
+      "loss": 0.1884,
+      "step": 15774
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9575156754587291,
+      "learning_rate": 1.983050185052815e-06,
+      "loss": 0.1631,
+      "step": 15775
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0866603608279728,
+      "learning_rate": 1.982065845714821e-06,
+      "loss": 0.1566,
+      "step": 15776
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7649475499030911,
+      "learning_rate": 1.981081723863748e-06,
+      "loss": 0.1656,
+      "step": 15777
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4407044728748781,
+      "learning_rate": 1.9800978195262833e-06,
+      "loss": 0.1635,
+      "step": 15778
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2012243933420368,
+      "learning_rate": 1.9791141327291207e-06,
+      "loss": 0.1667,
+      "step": 15779
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7866904569914929,
+      "learning_rate": 1.9781306634989373e-06,
+      "loss": 0.1476,
+      "step": 15780
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4004951623252793,
+      "learning_rate": 1.9771474118624123e-06,
+      "loss": 0.1752,
+      "step": 15781
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1472073072957027,
+      "learning_rate": 1.976164377846218e-06,
+      "loss": 0.166,
+      "step": 15782
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9439095519046051,
+      "learning_rate": 1.97518156147702e-06,
+      "loss": 0.1813,
+      "step": 15783
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2137528617256548,
+      "learning_rate": 1.974198962781475e-06,
+      "loss": 0.1753,
+      "step": 15784
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.939358251629492,
+      "learning_rate": 1.973216581786235e-06,
+      "loss": 0.1601,
+      "step": 15785
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0757297630588072,
+      "learning_rate": 1.972234418517951e-06,
+      "loss": 0.1591,
+      "step": 15786
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4805360150023117,
+      "learning_rate": 1.971252473003259e-06,
+      "loss": 0.1767,
+      "step": 15787
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0135733620995198,
+      "learning_rate": 1.9702707452687988e-06,
+      "loss": 0.14,
+      "step": 15788
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.087953767363629,
+      "learning_rate": 1.9692892353411963e-06,
+      "loss": 0.1531,
+      "step": 15789
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.67015170683124,
+      "learning_rate": 1.9683079432470774e-06,
+      "loss": 0.1513,
+      "step": 15790
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9810071214994488,
+      "learning_rate": 1.967326869013059e-06,
+      "loss": 0.1476,
+      "step": 15791
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0675283797681412,
+      "learning_rate": 1.966346012665754e-06,
+      "loss": 0.1575,
+      "step": 15792
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.121737893971538,
+      "learning_rate": 1.965365374231766e-06,
+      "loss": 0.1748,
+      "step": 15793
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2965253386253484,
+      "learning_rate": 1.964384953737699e-06,
+      "loss": 0.1592,
+      "step": 15794
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8148539120516977,
+      "learning_rate": 1.9634047512101405e-06,
+      "loss": 0.182,
+      "step": 15795
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1682681957562062,
+      "learning_rate": 1.9624247666756844e-06,
+      "loss": 0.1777,
+      "step": 15796
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8664849282242465,
+      "learning_rate": 1.9614450001609085e-06,
+      "loss": 0.1577,
+      "step": 15797
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.082203974334217,
+      "learning_rate": 1.960465451692394e-06,
+      "loss": 0.1579,
+      "step": 15798
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9962982984962556,
+      "learning_rate": 1.9594861212967055e-06,
+      "loss": 0.1612,
+      "step": 15799
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0910307551966145,
+      "learning_rate": 1.958507009000409e-06,
+      "loss": 0.1689,
+      "step": 15800
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.846953543876322,
+      "learning_rate": 1.957528114830065e-06,
+      "loss": 0.1661,
+      "step": 15801
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0591308349877249,
+      "learning_rate": 1.9565494388122273e-06,
+      "loss": 0.1529,
+      "step": 15802
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0679178149141837,
+      "learning_rate": 1.95557098097344e-06,
+      "loss": 0.1597,
+      "step": 15803
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.315835514316853,
+      "learning_rate": 1.9545927413402423e-06,
+      "loss": 0.1676,
+      "step": 15804
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9562754514141059,
+      "learning_rate": 1.953614719939173e-06,
+      "loss": 0.1579,
+      "step": 15805
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0302795637065507,
+      "learning_rate": 1.9526369167967575e-06,
+      "loss": 0.1626,
+      "step": 15806
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9573846588142638,
+      "learning_rate": 1.951659331939518e-06,
+      "loss": 0.1702,
+      "step": 15807
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5062813400680592,
+      "learning_rate": 1.9506819653939757e-06,
+      "loss": 0.1699,
+      "step": 15808
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.898213287994842,
+      "learning_rate": 1.9497048171866417e-06,
+      "loss": 0.1678,
+      "step": 15809
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8511959910939554,
+      "learning_rate": 1.948727887344016e-06,
+      "loss": 0.155,
+      "step": 15810
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9784540549217013,
+      "learning_rate": 1.947751175892605e-06,
+      "loss": 0.1587,
+      "step": 15811
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9690023808596446,
+      "learning_rate": 1.946774682858894e-06,
+      "loss": 0.1456,
+      "step": 15812
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.155590744200502,
+      "learning_rate": 1.9457984082693783e-06,
+      "loss": 0.1587,
+      "step": 15813
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1814968608821823,
+      "learning_rate": 1.9448223521505328e-06,
+      "loss": 0.1735,
+      "step": 15814
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0409770722575613,
+      "learning_rate": 1.9438465145288377e-06,
+      "loss": 0.1674,
+      "step": 15815
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.612792088595669,
+      "learning_rate": 1.9428708954307595e-06,
+      "loss": 0.1713,
+      "step": 15816
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9555743960260334,
+      "learning_rate": 1.941895494882763e-06,
+      "loss": 0.1508,
+      "step": 15817
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4384926799369615,
+      "learning_rate": 1.940920312911306e-06,
+      "loss": 0.1838,
+      "step": 15818
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0606797339626892,
+      "learning_rate": 1.9399453495428434e-06,
+      "loss": 0.1627,
+      "step": 15819
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8377247178765734,
+      "learning_rate": 1.9389706048038183e-06,
+      "loss": 0.1799,
+      "step": 15820
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.329021242357886,
+      "learning_rate": 1.9379960787206686e-06,
+      "loss": 0.1749,
+      "step": 15821
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2807704941709444,
+      "learning_rate": 1.9370217713198326e-06,
+      "loss": 0.1577,
+      "step": 15822
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2696927151281323,
+      "learning_rate": 1.936047682627734e-06,
+      "loss": 0.1498,
+      "step": 15823
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.34286567311086,
+      "learning_rate": 1.9350738126707978e-06,
+      "loss": 0.1778,
+      "step": 15824
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.102952604803538,
+      "learning_rate": 1.9341001614754407e-06,
+      "loss": 0.1737,
+      "step": 15825
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2905620541568963,
+      "learning_rate": 1.9331267290680744e-06,
+      "loss": 0.1787,
+      "step": 15826
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7752783882287622,
+      "learning_rate": 1.932153515475098e-06,
+      "loss": 0.1553,
+      "step": 15827
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.389049775025545,
+      "learning_rate": 1.9311805207229162e-06,
+      "loss": 0.1576,
+      "step": 15828
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.203796518234086,
+      "learning_rate": 1.9302077448379154e-06,
+      "loss": 0.1906,
+      "step": 15829
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7810701067378404,
+      "learning_rate": 1.929235187846488e-06,
+      "loss": 0.1539,
+      "step": 15830
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1441633726858527,
+      "learning_rate": 1.9282628497750098e-06,
+      "loss": 0.1673,
+      "step": 15831
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3399102124466988,
+      "learning_rate": 1.92729073064986e-06,
+      "loss": 0.194,
+      "step": 15832
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9996400920561656,
+      "learning_rate": 1.9263188304974023e-06,
+      "loss": 0.1648,
+      "step": 15833
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.816567835157871,
+      "learning_rate": 1.9253471493440036e-06,
+      "loss": 0.1708,
+      "step": 15834
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8033366766000336,
+      "learning_rate": 1.924375687216018e-06,
+      "loss": 0.1448,
+      "step": 15835
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0534861733157275,
+      "learning_rate": 1.9234044441398016e-06,
+      "loss": 0.169,
+      "step": 15836
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8614388275474186,
+      "learning_rate": 1.922433420141695e-06,
+      "loss": 0.1566,
+      "step": 15837
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9897933960131479,
+      "learning_rate": 1.921462615248035e-06,
+      "loss": 0.1361,
+      "step": 15838
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0033910146002347,
+      "learning_rate": 1.9204920294851613e-06,
+      "loss": 0.168,
+      "step": 15839
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.3643061597353467,
+      "learning_rate": 1.9195216628793956e-06,
+      "loss": 0.162,
+      "step": 15840
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3976734137083604,
+      "learning_rate": 1.9185515154570633e-06,
+      "loss": 0.18,
+      "step": 15841
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1903452435943236,
+      "learning_rate": 1.9175815872444748e-06,
+      "loss": 0.1673,
+      "step": 15842
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8597857795206585,
+      "learning_rate": 1.916611878267942e-06,
+      "loss": 0.1857,
+      "step": 15843
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2959628492227406,
+      "learning_rate": 1.915642388553769e-06,
+      "loss": 0.1528,
+      "step": 15844
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3997789619905479,
+      "learning_rate": 1.9146731181282554e-06,
+      "loss": 0.1762,
+      "step": 15845
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0091633763978645,
+      "learning_rate": 1.9137040670176878e-06,
+      "loss": 0.1719,
+      "step": 15846
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.299180758820901,
+      "learning_rate": 1.9127352352483563e-06,
+      "loss": 0.1726,
+      "step": 15847
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9054257427731401,
+      "learning_rate": 1.911766622846536e-06,
+      "loss": 0.1776,
+      "step": 15848
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.592958273371159,
+      "learning_rate": 1.9107982298385052e-06,
+      "loss": 0.1783,
+      "step": 15849
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.15568910717821,
+      "learning_rate": 1.9098300562505266e-06,
+      "loss": 0.164,
+      "step": 15850
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2317553042477192,
+      "learning_rate": 1.908862102108865e-06,
+      "loss": 0.1848,
+      "step": 15851
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.065679268129052,
+      "learning_rate": 1.9078943674397753e-06,
+      "loss": 0.1531,
+      "step": 15852
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8677385153275745,
+      "learning_rate": 1.9069268522695105e-06,
+      "loss": 0.1704,
+      "step": 15853
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.015003102554297,
+      "learning_rate": 1.9059595566243127e-06,
+      "loss": 0.1577,
+      "step": 15854
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5993881937232388,
+      "learning_rate": 1.9049924805304165e-06,
+      "loss": 0.1687,
+      "step": 15855
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9648294924172559,
+      "learning_rate": 1.9040256240140587e-06,
+      "loss": 0.1761,
+      "step": 15856
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.129006993288202,
+      "learning_rate": 1.9030589871014604e-06,
+      "loss": 0.1539,
+      "step": 15857
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0888111506199036,
+      "learning_rate": 1.9020925698188465e-06,
+      "loss": 0.1651,
+      "step": 15858
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.05744231234916,
+      "learning_rate": 1.9011263721924277e-06,
+      "loss": 0.1602,
+      "step": 15859
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.8526338413193164,
+      "learning_rate": 1.9001603942484127e-06,
+      "loss": 0.172,
+      "step": 15860
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3523800109758868,
+      "learning_rate": 1.8991946360130043e-06,
+      "loss": 0.1401,
+      "step": 15861
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.515723016981549,
+      "learning_rate": 1.8982290975124019e-06,
+      "loss": 0.1701,
+      "step": 15862
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7267760543292379,
+      "learning_rate": 1.8972637787727898e-06,
+      "loss": 0.1779,
+      "step": 15863
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.773154795440011,
+      "learning_rate": 1.8962986798203587e-06,
+      "loss": 0.1682,
+      "step": 15864
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0148911097830382,
+      "learning_rate": 1.8953338006812805e-06,
+      "loss": 0.1721,
+      "step": 15865
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.122399862038846,
+      "learning_rate": 1.8943691413817334e-06,
+      "loss": 0.1736,
+      "step": 15866
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1388130015125641,
+      "learning_rate": 1.8934047019478785e-06,
+      "loss": 0.149,
+      "step": 15867
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1411525588589864,
+      "learning_rate": 1.8924404824058816e-06,
+      "loss": 0.1763,
+      "step": 15868
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.338725253391659,
+      "learning_rate": 1.8914764827818921e-06,
+      "loss": 0.1514,
+      "step": 15869
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.542120667256205,
+      "learning_rate": 1.8905127031020598e-06,
+      "loss": 0.1648,
+      "step": 15870
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.957055240444867,
+      "learning_rate": 1.8895491433925328e-06,
+      "loss": 0.1638,
+      "step": 15871
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8326227194676635,
+      "learning_rate": 1.8885858036794401e-06,
+      "loss": 0.1607,
+      "step": 15872
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0202307760657778,
+      "learning_rate": 1.8876226839889177e-06,
+      "loss": 0.1709,
+      "step": 15873
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1402323985518499,
+      "learning_rate": 1.8866597843470858e-06,
+      "loss": 0.1642,
+      "step": 15874
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.010208304292357,
+      "learning_rate": 1.8856971047800687e-06,
+      "loss": 0.1636,
+      "step": 15875
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 26.547268409354285,
+      "learning_rate": 1.8847346453139726e-06,
+      "loss": 0.1706,
+      "step": 15876
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9852515941330995,
+      "learning_rate": 1.8837724059749074e-06,
+      "loss": 0.159,
+      "step": 15877
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0533560381335083,
+      "learning_rate": 1.8828103867889747e-06,
+      "loss": 0.1958,
+      "step": 15878
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3430404155874531,
+      "learning_rate": 1.8818485877822712e-06,
+      "loss": 0.1859,
+      "step": 15879
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8794215939911043,
+      "learning_rate": 1.8808870089808806e-06,
+      "loss": 0.1685,
+      "step": 15880
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.037638249002718,
+      "learning_rate": 1.879925650410892e-06,
+      "loss": 0.1607,
+      "step": 15881
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8697582929524619,
+      "learning_rate": 1.8789645120983746e-06,
+      "loss": 0.1792,
+      "step": 15882
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8982466851373708,
+      "learning_rate": 1.8780035940694075e-06,
+      "loss": 0.1577,
+      "step": 15883
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.061375290924336,
+      "learning_rate": 1.8770428963500475e-06,
+      "loss": 0.1484,
+      "step": 15884
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4971804168207266,
+      "learning_rate": 1.8760824189663618e-06,
+      "loss": 0.1775,
+      "step": 15885
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9099341373713296,
+      "learning_rate": 1.875122161944396e-06,
+      "loss": 0.1559,
+      "step": 15886
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0462905697176088,
+      "learning_rate": 1.8741621253102005e-06,
+      "loss": 0.1675,
+      "step": 15887
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7935935303208741,
+      "learning_rate": 1.8732023090898165e-06,
+      "loss": 0.1369,
+      "step": 15888
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9045384382704341,
+      "learning_rate": 1.8722427133092813e-06,
+      "loss": 0.173,
+      "step": 15889
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1845902438296145,
+      "learning_rate": 1.8712833379946217e-06,
+      "loss": 0.1821,
+      "step": 15890
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7466157451415563,
+      "learning_rate": 1.8703241831718578e-06,
+      "loss": 0.1655,
+      "step": 15891
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8891584234780563,
+      "learning_rate": 1.8693652488670121e-06,
+      "loss": 0.1685,
+      "step": 15892
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.973723244657113,
+      "learning_rate": 1.868406535106091e-06,
+      "loss": 0.1593,
+      "step": 15893
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1918050991675817,
+      "learning_rate": 1.8674480419151041e-06,
+      "loss": 0.1669,
+      "step": 15894
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8854809913998016,
+      "learning_rate": 1.8664897693200456e-06,
+      "loss": 0.1583,
+      "step": 15895
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0964839487023017,
+      "learning_rate": 1.8655317173469122e-06,
+      "loss": 0.1854,
+      "step": 15896
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8412304856386966,
+      "learning_rate": 1.86457388602169e-06,
+      "loss": 0.1756,
+      "step": 15897
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5080930598938767,
+      "learning_rate": 1.8636162753703636e-06,
+      "loss": 0.1666,
+      "step": 15898
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.05984641476992,
+      "learning_rate": 1.862658885418902e-06,
+      "loss": 0.1487,
+      "step": 15899
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0047321045727262,
+      "learning_rate": 1.8617017161932815e-06,
+      "loss": 0.1705,
+      "step": 15900
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9856203454509121,
+      "learning_rate": 1.8607447677194578e-06,
+      "loss": 0.1615,
+      "step": 15901
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7206410988649752,
+      "learning_rate": 1.8597880400233959e-06,
+      "loss": 0.1529,
+      "step": 15902
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1406816086087803,
+      "learning_rate": 1.8588315331310392e-06,
+      "loss": 0.1806,
+      "step": 15903
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4328388461061061,
+      "learning_rate": 1.8578752470683381e-06,
+      "loss": 0.1686,
+      "step": 15904
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.15321917515873,
+      "learning_rate": 1.8569191818612298e-06,
+      "loss": 0.1626,
+      "step": 15905
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.175305560323787,
+      "learning_rate": 1.8559633375356511e-06,
+      "loss": 0.1803,
+      "step": 15906
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.026447711263733,
+      "learning_rate": 1.8550077141175282e-06,
+      "loss": 0.1706,
+      "step": 15907
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8598763378612653,
+      "learning_rate": 1.8540523116327769e-06,
+      "loss": 0.1587,
+      "step": 15908
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9602544825284762,
+      "learning_rate": 1.8530971301073208e-06,
+      "loss": 0.1769,
+      "step": 15909
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1276646643628279,
+      "learning_rate": 1.8521421695670617e-06,
+      "loss": 0.1819,
+      "step": 15910
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3784525456178116,
+      "learning_rate": 1.8511874300379095e-06,
+      "loss": 0.1564,
+      "step": 15911
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9188987056733893,
+      "learning_rate": 1.8502329115457551e-06,
+      "loss": 0.185,
+      "step": 15912
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8884175776293851,
+      "learning_rate": 1.8492786141164943e-06,
+      "loss": 0.1616,
+      "step": 15913
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.864036007568502,
+      "learning_rate": 1.8483245377760106e-06,
+      "loss": 0.163,
+      "step": 15914
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.175478431383708,
+      "learning_rate": 1.847370682550187e-06,
+      "loss": 0.1706,
+      "step": 15915
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8877435476808375,
+      "learning_rate": 1.8464170484648924e-06,
+      "loss": 0.155,
+      "step": 15916
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2440300952858039,
+      "learning_rate": 1.8454636355459977e-06,
+      "loss": 0.152,
+      "step": 15917
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0778220695867444,
+      "learning_rate": 1.8445104438193595e-06,
+      "loss": 0.1672,
+      "step": 15918
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.217954161449602,
+      "learning_rate": 1.8435574733108397e-06,
+      "loss": 0.1745,
+      "step": 15919
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.905783275261828,
+      "learning_rate": 1.8426047240462807e-06,
+      "loss": 0.157,
+      "step": 15920
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0235188483867,
+      "learning_rate": 1.84165219605153e-06,
+      "loss": 0.1628,
+      "step": 15921
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7664994271883586,
+      "learning_rate": 1.840699889352423e-06,
+      "loss": 0.165,
+      "step": 15922
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.55969940166033,
+      "learning_rate": 1.8397478039747962e-06,
+      "loss": 0.1829,
+      "step": 15923
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9627591745734583,
+      "learning_rate": 1.83879593994447e-06,
+      "loss": 0.1737,
+      "step": 15924
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1817190829880238,
+      "learning_rate": 1.8378442972872635e-06,
+      "loss": 0.164,
+      "step": 15925
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2039858241344097,
+      "learning_rate": 1.8368928760289928e-06,
+      "loss": 0.1579,
+      "step": 15926
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9183717007030037,
+      "learning_rate": 1.8359416761954629e-06,
+      "loss": 0.1791,
+      "step": 15927
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0886857154910026,
+      "learning_rate": 1.8349906978124776e-06,
+      "loss": 0.1548,
+      "step": 15928
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0074054922942874,
+      "learning_rate": 1.8340399409058284e-06,
+      "loss": 0.1753,
+      "step": 15929
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2258662949395527,
+      "learning_rate": 1.833089405501307e-06,
+      "loss": 0.1464,
+      "step": 15930
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9872561371738571,
+      "learning_rate": 1.8321390916246961e-06,
+      "loss": 0.1626,
+      "step": 15931
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3703066745943606,
+      "learning_rate": 1.8311889993017772e-06,
+      "loss": 0.1834,
+      "step": 15932
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1405886391351094,
+      "learning_rate": 1.8302391285583153e-06,
+      "loss": 0.1906,
+      "step": 15933
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6044683790640693,
+      "learning_rate": 1.829289479420081e-06,
+      "loss": 0.152,
+      "step": 15934
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.029769477626035,
+      "learning_rate": 1.828340051912828e-06,
+      "loss": 0.1515,
+      "step": 15935
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3723365923561242,
+      "learning_rate": 1.8273908460623157e-06,
+      "loss": 0.1474,
+      "step": 15936
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.248137118999691,
+      "learning_rate": 1.8264418618942859e-06,
+      "loss": 0.1438,
+      "step": 15937
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.280726460462142,
+      "learning_rate": 1.8254930994344845e-06,
+      "loss": 0.1709,
+      "step": 15938
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1191582109101215,
+      "learning_rate": 1.8245445587086419e-06,
+      "loss": 0.1723,
+      "step": 15939
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2899830299526782,
+      "learning_rate": 1.82359623974249e-06,
+      "loss": 0.193,
+      "step": 15940
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0323262769938613,
+      "learning_rate": 1.8226481425617549e-06,
+      "loss": 0.1539,
+      "step": 15941
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5728960046365268,
+      "learning_rate": 1.821700267192148e-06,
+      "loss": 0.1543,
+      "step": 15942
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.888435356901117,
+      "learning_rate": 1.8207526136593857e-06,
+      "loss": 0.1849,
+      "step": 15943
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2068742130827914,
+      "learning_rate": 1.8198051819891672e-06,
+      "loss": 0.1574,
+      "step": 15944
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8943393072237987,
+      "learning_rate": 1.8188579722071985e-06,
+      "loss": 0.1575,
+      "step": 15945
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.218001554855009,
+      "learning_rate": 1.8179109843391673e-06,
+      "loss": 0.1686,
+      "step": 15946
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0580225505444398,
+      "learning_rate": 1.8169642184107628e-06,
+      "loss": 0.1771,
+      "step": 15947
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9884038518855041,
+      "learning_rate": 1.8160176744476654e-06,
+      "loss": 0.1588,
+      "step": 15948
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0281074804685377,
+      "learning_rate": 1.8150713524755536e-06,
+      "loss": 0.163,
+      "step": 15949
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8649855464767741,
+      "learning_rate": 1.8141252525200914e-06,
+      "loss": 0.1706,
+      "step": 15950
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5950583604159652,
+      "learning_rate": 1.813179374606946e-06,
+      "loss": 0.1543,
+      "step": 15951
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2086613904361754,
+      "learning_rate": 1.81223371876177e-06,
+      "loss": 0.1489,
+      "step": 15952
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7095853537929204,
+      "learning_rate": 1.8112882850102198e-06,
+      "loss": 0.1523,
+      "step": 15953
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9331269166408543,
+      "learning_rate": 1.8103430733779348e-06,
+      "loss": 0.1722,
+      "step": 15954
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8453545466230363,
+      "learning_rate": 1.809398083890559e-06,
+      "loss": 0.1549,
+      "step": 15955
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7681451063904245,
+      "learning_rate": 1.8084533165737195e-06,
+      "loss": 0.1457,
+      "step": 15956
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9178846288166791,
+      "learning_rate": 1.807508771453047e-06,
+      "loss": 0.1501,
+      "step": 15957
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0385827144853221,
+      "learning_rate": 1.8065644485541622e-06,
+      "loss": 0.1582,
+      "step": 15958
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8697295581069294,
+      "learning_rate": 1.8056203479026812e-06,
+      "loss": 0.1473,
+      "step": 15959
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2051188420799237,
+      "learning_rate": 1.8046764695242115e-06,
+      "loss": 0.1587,
+      "step": 15960
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0327908075621928,
+      "learning_rate": 1.8037328134443532e-06,
+      "loss": 0.175,
+      "step": 15961
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9823285672773258,
+      "learning_rate": 1.8027893796887075e-06,
+      "loss": 0.1754,
+      "step": 15962
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7991140537880537,
+      "learning_rate": 1.8018461682828603e-06,
+      "loss": 0.1396,
+      "step": 15963
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.347432924466073,
+      "learning_rate": 1.8009031792524012e-06,
+      "loss": 0.1891,
+      "step": 15964
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6531415450426012,
+      "learning_rate": 1.7999604126229043e-06,
+      "loss": 0.1518,
+      "step": 15965
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0830412428308247,
+      "learning_rate": 1.7990178684199444e-06,
+      "loss": 0.1662,
+      "step": 15966
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9304754757162109,
+      "learning_rate": 1.798075546669088e-06,
+      "loss": 0.1515,
+      "step": 15967
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7648335000274984,
+      "learning_rate": 1.797133447395898e-06,
+      "loss": 0.1643,
+      "step": 15968
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5204642842384637,
+      "learning_rate": 1.7961915706259236e-06,
+      "loss": 0.1595,
+      "step": 15969
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.230914612772511,
+      "learning_rate": 1.7952499163847192e-06,
+      "loss": 0.1545,
+      "step": 15970
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.048495018382882,
+      "learning_rate": 1.794308484697822e-06,
+      "loss": 0.1423,
+      "step": 15971
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.19952605245712,
+      "learning_rate": 1.7933672755907727e-06,
+      "loss": 0.1515,
+      "step": 15972
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2280860890863525,
+      "learning_rate": 1.7924262890890964e-06,
+      "loss": 0.1407,
+      "step": 15973
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0566678871977504,
+      "learning_rate": 1.7914855252183217e-06,
+      "loss": 0.193,
+      "step": 15974
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2322299380808484,
+      "learning_rate": 1.7905449840039645e-06,
+      "loss": 0.1444,
+      "step": 15975
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1100249588562576,
+      "learning_rate": 1.7896046654715427e-06,
+      "loss": 0.1737,
+      "step": 15976
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1424919104979556,
+      "learning_rate": 1.7886645696465566e-06,
+      "loss": 0.1688,
+      "step": 15977
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9677874819880176,
+      "learning_rate": 1.787724696554506e-06,
+      "loss": 0.1354,
+      "step": 15978
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.060213074988549,
+      "learning_rate": 1.7867850462208892e-06,
+      "loss": 0.1783,
+      "step": 15979
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8748895283896404,
+      "learning_rate": 1.785845618671188e-06,
+      "loss": 0.1462,
+      "step": 15980
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.25618502996236,
+      "learning_rate": 1.7849064139308925e-06,
+      "loss": 0.1594,
+      "step": 15981
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.127720954067132,
+      "learning_rate": 1.7839674320254718e-06,
+      "loss": 0.1536,
+      "step": 15982
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0095621148613478,
+      "learning_rate": 1.783028672980398e-06,
+      "loss": 0.1686,
+      "step": 15983
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0674805735190684,
+      "learning_rate": 1.7820901368211362e-06,
+      "loss": 0.1575,
+      "step": 15984
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9566168211011928,
+      "learning_rate": 1.7811518235731461e-06,
+      "loss": 0.1596,
+      "step": 15985
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3033046485908206,
+      "learning_rate": 1.780213733261874e-06,
+      "loss": 0.1595,
+      "step": 15986
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9332843058050407,
+      "learning_rate": 1.7792758659127706e-06,
+      "loss": 0.172,
+      "step": 15987
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0017301526569287,
+      "learning_rate": 1.7783382215512724e-06,
+      "loss": 0.1769,
+      "step": 15988
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2771970862900306,
+      "learning_rate": 1.7774008002028164e-06,
+      "loss": 0.1475,
+      "step": 15989
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9754875288852934,
+      "learning_rate": 1.7764636018928249e-06,
+      "loss": 0.1619,
+      "step": 15990
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4876323017582416,
+      "learning_rate": 1.7755266266467264e-06,
+      "loss": 0.1752,
+      "step": 15991
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1379979426252034,
+      "learning_rate": 1.7745898744899292e-06,
+      "loss": 0.162,
+      "step": 15992
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8945533397316004,
+      "learning_rate": 1.7736533454478466e-06,
+      "loss": 0.166,
+      "step": 15993
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.474577884249597,
+      "learning_rate": 1.7727170395458838e-06,
+      "loss": 0.1563,
+      "step": 15994
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1805580502190653,
+      "learning_rate": 1.7717809568094334e-06,
+      "loss": 0.1552,
+      "step": 15995
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8905053860600621,
+      "learning_rate": 1.7708450972638923e-06,
+      "loss": 0.1592,
+      "step": 15996
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1618938816030313,
+      "learning_rate": 1.76990946093464e-06,
+      "loss": 0.1844,
+      "step": 15997
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2644217815426884,
+      "learning_rate": 1.7689740478470608e-06,
+      "loss": 0.1661,
+      "step": 15998
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9554434364928839,
+      "learning_rate": 1.768038858026523e-06,
+      "loss": 0.163,
+      "step": 15999
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1010955555027215,
+      "learning_rate": 1.7671038914983963e-06,
+      "loss": 0.1675,
+      "step": 16000
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8927567601335374,
+      "learning_rate": 1.7661691482880416e-06,
+      "loss": 0.1633,
+      "step": 16001
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.202918742731831,
+      "learning_rate": 1.7652346284208167e-06,
+      "loss": 0.1921,
+      "step": 16002
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2242544638925155,
+      "learning_rate": 1.7643003319220642e-06,
+      "loss": 0.1741,
+      "step": 16003
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8326302787576414,
+      "learning_rate": 1.763366258817133e-06,
+      "loss": 0.1455,
+      "step": 16004
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3261346371041038,
+      "learning_rate": 1.762432409131355e-06,
+      "loss": 0.1838,
+      "step": 16005
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3388318854708003,
+      "learning_rate": 1.7614987828900654e-06,
+      "loss": 0.1504,
+      "step": 16006
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1517921068209622,
+      "learning_rate": 1.760565380118584e-06,
+      "loss": 0.1813,
+      "step": 16007
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9223647551231623,
+      "learning_rate": 1.7596322008422351e-06,
+      "loss": 0.1501,
+      "step": 16008
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0249862314487033,
+      "learning_rate": 1.7586992450863261e-06,
+      "loss": 0.1602,
+      "step": 16009
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3558351653151626,
+      "learning_rate": 1.7577665128761645e-06,
+      "loss": 0.1563,
+      "step": 16010
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0748815541391956,
+      "learning_rate": 1.7568340042370546e-06,
+      "loss": 0.1664,
+      "step": 16011
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0504259349670237,
+      "learning_rate": 1.755901719194285e-06,
+      "loss": 0.1619,
+      "step": 16012
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0688802098767785,
+      "learning_rate": 1.7549696577731502e-06,
+      "loss": 0.1584,
+      "step": 16013
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.414884983836496,
+      "learning_rate": 1.754037819998926e-06,
+      "loss": 0.1652,
+      "step": 16014
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3920470329979084,
+      "learning_rate": 1.753106205896895e-06,
+      "loss": 0.1533,
+      "step": 16015
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.166883450478275,
+      "learning_rate": 1.75217481549232e-06,
+      "loss": 0.1838,
+      "step": 16016
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0275778044647248,
+      "learning_rate": 1.7512436488104723e-06,
+      "loss": 0.1607,
+      "step": 16017
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5038447361122222,
+      "learning_rate": 1.7503127058766046e-06,
+      "loss": 0.1896,
+      "step": 16018
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9306661523910239,
+      "learning_rate": 1.74938198671597e-06,
+      "loss": 0.1754,
+      "step": 16019
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4884420143779518,
+      "learning_rate": 1.7484514913538154e-06,
+      "loss": 0.1673,
+      "step": 16020
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9758314584022224,
+      "learning_rate": 1.7475212198153823e-06,
+      "loss": 0.1535,
+      "step": 16021
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2001369707687648,
+      "learning_rate": 1.7465911721259e-06,
+      "loss": 0.178,
+      "step": 16022
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2680078302782236,
+      "learning_rate": 1.7456613483106e-06,
+      "loss": 0.1624,
+      "step": 16023
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9690828344942196,
+      "learning_rate": 1.7447317483947002e-06,
+      "loss": 0.1319,
+      "step": 16024
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2521543439308118,
+      "learning_rate": 1.7438023724034215e-06,
+      "loss": 0.1739,
+      "step": 16025
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8606284544479783,
+      "learning_rate": 1.7428732203619659e-06,
+      "loss": 0.1684,
+      "step": 16026
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2639439074812617,
+      "learning_rate": 1.741944292295541e-06,
+      "loss": 0.2058,
+      "step": 16027
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9448353201257036,
+      "learning_rate": 1.7410155882293434e-06,
+      "loss": 0.148,
+      "step": 16028
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6066023830778167,
+      "learning_rate": 1.7400871081885672e-06,
+      "loss": 0.1708,
+      "step": 16029
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.043947156009396,
+      "learning_rate": 1.7391588521983948e-06,
+      "loss": 0.162,
+      "step": 16030
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0654314480734575,
+      "learning_rate": 1.7382308202840027e-06,
+      "loss": 0.1717,
+      "step": 16031
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9800410320580712,
+      "learning_rate": 1.737303012470568e-06,
+      "loss": 0.1872,
+      "step": 16032
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0186507793410637,
+      "learning_rate": 1.7363754287832535e-06,
+      "loss": 0.174,
+      "step": 16033
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2723809059140396,
+      "learning_rate": 1.7354480692472253e-06,
+      "loss": 0.1666,
+      "step": 16034
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9192403831092457,
+      "learning_rate": 1.7345209338876324e-06,
+      "loss": 0.1606,
+      "step": 16035
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8880642758230413,
+      "learning_rate": 1.7335940227296254e-06,
+      "loss": 0.1649,
+      "step": 16036
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.191038489713937,
+      "learning_rate": 1.7326673357983482e-06,
+      "loss": 0.1748,
+      "step": 16037
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0796781843628873,
+      "learning_rate": 1.7317408731189378e-06,
+      "loss": 0.1598,
+      "step": 16038
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8984590128115462,
+      "learning_rate": 1.7308146347165212e-06,
+      "loss": 0.1815,
+      "step": 16039
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9255649584515351,
+      "learning_rate": 1.729888620616228e-06,
+      "loss": 0.1813,
+      "step": 16040
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1961173565730827,
+      "learning_rate": 1.7289628308431694e-06,
+      "loss": 0.1753,
+      "step": 16041
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.193927705094369,
+      "learning_rate": 1.7280372654224642e-06,
+      "loss": 0.1873,
+      "step": 16042
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8192238004138188,
+      "learning_rate": 1.7271119243792135e-06,
+      "loss": 0.1591,
+      "step": 16043
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.964331006140522,
+      "learning_rate": 1.726186807738518e-06,
+      "loss": 0.1526,
+      "step": 16044
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8146540744504316,
+      "learning_rate": 1.7252619155254734e-06,
+      "loss": 0.1507,
+      "step": 16045
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8120205092019961,
+      "learning_rate": 1.7243372477651688e-06,
+      "loss": 0.1432,
+      "step": 16046
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9761030170080665,
+      "learning_rate": 1.7234128044826836e-06,
+      "loss": 0.1664,
+      "step": 16047
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1021142028744921,
+      "learning_rate": 1.7224885857030916e-06,
+      "loss": 0.1598,
+      "step": 16048
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8608872145849567,
+      "learning_rate": 1.7215645914514668e-06,
+      "loss": 0.1622,
+      "step": 16049
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.107725886336896,
+      "learning_rate": 1.7206408217528669e-06,
+      "loss": 0.1642,
+      "step": 16050
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0829423681796864,
+      "learning_rate": 1.7197172766323556e-06,
+      "loss": 0.1549,
+      "step": 16051
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2654030905626539,
+      "learning_rate": 1.718793956114978e-06,
+      "loss": 0.1551,
+      "step": 16052
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7045139028307472,
+      "learning_rate": 1.717870860225782e-06,
+      "loss": 0.1527,
+      "step": 16053
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.109320034642993,
+      "learning_rate": 1.7169479889898065e-06,
+      "loss": 0.175,
+      "step": 16054
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.471551674901941,
+      "learning_rate": 1.7160253424320872e-06,
+      "loss": 0.1926,
+      "step": 16055
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8560448881427307,
+      "learning_rate": 1.7151029205776459e-06,
+      "loss": 0.181,
+      "step": 16056
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2365487452671833,
+      "learning_rate": 1.7141807234515085e-06,
+      "loss": 0.1671,
+      "step": 16057
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8256655601366389,
+      "learning_rate": 1.7132587510786846e-06,
+      "loss": 0.1469,
+      "step": 16058
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1887788220025923,
+      "learning_rate": 1.7123370034841869e-06,
+      "loss": 0.1688,
+      "step": 16059
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3846537570490054,
+      "learning_rate": 1.7114154806930138e-06,
+      "loss": 0.1707,
+      "step": 16060
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0174836563541019,
+      "learning_rate": 1.7104941827301668e-06,
+      "loss": 0.1723,
+      "step": 16061
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.233727975926785,
+      "learning_rate": 1.7095731096206313e-06,
+      "loss": 0.182,
+      "step": 16062
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9338202871197535,
+      "learning_rate": 1.7086522613893918e-06,
+      "loss": 0.1656,
+      "step": 16063
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5330592853721912,
+      "learning_rate": 1.7077316380614317e-06,
+      "loss": 0.1568,
+      "step": 16064
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.663600549880783,
+      "learning_rate": 1.7068112396617164e-06,
+      "loss": 0.1489,
+      "step": 16065
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2309622613220172,
+      "learning_rate": 1.7058910662152173e-06,
+      "loss": 0.1612,
+      "step": 16066
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2666130678712084,
+      "learning_rate": 1.7049711177468896e-06,
+      "loss": 0.1541,
+      "step": 16067
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9627847158187706,
+      "learning_rate": 1.7040513942816905e-06,
+      "loss": 0.1571,
+      "step": 16068
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.041748171121705,
+      "learning_rate": 1.703131895844563e-06,
+      "loss": 0.1844,
+      "step": 16069
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0538437416835476,
+      "learning_rate": 1.7022126224604529e-06,
+      "loss": 0.1931,
+      "step": 16070
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8196593035973092,
+      "learning_rate": 1.7012935741542925e-06,
+      "loss": 0.1474,
+      "step": 16071
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5544017080113712,
+      "learning_rate": 1.700374750951016e-06,
+      "loss": 0.1739,
+      "step": 16072
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6594564457538314,
+      "learning_rate": 1.6994561528755404e-06,
+      "loss": 0.1629,
+      "step": 16073
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.716203363591545,
+      "learning_rate": 1.698537779952788e-06,
+      "loss": 0.1574,
+      "step": 16074
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4843481543008163,
+      "learning_rate": 1.6976196322076655e-06,
+      "loss": 0.1672,
+      "step": 16075
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2056535835474138,
+      "learning_rate": 1.6967017096650807e-06,
+      "loss": 0.1832,
+      "step": 16076
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9487721823636202,
+      "learning_rate": 1.6957840123499292e-06,
+      "loss": 0.1626,
+      "step": 16077
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8886952662815739,
+      "learning_rate": 1.6948665402871067e-06,
+      "loss": 0.1518,
+      "step": 16078
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4301649119446862,
+      "learning_rate": 1.6939492935014966e-06,
+      "loss": 0.1574,
+      "step": 16079
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0618141135360368,
+      "learning_rate": 1.6930322720179816e-06,
+      "loss": 0.1476,
+      "step": 16080
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.076064486726175,
+      "learning_rate": 1.6921154758614378e-06,
+      "loss": 0.164,
+      "step": 16081
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3984045430593135,
+      "learning_rate": 1.6911989050567279e-06,
+      "loss": 0.1682,
+      "step": 16082
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9973740840239897,
+      "learning_rate": 1.6902825596287198e-06,
+      "loss": 0.1669,
+      "step": 16083
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.190269199690928,
+      "learning_rate": 1.6893664396022646e-06,
+      "loss": 0.1423,
+      "step": 16084
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.473604726310278,
+      "learning_rate": 1.688450545002216e-06,
+      "loss": 0.1795,
+      "step": 16085
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0436960291378419,
+      "learning_rate": 1.6875348758534127e-06,
+      "loss": 0.1685,
+      "step": 16086
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3857159745313614,
+      "learning_rate": 1.6866194321806984e-06,
+      "loss": 0.1651,
+      "step": 16087
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8033380877779788,
+      "learning_rate": 1.6857042140088996e-06,
+      "loss": 0.1655,
+      "step": 16088
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9083806621123651,
+      "learning_rate": 1.6847892213628436e-06,
+      "loss": 0.1639,
+      "step": 16089
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0781586943612742,
+      "learning_rate": 1.6838744542673492e-06,
+      "loss": 0.1603,
+      "step": 16090
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9771201956290885,
+      "learning_rate": 1.6829599127472318e-06,
+      "loss": 0.1743,
+      "step": 16091
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.060528233189165,
+      "learning_rate": 1.6820455968272953e-06,
+      "loss": 0.1684,
+      "step": 16092
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8894300227873709,
+      "learning_rate": 1.681131506532343e-06,
+      "loss": 0.1536,
+      "step": 16093
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3933629571467805,
+      "learning_rate": 1.6802176418871664e-06,
+      "loss": 0.1716,
+      "step": 16094
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9816041566511617,
+      "learning_rate": 1.6793040029165596e-06,
+      "loss": 0.1823,
+      "step": 16095
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2379254151826307,
+      "learning_rate": 1.6783905896452978e-06,
+      "loss": 0.1749,
+      "step": 16096
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0753028944850174,
+      "learning_rate": 1.677477402098162e-06,
+      "loss": 0.1633,
+      "step": 16097
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9176788598790777,
+      "learning_rate": 1.6765644402999216e-06,
+      "loss": 0.1544,
+      "step": 16098
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2709621204145147,
+      "learning_rate": 1.675651704275344e-06,
+      "loss": 0.1679,
+      "step": 16099
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1446384656419066,
+      "learning_rate": 1.674739194049183e-06,
+      "loss": 0.1589,
+      "step": 16100
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.02180679059322,
+      "learning_rate": 1.673826909646189e-06,
+      "loss": 0.1534,
+      "step": 16101
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2679799246233314,
+      "learning_rate": 1.6729148510911142e-06,
+      "loss": 0.1636,
+      "step": 16102
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9948990950616705,
+      "learning_rate": 1.672003018408691e-06,
+      "loss": 0.1496,
+      "step": 16103
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.900105132623221,
+      "learning_rate": 1.6710914116236588e-06,
+      "loss": 0.1585,
+      "step": 16104
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9949865804409496,
+      "learning_rate": 1.6701800307607397e-06,
+      "loss": 0.1584,
+      "step": 16105
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.156298996343876,
+      "learning_rate": 1.6692688758446574e-06,
+      "loss": 0.174,
+      "step": 16106
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9389002868080762,
+      "learning_rate": 1.6683579469001287e-06,
+      "loss": 0.1793,
+      "step": 16107
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9776814651404512,
+      "learning_rate": 1.6674472439518629e-06,
+      "loss": 0.1497,
+      "step": 16108
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0709974373621516,
+      "learning_rate": 1.6665367670245592e-06,
+      "loss": 0.1841,
+      "step": 16109
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8930890109533807,
+      "learning_rate": 1.6656265161429186e-06,
+      "loss": 0.1682,
+      "step": 16110
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8296255055125633,
+      "learning_rate": 1.664716491331626e-06,
+      "loss": 0.1513,
+      "step": 16111
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9478955423947181,
+      "learning_rate": 1.6638066926153728e-06,
+      "loss": 0.1401,
+      "step": 16112
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9566375845034522,
+      "learning_rate": 1.6628971200188316e-06,
+      "loss": 0.1619,
+      "step": 16113
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7919743312019453,
+      "learning_rate": 1.66198777356668e-06,
+      "loss": 0.1513,
+      "step": 16114
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0115909183001128,
+      "learning_rate": 1.6610786532835776e-06,
+      "loss": 0.1706,
+      "step": 16115
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.92772352473913,
+      "learning_rate": 1.660169759194188e-06,
+      "loss": 0.1635,
+      "step": 16116
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1785822029161883,
+      "learning_rate": 1.6592610913231665e-06,
+      "loss": 0.1683,
+      "step": 16117
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9009097664645848,
+      "learning_rate": 1.6583526496951573e-06,
+      "loss": 0.1604,
+      "step": 16118
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8481746888363388,
+      "learning_rate": 1.657444434334805e-06,
+      "loss": 0.1726,
+      "step": 16119
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0707301830080818,
+      "learning_rate": 1.656536445266742e-06,
+      "loss": 0.185,
+      "step": 16120
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1814006659495366,
+      "learning_rate": 1.655628682515602e-06,
+      "loss": 0.1763,
+      "step": 16121
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2437639723178673,
+      "learning_rate": 1.654721146106002e-06,
+      "loss": 0.1796,
+      "step": 16122
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7788299770768288,
+      "learning_rate": 1.6538138360625633e-06,
+      "loss": 0.1792,
+      "step": 16123
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0035397369570327,
+      "learning_rate": 1.652906752409894e-06,
+      "loss": 0.1646,
+      "step": 16124
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0281512783310551,
+      "learning_rate": 1.6519998951726045e-06,
+      "loss": 0.1858,
+      "step": 16125
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8408852550415522,
+      "learning_rate": 1.6510932643752863e-06,
+      "loss": 0.1609,
+      "step": 16126
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0078754787295523,
+      "learning_rate": 1.6501868600425374e-06,
+      "loss": 0.1689,
+      "step": 16127
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8816314300651446,
+      "learning_rate": 1.6492806821989393e-06,
+      "loss": 0.141,
+      "step": 16128
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0799580233562474,
+      "learning_rate": 1.6483747308690768e-06,
+      "loss": 0.1734,
+      "step": 16129
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.128032995538658,
+      "learning_rate": 1.6474690060775178e-06,
+      "loss": 0.1487,
+      "step": 16130
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2278270712864054,
+      "learning_rate": 1.6465635078488372e-06,
+      "loss": 0.1667,
+      "step": 16131
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.894278260754333,
+      "learning_rate": 1.6456582362075911e-06,
+      "loss": 0.1604,
+      "step": 16132
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8471791910866014,
+      "learning_rate": 1.6447531911783365e-06,
+      "loss": 0.1685,
+      "step": 16133
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4038276898923099,
+      "learning_rate": 1.6438483727856268e-06,
+      "loss": 0.166,
+      "step": 16134
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8586239073799874,
+      "learning_rate": 1.6429437810539982e-06,
+      "loss": 0.1607,
+      "step": 16135
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.769070945114241,
+      "learning_rate": 1.6420394160079955e-06,
+      "loss": 0.1649,
+      "step": 16136
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3100985933899716,
+      "learning_rate": 1.6411352776721423e-06,
+      "loss": 0.2015,
+      "step": 16137
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2512570673071481,
+      "learning_rate": 1.640231366070969e-06,
+      "loss": 0.1608,
+      "step": 16138
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8822105121244441,
+      "learning_rate": 1.6393276812289905e-06,
+      "loss": 0.1631,
+      "step": 16139
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8806674958039984,
+      "learning_rate": 1.6384242231707203e-06,
+      "loss": 0.1791,
+      "step": 16140
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8521437055834901,
+      "learning_rate": 1.6375209919206657e-06,
+      "loss": 0.156,
+      "step": 16141
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.863487539470905,
+      "learning_rate": 1.6366179875033284e-06,
+      "loss": 0.1528,
+      "step": 16142
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3976787853176753,
+      "learning_rate": 1.6357152099431984e-06,
+      "loss": 0.169,
+      "step": 16143
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.139814567658281,
+      "learning_rate": 1.6348126592647684e-06,
+      "loss": 0.192,
+      "step": 16144
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1335595840636001,
+      "learning_rate": 1.6339103354925146e-06,
+      "loss": 0.1688,
+      "step": 16145
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8440126829948659,
+      "learning_rate": 1.6330082386509182e-06,
+      "loss": 0.1534,
+      "step": 16146
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8508049215827809,
+      "learning_rate": 1.6321063687644435e-06,
+      "loss": 0.1768,
+      "step": 16147
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.282254408840473,
+      "learning_rate": 1.631204725857558e-06,
+      "loss": 0.1668,
+      "step": 16148
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0052396010512652,
+      "learning_rate": 1.6303033099547149e-06,
+      "loss": 0.1659,
+      "step": 16149
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9286821198774534,
+      "learning_rate": 1.629402121080368e-06,
+      "loss": 0.1649,
+      "step": 16150
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7312340130080666,
+      "learning_rate": 1.6285011592589628e-06,
+      "loss": 0.153,
+      "step": 16151
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0912525370551018,
+      "learning_rate": 1.6276004245149346e-06,
+      "loss": 0.1601,
+      "step": 16152
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2664795393962827,
+      "learning_rate": 1.6266999168727204e-06,
+      "loss": 0.1802,
+      "step": 16153
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0799495507810202,
+      "learning_rate": 1.6257996363567408e-06,
+      "loss": 0.1593,
+      "step": 16154
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9822469573275092,
+      "learning_rate": 1.6248995829914216e-06,
+      "loss": 0.1661,
+      "step": 16155
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8909355640364796,
+      "learning_rate": 1.6239997568011723e-06,
+      "loss": 0.1796,
+      "step": 16156
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9762448493305782,
+      "learning_rate": 1.6231001578104045e-06,
+      "loss": 0.1688,
+      "step": 16157
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8456224303249009,
+      "learning_rate": 1.6222007860435153e-06,
+      "loss": 0.1545,
+      "step": 16158
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.097065634578124,
+      "learning_rate": 1.621301641524904e-06,
+      "loss": 0.1919,
+      "step": 16159
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2272594527185567,
+      "learning_rate": 1.6204027242789577e-06,
+      "loss": 0.1747,
+      "step": 16160
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4027279958659618,
+      "learning_rate": 1.619504034330064e-06,
+      "loss": 0.1336,
+      "step": 16161
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.18531326701348,
+      "learning_rate": 1.618605571702595e-06,
+      "loss": 0.1574,
+      "step": 16162
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2249771958511868,
+      "learning_rate": 1.617707336420925e-06,
+      "loss": 0.1596,
+      "step": 16163
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0243959933280622,
+      "learning_rate": 1.6168093285094144e-06,
+      "loss": 0.1816,
+      "step": 16164
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1661606544666774,
+      "learning_rate": 1.6159115479924259e-06,
+      "loss": 0.1564,
+      "step": 16165
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.535577913582053,
+      "learning_rate": 1.615013994894309e-06,
+      "loss": 0.1746,
+      "step": 16166
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.085710298402728,
+      "learning_rate": 1.6141166692394106e-06,
+      "loss": 0.134,
+      "step": 16167
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0799809442745136,
+      "learning_rate": 1.6132195710520716e-06,
+      "loss": 0.1563,
+      "step": 16168
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2622437343108193,
+      "learning_rate": 1.6123227003566267e-06,
+      "loss": 0.1582,
+      "step": 16169
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0677073332003861,
+      "learning_rate": 1.6114260571774031e-06,
+      "loss": 0.1644,
+      "step": 16170
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2140411169479177,
+      "learning_rate": 1.6105296415387194e-06,
+      "loss": 0.1616,
+      "step": 16171
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9562483027738354,
+      "learning_rate": 1.609633453464895e-06,
+      "loss": 0.1614,
+      "step": 16172
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2325748339085705,
+      "learning_rate": 1.6087374929802346e-06,
+      "loss": 0.1424,
+      "step": 16173
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0057048813635803,
+      "learning_rate": 1.6078417601090457e-06,
+      "loss": 0.1744,
+      "step": 16174
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0863081131932661,
+      "learning_rate": 1.606946254875621e-06,
+      "loss": 0.1825,
+      "step": 16175
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0069948602074392,
+      "learning_rate": 1.6060509773042533e-06,
+      "loss": 0.1572,
+      "step": 16176
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8300856535641151,
+      "learning_rate": 1.6051559274192275e-06,
+      "loss": 0.183,
+      "step": 16177
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8124015039311869,
+      "learning_rate": 1.604261105244823e-06,
+      "loss": 0.1743,
+      "step": 16178
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0108753370218027,
+      "learning_rate": 1.6033665108053075e-06,
+      "loss": 0.1524,
+      "step": 16179
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1061628747426537,
+      "learning_rate": 1.6024721441249525e-06,
+      "loss": 0.1644,
+      "step": 16180
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.137413127806049,
+      "learning_rate": 1.6015780052280128e-06,
+      "loss": 0.1657,
+      "step": 16181
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2868473563497702,
+      "learning_rate": 1.6006840941387458e-06,
+      "loss": 0.1818,
+      "step": 16182
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2457052080057864,
+      "learning_rate": 1.5997904108813944e-06,
+      "loss": 0.1723,
+      "step": 16183
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2843137230704258,
+      "learning_rate": 1.5988969554802058e-06,
+      "loss": 0.1634,
+      "step": 16184
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8464560307722944,
+      "learning_rate": 1.5980037279594097e-06,
+      "loss": 0.1469,
+      "step": 16185
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0749915701214785,
+      "learning_rate": 1.5971107283432363e-06,
+      "loss": 0.1841,
+      "step": 16186
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1146209065453505,
+      "learning_rate": 1.5962179566559112e-06,
+      "loss": 0.1895,
+      "step": 16187
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9494386154442932,
+      "learning_rate": 1.5953254129216467e-06,
+      "loss": 0.1494,
+      "step": 16188
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.082046499912269,
+      "learning_rate": 1.594433097164657e-06,
+      "loss": 0.1761,
+      "step": 16189
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0094160078152505,
+      "learning_rate": 1.593541009409143e-06,
+      "loss": 0.1646,
+      "step": 16190
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3228427682959791,
+      "learning_rate": 1.592649149679305e-06,
+      "loss": 0.1611,
+      "step": 16191
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2135503694029435,
+      "learning_rate": 1.5917575179993328e-06,
+      "loss": 0.1604,
+      "step": 16192
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4004802440515558,
+      "learning_rate": 1.5908661143934112e-06,
+      "loss": 0.1517,
+      "step": 16193
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8070297986487432,
+      "learning_rate": 1.5899749388857222e-06,
+      "loss": 0.1672,
+      "step": 16194
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9741897108941053,
+      "learning_rate": 1.5890839915004398e-06,
+      "loss": 0.1729,
+      "step": 16195
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.948612681485556,
+      "learning_rate": 1.5881932722617277e-06,
+      "loss": 0.1833,
+      "step": 16196
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.751015180435459,
+      "learning_rate": 1.5873027811937491e-06,
+      "loss": 0.1689,
+      "step": 16197
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8035718788803883,
+      "learning_rate": 1.5864125183206569e-06,
+      "loss": 0.1714,
+      "step": 16198
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8762858502283523,
+      "learning_rate": 1.5855224836666016e-06,
+      "loss": 0.1465,
+      "step": 16199
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2879099555445566,
+      "learning_rate": 1.584632677255723e-06,
+      "loss": 0.1748,
+      "step": 16200
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0847853329598949,
+      "learning_rate": 1.5837430991121594e-06,
+      "loss": 0.1581,
+      "step": 16201
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1504659150239045,
+      "learning_rate": 1.5828537492600382e-06,
+      "loss": 0.1577,
+      "step": 16202
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.439135514835605,
+      "learning_rate": 1.5819646277234834e-06,
+      "loss": 0.166,
+      "step": 16203
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.593461322594299,
+      "learning_rate": 1.581075734526617e-06,
+      "loss": 0.1686,
+      "step": 16204
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0218168743453442,
+      "learning_rate": 1.580187069693544e-06,
+      "loss": 0.1435,
+      "step": 16205
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.13250068726926,
+      "learning_rate": 1.5792986332483739e-06,
+      "loss": 0.17,
+      "step": 16206
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1497848504486927,
+      "learning_rate": 1.578410425215202e-06,
+      "loss": 0.1874,
+      "step": 16207
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9860777098312422,
+      "learning_rate": 1.577522445618126e-06,
+      "loss": 0.1784,
+      "step": 16208
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0653169406941323,
+      "learning_rate": 1.576634694481227e-06,
+      "loss": 0.1772,
+      "step": 16209
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3828187860236651,
+      "learning_rate": 1.575747171828589e-06,
+      "loss": 0.1707,
+      "step": 16210
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0105506162693179,
+      "learning_rate": 1.5748598776842838e-06,
+      "loss": 0.175,
+      "step": 16211
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0796724874430454,
+      "learning_rate": 1.5739728120723795e-06,
+      "loss": 0.1764,
+      "step": 16212
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8717548247982144,
+      "learning_rate": 1.573085975016938e-06,
+      "loss": 0.1559,
+      "step": 16213
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9807085943349985,
+      "learning_rate": 1.5721993665420187e-06,
+      "loss": 0.1687,
+      "step": 16214
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.243518809958635,
+      "learning_rate": 1.5713129866716647e-06,
+      "loss": 0.1695,
+      "step": 16215
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0002427361868018,
+      "learning_rate": 1.5704268354299246e-06,
+      "loss": 0.1571,
+      "step": 16216
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0513130593254636,
+      "learning_rate": 1.56954091284083e-06,
+      "loss": 0.1683,
+      "step": 16217
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1471167807414897,
+      "learning_rate": 1.5686552189284177e-06,
+      "loss": 0.172,
+      "step": 16218
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8267405172081425,
+      "learning_rate": 1.5677697537167048e-06,
+      "loss": 0.1513,
+      "step": 16219
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7890938430215125,
+      "learning_rate": 1.5668845172297143e-06,
+      "loss": 0.1588,
+      "step": 16220
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8541591590053499,
+      "learning_rate": 1.5659995094914603e-06,
+      "loss": 0.1799,
+      "step": 16221
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0495622903964184,
+      "learning_rate": 1.565114730525944e-06,
+      "loss": 0.1556,
+      "step": 16222
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0599533152387104,
+      "learning_rate": 1.564230180357168e-06,
+      "loss": 0.169,
+      "step": 16223
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1758205426339616,
+      "learning_rate": 1.5633458590091233e-06,
+      "loss": 0.1818,
+      "step": 16224
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3566302422746803,
+      "learning_rate": 1.5624617665058005e-06,
+      "loss": 0.1603,
+      "step": 16225
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0060617851367233,
+      "learning_rate": 1.5615779028711775e-06,
+      "loss": 0.1576,
+      "step": 16226
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6034621808942282,
+      "learning_rate": 1.5606942681292326e-06,
+      "loss": 0.1638,
+      "step": 16227
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0066324518206973,
+      "learning_rate": 1.55981086230393e-06,
+      "loss": 0.1641,
+      "step": 16228
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0712368190292123,
+      "learning_rate": 1.5589276854192336e-06,
+      "loss": 0.1839,
+      "step": 16229
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9567927807699808,
+      "learning_rate": 1.5580447374991003e-06,
+      "loss": 0.1688,
+      "step": 16230
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3476202419921466,
+      "learning_rate": 1.557162018567484e-06,
+      "loss": 0.1629,
+      "step": 16231
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8950686877683366,
+      "learning_rate": 1.5562795286483212e-06,
+      "loss": 0.1457,
+      "step": 16232
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9267746323940856,
+      "learning_rate": 1.555397267765556e-06,
+      "loss": 0.1918,
+      "step": 16233
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0000652875899563,
+      "learning_rate": 1.5545152359431149e-06,
+      "loss": 0.1721,
+      "step": 16234
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9567611863206363,
+      "learning_rate": 1.5536334332049274e-06,
+      "loss": 0.1734,
+      "step": 16235
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0972798771214651,
+      "learning_rate": 1.5527518595749068e-06,
+      "loss": 0.1634,
+      "step": 16236
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.130250553330175,
+      "learning_rate": 1.551870515076972e-06,
+      "loss": 0.172,
+      "step": 16237
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0199759185793813,
+      "learning_rate": 1.5509893997350245e-06,
+      "loss": 0.1778,
+      "step": 16238
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.033979382796841,
+      "learning_rate": 1.5501085135729666e-06,
+      "loss": 0.1737,
+      "step": 16239
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.348123338689501,
+      "learning_rate": 1.5492278566146945e-06,
+      "loss": 0.1746,
+      "step": 16240
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.066130979899349,
+      "learning_rate": 1.5483474288840927e-06,
+      "loss": 0.196,
+      "step": 16241
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9158187549060963,
+      "learning_rate": 1.5474672304050454e-06,
+      "loss": 0.1575,
+      "step": 16242
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1187810995278062,
+      "learning_rate": 1.5465872612014255e-06,
+      "loss": 0.172,
+      "step": 16243
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.019799657851233,
+      "learning_rate": 1.545707521297105e-06,
+      "loss": 0.1769,
+      "step": 16244
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0459380016808018,
+      "learning_rate": 1.5448280107159442e-06,
+      "loss": 0.1888,
+      "step": 16245
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.187399423034828,
+      "learning_rate": 1.5439487294818002e-06,
+      "loss": 0.1361,
+      "step": 16246
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0301682592357948,
+      "learning_rate": 1.543069677618525e-06,
+      "loss": 0.1436,
+      "step": 16247
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1966512262807896,
+      "learning_rate": 1.5421908551499653e-06,
+      "loss": 0.1637,
+      "step": 16248
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.800092552120167,
+      "learning_rate": 1.5413122620999533e-06,
+      "loss": 0.161,
+      "step": 16249
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8528766108510961,
+      "learning_rate": 1.540433898492326e-06,
+      "loss": 0.1634,
+      "step": 16250
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9868816636905955,
+      "learning_rate": 1.539555764350905e-06,
+      "loss": 0.167,
+      "step": 16251
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.856926643725097,
+      "learning_rate": 1.5386778596995144e-06,
+      "loss": 0.1434,
+      "step": 16252
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1659977976533071,
+      "learning_rate": 1.5378001845619616e-06,
+      "loss": 0.1716,
+      "step": 16253
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4201134946411582,
+      "learning_rate": 1.53692273896206e-06,
+      "loss": 0.1706,
+      "step": 16254
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1415900722887473,
+      "learning_rate": 1.5360455229236049e-06,
+      "loss": 0.1813,
+      "step": 16255
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1716332672969776,
+      "learning_rate": 1.5351685364703916e-06,
+      "loss": 0.1597,
+      "step": 16256
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9427550839380094,
+      "learning_rate": 1.5342917796262136e-06,
+      "loss": 0.1502,
+      "step": 16257
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8506175397006018,
+      "learning_rate": 1.533415252414846e-06,
+      "loss": 0.1593,
+      "step": 16258
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7706683930731257,
+      "learning_rate": 1.5325389548600711e-06,
+      "loss": 0.1607,
+      "step": 16259
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1888829820220763,
+      "learning_rate": 1.531662886985652e-06,
+      "loss": 0.1752,
+      "step": 16260
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6787971741140937,
+      "learning_rate": 1.5307870488153586e-06,
+      "loss": 0.1642,
+      "step": 16261
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.936195924156837,
+      "learning_rate": 1.529911440372942e-06,
+      "loss": 0.1665,
+      "step": 16262
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.5813626304960975,
+      "learning_rate": 1.5290360616821564e-06,
+      "loss": 0.167,
+      "step": 16263
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9985641711965435,
+      "learning_rate": 1.5281609127667451e-06,
+      "loss": 0.1679,
+      "step": 16264
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.9953769021880143,
+      "learning_rate": 1.5272859936504513e-06,
+      "loss": 0.1587,
+      "step": 16265
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.899874692531944,
+      "learning_rate": 1.5264113043569994e-06,
+      "loss": 0.1555,
+      "step": 16266
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1861633024929679,
+      "learning_rate": 1.5255368449101226e-06,
+      "loss": 0.18,
+      "step": 16267
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.033925166327829,
+      "learning_rate": 1.5246626153335364e-06,
+      "loss": 0.1426,
+      "step": 16268
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.049961418295712,
+      "learning_rate": 1.5237886156509563e-06,
+      "loss": 0.1673,
+      "step": 16269
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0985176283197147,
+      "learning_rate": 1.5229148458860865e-06,
+      "loss": 0.1601,
+      "step": 16270
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8934168171267731,
+      "learning_rate": 1.5220413060626327e-06,
+      "loss": 0.1644,
+      "step": 16271
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9981608465218564,
+      "learning_rate": 1.5211679962042858e-06,
+      "loss": 0.1549,
+      "step": 16272
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8914455120721887,
+      "learning_rate": 1.5202949163347348e-06,
+      "loss": 0.1378,
+      "step": 16273
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6112780017756578,
+      "learning_rate": 1.519422066477666e-06,
+      "loss": 0.1838,
+      "step": 16274
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8425598172314763,
+      "learning_rate": 1.5185494466567508e-06,
+      "loss": 0.1734,
+      "step": 16275
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0261225190109788,
+      "learning_rate": 1.5176770568956623e-06,
+      "loss": 0.1726,
+      "step": 16276
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1873182046443194,
+      "learning_rate": 1.5168048972180605e-06,
+      "loss": 0.1688,
+      "step": 16277
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3521035987276042,
+      "learning_rate": 1.5159329676476075e-06,
+      "loss": 0.1623,
+      "step": 16278
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2750024973858918,
+      "learning_rate": 1.5150612682079502e-06,
+      "loss": 0.1844,
+      "step": 16279
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3258702215505207,
+      "learning_rate": 1.5141897989227372e-06,
+      "loss": 0.1674,
+      "step": 16280
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9289221783042039,
+      "learning_rate": 1.513318559815603e-06,
+      "loss": 0.1771,
+      "step": 16281
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.074167772829532,
+      "learning_rate": 1.512447550910181e-06,
+      "loss": 0.1811,
+      "step": 16282
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9460771136548177,
+      "learning_rate": 1.5115767722301e-06,
+      "loss": 0.1523,
+      "step": 16283
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.925950211310057,
+      "learning_rate": 1.51070622379898e-06,
+      "loss": 0.1527,
+      "step": 16284
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0862374172733744,
+      "learning_rate": 1.509835905640431e-06,
+      "loss": 0.1697,
+      "step": 16285
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2381593824256332,
+      "learning_rate": 1.5089658177780653e-06,
+      "loss": 0.1697,
+      "step": 16286
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3238266766665978,
+      "learning_rate": 1.5080959602354783e-06,
+      "loss": 0.1588,
+      "step": 16287
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.170935256463546,
+      "learning_rate": 1.5072263330362713e-06,
+      "loss": 0.1631,
+      "step": 16288
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.085728386526513,
+      "learning_rate": 1.5063569362040265e-06,
+      "loss": 0.1783,
+      "step": 16289
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8514572170467741,
+      "learning_rate": 1.5054877697623305e-06,
+      "loss": 0.1727,
+      "step": 16290
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.031428838828624,
+      "learning_rate": 1.50461883373476e-06,
+      "loss": 0.1678,
+      "step": 16291
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1681429510390187,
+      "learning_rate": 1.5037501281448819e-06,
+      "loss": 0.1705,
+      "step": 16292
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.383254386252376,
+      "learning_rate": 1.5028816530162627e-06,
+      "loss": 0.1611,
+      "step": 16293
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0569013766570596,
+      "learning_rate": 1.5020134083724568e-06,
+      "loss": 0.1687,
+      "step": 16294
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.900329262047711,
+      "learning_rate": 1.5011453942370191e-06,
+      "loss": 0.1653,
+      "step": 16295
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2811474221674615,
+      "learning_rate": 1.5002776106334904e-06,
+      "loss": 0.1677,
+      "step": 16296
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1071821925985452,
+      "learning_rate": 1.4994100575854143e-06,
+      "loss": 0.1496,
+      "step": 16297
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7783188030566939,
+      "learning_rate": 1.4985427351163184e-06,
+      "loss": 0.1746,
+      "step": 16298
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3811377289457276,
+      "learning_rate": 1.4976756432497309e-06,
+      "loss": 0.1598,
+      "step": 16299
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7740242905028717,
+      "learning_rate": 1.4968087820091714e-06,
+      "loss": 0.1797,
+      "step": 16300
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2751691210341451,
+      "learning_rate": 1.495942151418156e-06,
+      "loss": 0.176,
+      "step": 16301
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8551080958388871,
+      "learning_rate": 1.495075751500188e-06,
+      "loss": 0.1475,
+      "step": 16302
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1155180459959604,
+      "learning_rate": 1.4942095822787738e-06,
+      "loss": 0.1723,
+      "step": 16303
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9582474774941909,
+      "learning_rate": 1.4933436437774017e-06,
+      "loss": 0.1623,
+      "step": 16304
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1581963477224246,
+      "learning_rate": 1.4924779360195662e-06,
+      "loss": 0.18,
+      "step": 16305
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1922788818545937,
+      "learning_rate": 1.4916124590287451e-06,
+      "loss": 0.1693,
+      "step": 16306
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7897651228922262,
+      "learning_rate": 1.4907472128284185e-06,
+      "loss": 0.1646,
+      "step": 16307
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2555341889630993,
+      "learning_rate": 1.489882197442053e-06,
+      "loss": 0.187,
+      "step": 16308
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7526073707918681,
+      "learning_rate": 1.4890174128931123e-06,
+      "loss": 0.1391,
+      "step": 16309
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.70102190020427,
+      "learning_rate": 1.4881528592050576e-06,
+      "loss": 0.164,
+      "step": 16310
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1421019175027005,
+      "learning_rate": 1.4872885364013357e-06,
+      "loss": 0.1652,
+      "step": 16311
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3468381523610138,
+      "learning_rate": 1.4864244445053956e-06,
+      "loss": 0.1671,
+      "step": 16312
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8954079863094079,
+      "learning_rate": 1.4855605835406695e-06,
+      "loss": 0.1435,
+      "step": 16313
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6833123666950103,
+      "learning_rate": 1.4846969535305967e-06,
+      "loss": 0.179,
+      "step": 16314
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1499521503608294,
+      "learning_rate": 1.4838335544985982e-06,
+      "loss": 0.1595,
+      "step": 16315
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.255100924441286,
+      "learning_rate": 1.4829703864680945e-06,
+      "loss": 0.1686,
+      "step": 16316
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.269949061927699,
+      "learning_rate": 1.482107449462501e-06,
+      "loss": 0.174,
+      "step": 16317
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0630711296146427,
+      "learning_rate": 1.4812447435052258e-06,
+      "loss": 0.1379,
+      "step": 16318
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1507195675909738,
+      "learning_rate": 1.4803822686196657e-06,
+      "loss": 0.1676,
+      "step": 16319
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0780534179912902,
+      "learning_rate": 1.4795200248292207e-06,
+      "loss": 0.1674,
+      "step": 16320
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1352401978328097,
+      "learning_rate": 1.4786580121572736e-06,
+      "loss": 0.1763,
+      "step": 16321
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2704926559864493,
+      "learning_rate": 1.477796230627211e-06,
+      "loss": 0.1644,
+      "step": 16322
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3269094679941507,
+      "learning_rate": 1.476934680262405e-06,
+      "loss": 0.1408,
+      "step": 16323
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9223432723723342,
+      "learning_rate": 1.4760733610862298e-06,
+      "loss": 0.1999,
+      "step": 16324
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0167153221285,
+      "learning_rate": 1.475212273122043e-06,
+      "loss": 0.154,
+      "step": 16325
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1643049382379205,
+      "learning_rate": 1.474351416393206e-06,
+      "loss": 0.1609,
+      "step": 16326
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3266734784664354,
+      "learning_rate": 1.47349079092307e-06,
+      "loss": 0.1696,
+      "step": 16327
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.965325287003813,
+      "learning_rate": 1.4726303967349754e-06,
+      "loss": 0.1585,
+      "step": 16328
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8263306753742028,
+      "learning_rate": 1.4717702338522654e-06,
+      "loss": 0.2012,
+      "step": 16329
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.655930629079804,
+      "learning_rate": 1.4709103022982673e-06,
+      "loss": 0.2196,
+      "step": 16330
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8648245666271341,
+      "learning_rate": 1.4700506020963113e-06,
+      "loss": 0.1682,
+      "step": 16331
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1845845429985948,
+      "learning_rate": 1.4691911332697118e-06,
+      "loss": 0.1737,
+      "step": 16332
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0154675997457938,
+      "learning_rate": 1.468331895841787e-06,
+      "loss": 0.1662,
+      "step": 16333
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0646982504956448,
+      "learning_rate": 1.4674728898358391e-06,
+      "loss": 0.1612,
+      "step": 16334
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.929233799062197,
+      "learning_rate": 1.466614115275171e-06,
+      "loss": 0.1785,
+      "step": 16335
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8890200594074377,
+      "learning_rate": 1.4657555721830775e-06,
+      "loss": 0.1637,
+      "step": 16336
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.407647389509816,
+      "learning_rate": 1.4648972605828482e-06,
+      "loss": 0.1699,
+      "step": 16337
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9429287949318823,
+      "learning_rate": 1.4640391804977605e-06,
+      "loss": 0.1524,
+      "step": 16338
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.389627543861334,
+      "learning_rate": 1.4631813319510945e-06,
+      "loss": 0.1587,
+      "step": 16339
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8792147979013831,
+      "learning_rate": 1.462323714966114e-06,
+      "loss": 0.1748,
+      "step": 16340
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1173411674101408,
+      "learning_rate": 1.461466329566088e-06,
+      "loss": 0.1689,
+      "step": 16341
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1589594154477718,
+      "learning_rate": 1.460609175774268e-06,
+      "loss": 0.1778,
+      "step": 16342
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0335774583431516,
+      "learning_rate": 1.4597522536139052e-06,
+      "loss": 0.143,
+      "step": 16343
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2014210515526043,
+      "learning_rate": 1.458895563108248e-06,
+      "loss": 0.1636,
+      "step": 16344
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2414386216197213,
+      "learning_rate": 1.4580391042805287e-06,
+      "loss": 0.1932,
+      "step": 16345
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2599250563452244,
+      "learning_rate": 1.4571828771539843e-06,
+      "loss": 0.1674,
+      "step": 16346
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1791440023288944,
+      "learning_rate": 1.4563268817518327e-06,
+      "loss": 0.1776,
+      "step": 16347
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0322052855841803,
+      "learning_rate": 1.4554711180973003e-06,
+      "loss": 0.1613,
+      "step": 16348
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2190450616956596,
+      "learning_rate": 1.4546155862135946e-06,
+      "loss": 0.1624,
+      "step": 16349
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.610583375351424,
+      "learning_rate": 1.4537602861239253e-06,
+      "loss": 0.1605,
+      "step": 16350
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.11298369910105,
+      "learning_rate": 1.452905217851489e-06,
+      "loss": 0.1929,
+      "step": 16351
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9145769935297287,
+      "learning_rate": 1.452050381419481e-06,
+      "loss": 0.1699,
+      "step": 16352
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.458729223003803,
+      "learning_rate": 1.4511957768510897e-06,
+      "loss": 0.1594,
+      "step": 16353
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1644354573343156,
+      "learning_rate": 1.4503414041694985e-06,
+      "loss": 0.1862,
+      "step": 16354
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9109531000011147,
+      "learning_rate": 1.4494872633978763e-06,
+      "loss": 0.1455,
+      "step": 16355
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.853462469807387,
+      "learning_rate": 1.4486333545593978e-06,
+      "loss": 0.1534,
+      "step": 16356
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6192208694248555,
+      "learning_rate": 1.4477796776772202e-06,
+      "loss": 0.1783,
+      "step": 16357
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1859687710840425,
+      "learning_rate": 1.4469262327745038e-06,
+      "loss": 0.1566,
+      "step": 16358
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1934652652801543,
+      "learning_rate": 1.4460730198743945e-06,
+      "loss": 0.1648,
+      "step": 16359
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8857142797389227,
+      "learning_rate": 1.445220039000037e-06,
+      "loss": 0.1575,
+      "step": 16360
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8879561711602736,
+      "learning_rate": 1.444367290174573e-06,
+      "loss": 0.1495,
+      "step": 16361
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9377892019061446,
+      "learning_rate": 1.4435147734211252e-06,
+      "loss": 0.1924,
+      "step": 16362
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.100278086878258,
+      "learning_rate": 1.442662488762826e-06,
+      "loss": 0.1658,
+      "step": 16363
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9895730310143408,
+      "learning_rate": 1.441810436222788e-06,
+      "loss": 0.1758,
+      "step": 16364
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2782203378181645,
+      "learning_rate": 1.4409586158241272e-06,
+      "loss": 0.1713,
+      "step": 16365
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0976176848106227,
+      "learning_rate": 1.4401070275899442e-06,
+      "loss": 0.1675,
+      "step": 16366
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9710678328379625,
+      "learning_rate": 1.4392556715433447e-06,
+      "loss": 0.1641,
+      "step": 16367
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3605252867567799,
+      "learning_rate": 1.438404547707417e-06,
+      "loss": 0.1721,
+      "step": 16368
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1191812979827758,
+      "learning_rate": 1.4375536561052483e-06,
+      "loss": 0.167,
+      "step": 16369
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9856198196977122,
+      "learning_rate": 1.4367029967599211e-06,
+      "loss": 0.1886,
+      "step": 16370
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1024483254210755,
+      "learning_rate": 1.4358525696945104e-06,
+      "loss": 0.1536,
+      "step": 16371
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8868452274281557,
+      "learning_rate": 1.4350023749320807e-06,
+      "loss": 0.1692,
+      "step": 16372
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0929098488958067,
+      "learning_rate": 1.4341524124956974e-06,
+      "loss": 0.1664,
+      "step": 16373
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1365025851174004,
+      "learning_rate": 1.4333026824084116e-06,
+      "loss": 0.1412,
+      "step": 16374
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0641312250521482,
+      "learning_rate": 1.4324531846932766e-06,
+      "loss": 0.1577,
+      "step": 16375
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1004909593029202,
+      "learning_rate": 1.4316039193733299e-06,
+      "loss": 0.1544,
+      "step": 16376
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.044385625193077,
+      "learning_rate": 1.4307548864716137e-06,
+      "loss": 0.1674,
+      "step": 16377
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4788242272864243,
+      "learning_rate": 1.4299060860111536e-06,
+      "loss": 0.1409,
+      "step": 16378
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.976932708881975,
+      "learning_rate": 1.4290575180149735e-06,
+      "loss": 0.155,
+      "step": 16379
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4220348323583278,
+      "learning_rate": 1.4282091825060963e-06,
+      "loss": 0.1681,
+      "step": 16380
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.172753518856644,
+      "learning_rate": 1.4273610795075255e-06,
+      "loss": 0.168,
+      "step": 16381
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0594369986964254,
+      "learning_rate": 1.4265132090422718e-06,
+      "loss": 0.1767,
+      "step": 16382
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9537846912091068,
+      "learning_rate": 1.42566557113333e-06,
+      "loss": 0.1582,
+      "step": 16383
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1865624630619633,
+      "learning_rate": 1.4248181658036964e-06,
+      "loss": 0.1691,
+      "step": 16384
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.086406895496996,
+      "learning_rate": 1.4239709930763513e-06,
+      "loss": 0.1747,
+      "step": 16385
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1570947276293209,
+      "learning_rate": 1.4231240529742774e-06,
+      "loss": 0.1715,
+      "step": 16386
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9852028749587856,
+      "learning_rate": 1.4222773455204486e-06,
+      "loss": 0.1396,
+      "step": 16387
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.275628718580738,
+      "learning_rate": 1.4214308707378333e-06,
+      "loss": 0.176,
+      "step": 16388
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2817426416927773,
+      "learning_rate": 1.4205846286493875e-06,
+      "loss": 0.1856,
+      "step": 16389
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0207781425195546,
+      "learning_rate": 1.4197386192780715e-06,
+      "loss": 0.1574,
+      "step": 16390
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.991995452101157,
+      "learning_rate": 1.4188928426468263e-06,
+      "loss": 0.1583,
+      "step": 16391
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4650315505964862,
+      "learning_rate": 1.418047298778601e-06,
+      "loss": 0.1735,
+      "step": 16392
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9804090300175904,
+      "learning_rate": 1.4172019876963249e-06,
+      "loss": 0.1867,
+      "step": 16393
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0081826340989326,
+      "learning_rate": 1.4163569094229311e-06,
+      "loss": 0.1488,
+      "step": 16394
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0979656503210717,
+      "learning_rate": 1.4155120639813392e-06,
+      "loss": 0.1768,
+      "step": 16395
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.648498219436163,
+      "learning_rate": 1.414667451394468e-06,
+      "loss": 0.1893,
+      "step": 16396
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8907662091834141,
+      "learning_rate": 1.4138230716852285e-06,
+      "loss": 0.189,
+      "step": 16397
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.169590311793009,
+      "learning_rate": 1.4129789248765214e-06,
+      "loss": 0.1481,
+      "step": 16398
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8896912266362786,
+      "learning_rate": 1.4121350109912479e-06,
+      "loss": 0.1561,
+      "step": 16399
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0289256443235757,
+      "learning_rate": 1.4112913300522946e-06,
+      "loss": 0.1538,
+      "step": 16400
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1660382935985822,
+      "learning_rate": 1.4104478820825518e-06,
+      "loss": 0.153,
+      "step": 16401
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0220030347435074,
+      "learning_rate": 1.4096046671048935e-06,
+      "loss": 0.1428,
+      "step": 16402
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7746742346123936,
+      "learning_rate": 1.4087616851421959e-06,
+      "loss": 0.1324,
+      "step": 16403
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.068496271057554,
+      "learning_rate": 1.4079189362173196e-06,
+      "loss": 0.1699,
+      "step": 16404
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0600589036312207,
+      "learning_rate": 1.4070764203531283e-06,
+      "loss": 0.155,
+      "step": 16405
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0528632347878315,
+      "learning_rate": 1.4062341375724742e-06,
+      "loss": 0.1697,
+      "step": 16406
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.084378023728591,
+      "learning_rate": 1.4053920878982074e-06,
+      "loss": 0.1613,
+      "step": 16407
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3434346377155673,
+      "learning_rate": 1.4045502713531623e-06,
+      "loss": 0.1797,
+      "step": 16408
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.131817937825781,
+      "learning_rate": 1.4037086879601803e-06,
+      "loss": 0.162,
+      "step": 16409
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.779090735975815,
+      "learning_rate": 1.4028673377420821e-06,
+      "loss": 0.1587,
+      "step": 16410
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.631104786572298,
+      "learning_rate": 1.402026220721695e-06,
+      "loss": 0.1926,
+      "step": 16411
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1120761340682184,
+      "learning_rate": 1.4011853369218308e-06,
+      "loss": 0.1774,
+      "step": 16412
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8984433358696586,
+      "learning_rate": 1.4003446863653004e-06,
+      "loss": 0.1529,
+      "step": 16413
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9389321223117489,
+      "learning_rate": 1.3995042690749072e-06,
+      "loss": 0.1876,
+      "step": 16414
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.002293730084705,
+      "learning_rate": 1.3986640850734444e-06,
+      "loss": 0.1472,
+      "step": 16415
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9460194544172318,
+      "learning_rate": 1.3978241343837073e-06,
+      "loss": 0.178,
+      "step": 16416
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9586613850300995,
+      "learning_rate": 1.396984417028473e-06,
+      "loss": 0.1549,
+      "step": 16417
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3849114738939412,
+      "learning_rate": 1.3961449330305255e-06,
+      "loss": 0.16,
+      "step": 16418
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1897351890626975,
+      "learning_rate": 1.3953056824126298e-06,
+      "loss": 0.1598,
+      "step": 16419
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1119253441096333,
+      "learning_rate": 1.3944666651975559e-06,
+      "loss": 0.1643,
+      "step": 16420
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9219902980961129,
+      "learning_rate": 1.3936278814080572e-06,
+      "loss": 0.1604,
+      "step": 16421
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2686997809339151,
+      "learning_rate": 1.3927893310668883e-06,
+      "loss": 0.1657,
+      "step": 16422
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1503442654242535,
+      "learning_rate": 1.3919510141967951e-06,
+      "loss": 0.1586,
+      "step": 16423
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0609452628827456,
+      "learning_rate": 1.3911129308205196e-06,
+      "loss": 0.1798,
+      "step": 16424
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9015507948091649,
+      "learning_rate": 1.390275080960789e-06,
+      "loss": 0.1503,
+      "step": 16425
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1521551730803656,
+      "learning_rate": 1.3894374646403363e-06,
+      "loss": 0.1621,
+      "step": 16426
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.111762810036442,
+      "learning_rate": 1.3886000818818758e-06,
+      "loss": 0.1691,
+      "step": 16427
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0080781136444183,
+      "learning_rate": 1.3877629327081266e-06,
+      "loss": 0.1465,
+      "step": 16428
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.299355270870439,
+      "learning_rate": 1.3869260171417919e-06,
+      "loss": 0.166,
+      "step": 16429
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1388745378816476,
+      "learning_rate": 1.3860893352055782e-06,
+      "loss": 0.1801,
+      "step": 16430
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1660929089576133,
+      "learning_rate": 1.3852528869221759e-06,
+      "loss": 0.153,
+      "step": 16431
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2058432239074868,
+      "learning_rate": 1.3844166723142748e-06,
+      "loss": 0.1906,
+      "step": 16432
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.496859015835789,
+      "learning_rate": 1.3835806914045602e-06,
+      "loss": 0.1554,
+      "step": 16433
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.102976200306153,
+      "learning_rate": 1.3827449442157049e-06,
+      "loss": 0.1776,
+      "step": 16434
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8612700307948432,
+      "learning_rate": 1.3819094307703807e-06,
+      "loss": 0.162,
+      "step": 16435
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2124658733496405,
+      "learning_rate": 1.3810741510912485e-06,
+      "loss": 0.1533,
+      "step": 16436
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2087711356790787,
+      "learning_rate": 1.380239105200969e-06,
+      "loss": 0.1724,
+      "step": 16437
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8197129714176736,
+      "learning_rate": 1.3794042931221873e-06,
+      "loss": 0.1483,
+      "step": 16438
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2090348799586166,
+      "learning_rate": 1.3785697148775522e-06,
+      "loss": 0.1589,
+      "step": 16439
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9053956920846441,
+      "learning_rate": 1.3777353704897002e-06,
+      "loss": 0.1658,
+      "step": 16440
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.970638279591743,
+      "learning_rate": 1.376901259981266e-06,
+      "loss": 0.1563,
+      "step": 16441
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8254486572617799,
+      "learning_rate": 1.3760673833748684e-06,
+      "loss": 0.1497,
+      "step": 16442
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0612951183055963,
+      "learning_rate": 1.3752337406931338e-06,
+      "loss": 0.1393,
+      "step": 16443
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.129746663017258,
+      "learning_rate": 1.3744003319586685e-06,
+      "loss": 0.162,
+      "step": 16444
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.166170052999444,
+      "learning_rate": 1.3735671571940835e-06,
+      "loss": 0.1663,
+      "step": 16445
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9240405167721508,
+      "learning_rate": 1.3727342164219736e-06,
+      "loss": 0.1682,
+      "step": 16446
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8770970913366118,
+      "learning_rate": 1.371901509664939e-06,
+      "loss": 0.1839,
+      "step": 16447
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3908562774318896,
+      "learning_rate": 1.3710690369455605e-06,
+      "loss": 0.1514,
+      "step": 16448
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5732484290072952,
+      "learning_rate": 1.3702367982864218e-06,
+      "loss": 0.1551,
+      "step": 16449
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0855678285096475,
+      "learning_rate": 1.3694047937100985e-06,
+      "loss": 0.1643,
+      "step": 16450
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2534297475243648,
+      "learning_rate": 1.368573023239157e-06,
+      "loss": 0.1929,
+      "step": 16451
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9189340263405366,
+      "learning_rate": 1.3677414868961615e-06,
+      "loss": 0.1644,
+      "step": 16452
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0245879139658436,
+      "learning_rate": 1.3669101847036625e-06,
+      "loss": 0.1689,
+      "step": 16453
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.392745604846661,
+      "learning_rate": 1.3660791166842158e-06,
+      "loss": 0.1812,
+      "step": 16454
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4550152590012073,
+      "learning_rate": 1.3652482828603575e-06,
+      "loss": 0.155,
+      "step": 16455
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0442879814309745,
+      "learning_rate": 1.3644176832546296e-06,
+      "loss": 0.184,
+      "step": 16456
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1630175865305374,
+      "learning_rate": 1.3635873178895587e-06,
+      "loss": 0.1712,
+      "step": 16457
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1786859632173177,
+      "learning_rate": 1.3627571867876689e-06,
+      "loss": 0.1613,
+      "step": 16458
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.825323167134559,
+      "learning_rate": 1.3619272899714776e-06,
+      "loss": 0.1654,
+      "step": 16459
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2754532680519124,
+      "learning_rate": 1.3610976274634991e-06,
+      "loss": 0.1724,
+      "step": 16460
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1863324872811745,
+      "learning_rate": 1.3602681992862333e-06,
+      "loss": 0.1731,
+      "step": 16461
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9164563337067313,
+      "learning_rate": 1.359439005462183e-06,
+      "loss": 0.1614,
+      "step": 16462
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4849730806306138,
+      "learning_rate": 1.3586100460138352e-06,
+      "loss": 0.1425,
+      "step": 16463
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9652478128597369,
+      "learning_rate": 1.3577813209636803e-06,
+      "loss": 0.1739,
+      "step": 16464
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.041259915146661,
+      "learning_rate": 1.3569528303341927e-06,
+      "loss": 0.1763,
+      "step": 16465
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0776118097768106,
+      "learning_rate": 1.356124574147848e-06,
+      "loss": 0.1636,
+      "step": 16466
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4982116007172273,
+      "learning_rate": 1.3552965524271144e-06,
+      "loss": 0.1715,
+      "step": 16467
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.091139820413839,
+      "learning_rate": 1.3544687651944476e-06,
+      "loss": 0.1667,
+      "step": 16468
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0532924603589213,
+      "learning_rate": 1.3536412124723075e-06,
+      "loss": 0.1559,
+      "step": 16469
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7324418230947063,
+      "learning_rate": 1.3528138942831337e-06,
+      "loss": 0.1611,
+      "step": 16470
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2857408269894968,
+      "learning_rate": 1.351986810649375e-06,
+      "loss": 0.1643,
+      "step": 16471
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.007133004425682,
+      "learning_rate": 1.35115996159346e-06,
+      "loss": 0.1704,
+      "step": 16472
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9087690813649096,
+      "learning_rate": 1.3503333471378211e-06,
+      "loss": 0.168,
+      "step": 16473
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3235653985397124,
+      "learning_rate": 1.3495069673048778e-06,
+      "loss": 0.1738,
+      "step": 16474
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0334826159258361,
+      "learning_rate": 1.3486808221170455e-06,
+      "loss": 0.1568,
+      "step": 16475
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1220531438873298,
+      "learning_rate": 1.3478549115967344e-06,
+      "loss": 0.1461,
+      "step": 16476
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7370009339817567,
+      "learning_rate": 1.3470292357663506e-06,
+      "loss": 0.1513,
+      "step": 16477
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0200812436144857,
+      "learning_rate": 1.3462037946482842e-06,
+      "loss": 0.1608,
+      "step": 16478
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8658443215049889,
+      "learning_rate": 1.3453785882649317e-06,
+      "loss": 0.1708,
+      "step": 16479
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9038392745163156,
+      "learning_rate": 1.3445536166386708e-06,
+      "loss": 0.1551,
+      "step": 16480
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8945414551418993,
+      "learning_rate": 1.3437288797918858e-06,
+      "loss": 0.1659,
+      "step": 16481
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2513579274863056,
+      "learning_rate": 1.3429043777469397e-06,
+      "loss": 0.1573,
+      "step": 16482
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.032128921002778,
+      "learning_rate": 1.3420801105262026e-06,
+      "loss": 0.1572,
+      "step": 16483
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9609598056949419,
+      "learning_rate": 1.3412560781520334e-06,
+      "loss": 0.1575,
+      "step": 16484
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9334010175995588,
+      "learning_rate": 1.3404322806467796e-06,
+      "loss": 0.1288,
+      "step": 16485
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3460503358341827,
+      "learning_rate": 1.339608718032791e-06,
+      "loss": 0.1533,
+      "step": 16486
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8418329982109378,
+      "learning_rate": 1.3387853903324032e-06,
+      "loss": 0.1772,
+      "step": 16487
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.667359050992339,
+      "learning_rate": 1.337962297567954e-06,
+      "loss": 0.1493,
+      "step": 16488
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0445679883208367,
+      "learning_rate": 1.3371394397617644e-06,
+      "loss": 0.1637,
+      "step": 16489
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8907911590724815,
+      "learning_rate": 1.3363168169361574e-06,
+      "loss": 0.1643,
+      "step": 16490
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0557707181235574,
+      "learning_rate": 1.3354944291134452e-06,
+      "loss": 0.1616,
+      "step": 16491
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8614216558652774,
+      "learning_rate": 1.3346722763159358e-06,
+      "loss": 0.1483,
+      "step": 16492
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.469119758561671,
+      "learning_rate": 1.3338503585659302e-06,
+      "loss": 0.1802,
+      "step": 16493
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8268722157076108,
+      "learning_rate": 1.3330286758857258e-06,
+      "loss": 0.1535,
+      "step": 16494
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.978780111321717,
+      "learning_rate": 1.3322072282976051e-06,
+      "loss": 0.1771,
+      "step": 16495
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7792839909962566,
+      "learning_rate": 1.3313860158238556e-06,
+      "loss": 0.156,
+      "step": 16496
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.130327790411797,
+      "learning_rate": 1.3305650384867475e-06,
+      "loss": 0.1761,
+      "step": 16497
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9698512769898063,
+      "learning_rate": 1.329744296308555e-06,
+      "loss": 0.1744,
+      "step": 16498
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7540272536679848,
+      "learning_rate": 1.3289237893115348e-06,
+      "loss": 0.1715,
+      "step": 16499
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9543124083561036,
+      "learning_rate": 1.3281035175179503e-06,
+      "loss": 0.1716,
+      "step": 16500
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0794734550478802,
+      "learning_rate": 1.3272834809500446e-06,
+      "loss": 0.176,
+      "step": 16501
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.986549355968139,
+      "learning_rate": 1.3264636796300646e-06,
+      "loss": 0.1717,
+      "step": 16502
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9899451300872415,
+      "learning_rate": 1.325644113580249e-06,
+      "loss": 0.1564,
+      "step": 16503
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1125736143872642,
+      "learning_rate": 1.3248247828228244e-06,
+      "loss": 0.1646,
+      "step": 16504
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1028198395225064,
+      "learning_rate": 1.324005687380021e-06,
+      "loss": 0.1556,
+      "step": 16505
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9268461161989021,
+      "learning_rate": 1.32318682727405e-06,
+      "loss": 0.1824,
+      "step": 16506
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2747726043601,
+      "learning_rate": 1.322368202527129e-06,
+      "loss": 0.1638,
+      "step": 16507
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.851572183953692,
+      "learning_rate": 1.321549813161458e-06,
+      "loss": 0.1739,
+      "step": 16508
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2012203043166474,
+      "learning_rate": 1.3207316591992392e-06,
+      "loss": 0.1578,
+      "step": 16509
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.631785569122105,
+      "learning_rate": 1.3199137406626639e-06,
+      "loss": 0.1582,
+      "step": 16510
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8110969696411472,
+      "learning_rate": 1.319096057573921e-06,
+      "loss": 0.1816,
+      "step": 16511
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4847912761305304,
+      "learning_rate": 1.3182786099551848e-06,
+      "loss": 0.1669,
+      "step": 16512
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0646581402910877,
+      "learning_rate": 1.3174613978286355e-06,
+      "loss": 0.1706,
+      "step": 16513
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5209956250155796,
+      "learning_rate": 1.3166444212164331e-06,
+      "loss": 0.1956,
+      "step": 16514
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0073009213634985,
+      "learning_rate": 1.3158276801407432e-06,
+      "loss": 0.1748,
+      "step": 16515
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9904598112902595,
+      "learning_rate": 1.3150111746237159e-06,
+      "loss": 0.1339,
+      "step": 16516
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4137658796525223,
+      "learning_rate": 1.3141949046875025e-06,
+      "loss": 0.1742,
+      "step": 16517
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1149818676663423,
+      "learning_rate": 1.3133788703542417e-06,
+      "loss": 0.1771,
+      "step": 16518
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8472715236112703,
+      "learning_rate": 1.3125630716460692e-06,
+      "loss": 0.1746,
+      "step": 16519
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1760004963805697,
+      "learning_rate": 1.3117475085851173e-06,
+      "loss": 0.1757,
+      "step": 16520
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0916298607265162,
+      "learning_rate": 1.3109321811935017e-06,
+      "loss": 0.1651,
+      "step": 16521
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0303127188530168,
+      "learning_rate": 1.3101170894933436e-06,
+      "loss": 0.1641,
+      "step": 16522
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9849398321369848,
+      "learning_rate": 1.3093022335067485e-06,
+      "loss": 0.1786,
+      "step": 16523
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.926636407257029,
+      "learning_rate": 1.3084876132558233e-06,
+      "loss": 0.1583,
+      "step": 16524
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0249330189845316,
+      "learning_rate": 1.3076732287626603e-06,
+      "loss": 0.1666,
+      "step": 16525
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8608635334435888,
+      "learning_rate": 1.306859080049353e-06,
+      "loss": 0.163,
+      "step": 16526
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.077394151696253,
+      "learning_rate": 1.3060451671379837e-06,
+      "loss": 0.1793,
+      "step": 16527
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9431214779085131,
+      "learning_rate": 1.3052314900506292e-06,
+      "loss": 0.1718,
+      "step": 16528
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0518179327905288,
+      "learning_rate": 1.3044180488093616e-06,
+      "loss": 0.1525,
+      "step": 16529
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3463257700345999,
+      "learning_rate": 1.303604843436248e-06,
+      "loss": 0.1587,
+      "step": 16530
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0130415107260289,
+      "learning_rate": 1.3027918739533429e-06,
+      "loss": 0.1584,
+      "step": 16531
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2403872414203043,
+      "learning_rate": 1.3019791403826998e-06,
+      "loss": 0.158,
+      "step": 16532
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9707677779447362,
+      "learning_rate": 1.3011666427463631e-06,
+      "loss": 0.1448,
+      "step": 16533
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4342116326299943,
+      "learning_rate": 1.3003543810663744e-06,
+      "loss": 0.1634,
+      "step": 16534
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8825279913771847,
+      "learning_rate": 1.2995423553647623e-06,
+      "loss": 0.1418,
+      "step": 16535
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8585049453063847,
+      "learning_rate": 1.2987305656635541e-06,
+      "loss": 0.1763,
+      "step": 16536
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8345181790410855,
+      "learning_rate": 1.297919011984774e-06,
+      "loss": 0.192,
+      "step": 16537
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0837015041442009,
+      "learning_rate": 1.2971076943504302e-06,
+      "loss": 0.1513,
+      "step": 16538
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3773116668194152,
+      "learning_rate": 1.296296612782534e-06,
+      "loss": 0.1754,
+      "step": 16539
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0696309455674147,
+      "learning_rate": 1.2954857673030807e-06,
+      "loss": 0.1589,
+      "step": 16540
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9033196322937592,
+      "learning_rate": 1.2946751579340699e-06,
+      "loss": 0.1849,
+      "step": 16541
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9471419142153938,
+      "learning_rate": 1.293864784697486e-06,
+      "loss": 0.1392,
+      "step": 16542
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8082640463443478,
+      "learning_rate": 1.2930546476153128e-06,
+      "loss": 0.1456,
+      "step": 16543
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1840965734810238,
+      "learning_rate": 1.2922447467095222e-06,
+      "loss": 0.1835,
+      "step": 16544
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5552907721291551,
+      "learning_rate": 1.2914350820020837e-06,
+      "loss": 0.1479,
+      "step": 16545
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2297661172025343,
+      "learning_rate": 1.290625653514962e-06,
+      "loss": 0.157,
+      "step": 16546
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8249973525995244,
+      "learning_rate": 1.2898164612701125e-06,
+      "loss": 0.1489,
+      "step": 16547
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5170565974582284,
+      "learning_rate": 1.2890075052894812e-06,
+      "loss": 0.1853,
+      "step": 16548
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9236007876873691,
+      "learning_rate": 1.2881987855950162e-06,
+      "loss": 0.1636,
+      "step": 16549
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9777282906763574,
+      "learning_rate": 1.2873903022086487e-06,
+      "loss": 0.155,
+      "step": 16550
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1129280526793448,
+      "learning_rate": 1.2865820551523134e-06,
+      "loss": 0.1848,
+      "step": 16551
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.995012290601466,
+      "learning_rate": 1.2857740444479306e-06,
+      "loss": 0.1923,
+      "step": 16552
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.014297480330186,
+      "learning_rate": 1.2849662701174204e-06,
+      "loss": 0.1802,
+      "step": 16553
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3572272500601252,
+      "learning_rate": 1.28415873218269e-06,
+      "loss": 0.1404,
+      "step": 16554
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5575915436378995,
+      "learning_rate": 1.2833514306656468e-06,
+      "loss": 0.1473,
+      "step": 16555
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1382318316054767,
+      "learning_rate": 1.2825443655881897e-06,
+      "loss": 0.1705,
+      "step": 16556
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9193602383948707,
+      "learning_rate": 1.2817375369722074e-06,
+      "loss": 0.1576,
+      "step": 16557
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8721215551729223,
+      "learning_rate": 1.2809309448395891e-06,
+      "loss": 0.1705,
+      "step": 16558
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.222139036368183,
+      "learning_rate": 1.2801245892122095e-06,
+      "loss": 0.169,
+      "step": 16559
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0006821267173502,
+      "learning_rate": 1.2793184701119444e-06,
+      "loss": 0.1468,
+      "step": 16560
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8754388516894127,
+      "learning_rate": 1.2785125875606563e-06,
+      "loss": 0.1743,
+      "step": 16561
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8966656140349384,
+      "learning_rate": 1.277706941580208e-06,
+      "loss": 0.1775,
+      "step": 16562
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2780190755664806,
+      "learning_rate": 1.2769015321924506e-06,
+      "loss": 0.1712,
+      "step": 16563
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9360785334190713,
+      "learning_rate": 1.2760963594192332e-06,
+      "loss": 0.1488,
+      "step": 16564
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.910441612685864,
+      "learning_rate": 1.2752914232823942e-06,
+      "loss": 0.1678,
+      "step": 16565
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5132061912551134,
+      "learning_rate": 1.2744867238037695e-06,
+      "loss": 0.1779,
+      "step": 16566
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0221427405596983,
+      "learning_rate": 1.2736822610051825e-06,
+      "loss": 0.1525,
+      "step": 16567
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9288590580525012,
+      "learning_rate": 1.2728780349084603e-06,
+      "loss": 0.1559,
+      "step": 16568
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1899146088344454,
+      "learning_rate": 1.272074045535412e-06,
+      "loss": 0.1549,
+      "step": 16569
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.185011658713729,
+      "learning_rate": 1.271270292907849e-06,
+      "loss": 0.1754,
+      "step": 16570
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.110425755207929,
+      "learning_rate": 1.270466777047572e-06,
+      "loss": 0.1713,
+      "step": 16571
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8477743760141506,
+      "learning_rate": 1.2696634979763757e-06,
+      "loss": 0.145,
+      "step": 16572
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.876659589143047,
+      "learning_rate": 1.2688604557160523e-06,
+      "loss": 0.1719,
+      "step": 16573
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.931464501268706,
+      "learning_rate": 1.26805765028838e-06,
+      "loss": 0.1742,
+      "step": 16574
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0268481379315793,
+      "learning_rate": 1.2672550817151397e-06,
+      "loss": 0.1778,
+      "step": 16575
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9313058948325292,
+      "learning_rate": 1.2664527500180956e-06,
+      "loss": 0.1481,
+      "step": 16576
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.041165067742634,
+      "learning_rate": 1.2656506552190163e-06,
+      "loss": 0.1793,
+      "step": 16577
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3093192642434583,
+      "learning_rate": 1.264848797339655e-06,
+      "loss": 0.1745,
+      "step": 16578
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1592795347577627,
+      "learning_rate": 1.2640471764017625e-06,
+      "loss": 0.165,
+      "step": 16579
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5147033431066657,
+      "learning_rate": 1.2632457924270835e-06,
+      "loss": 0.1658,
+      "step": 16580
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9477516005638641,
+      "learning_rate": 1.2624446454373596e-06,
+      "loss": 0.162,
+      "step": 16581
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6145234740898102,
+      "learning_rate": 1.2616437354543142e-06,
+      "loss": 0.1783,
+      "step": 16582
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9709781790920385,
+      "learning_rate": 1.2608430624996793e-06,
+      "loss": 0.1613,
+      "step": 16583
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0059251190145087,
+      "learning_rate": 1.2600426265951671e-06,
+      "loss": 0.1655,
+      "step": 16584
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.852242224906641,
+      "learning_rate": 1.2592424277624948e-06,
+      "loss": 0.1798,
+      "step": 16585
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9307410637537811,
+      "learning_rate": 1.2584424660233641e-06,
+      "loss": 0.1717,
+      "step": 16586
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3261276227772267,
+      "learning_rate": 1.2576427413994764e-06,
+      "loss": 0.1568,
+      "step": 16587
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1057900784109742,
+      "learning_rate": 1.2568432539125207e-06,
+      "loss": 0.1714,
+      "step": 16588
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0409802751492843,
+      "learning_rate": 1.256044003584186e-06,
+      "loss": 0.1449,
+      "step": 16589
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.189379870667062,
+      "learning_rate": 1.255244990436153e-06,
+      "loss": 0.1788,
+      "step": 16590
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9420416050470736,
+      "learning_rate": 1.2544462144900926e-06,
+      "loss": 0.1716,
+      "step": 16591
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7131804242727195,
+      "learning_rate": 1.253647675767674e-06,
+      "loss": 0.1671,
+      "step": 16592
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8651428576530652,
+      "learning_rate": 1.2528493742905533e-06,
+      "loss": 0.1541,
+      "step": 16593
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9329106635611252,
+      "learning_rate": 1.252051310080391e-06,
+      "loss": 0.1493,
+      "step": 16594
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8348058446150314,
+      "learning_rate": 1.2512534831588285e-06,
+      "loss": 0.1701,
+      "step": 16595
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3710651965857372,
+      "learning_rate": 1.2504558935475108e-06,
+      "loss": 0.1533,
+      "step": 16596
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.540429321961662,
+      "learning_rate": 1.2496585412680696e-06,
+      "loss": 0.1754,
+      "step": 16597
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9723673657195523,
+      "learning_rate": 1.2488614263421338e-06,
+      "loss": 0.1668,
+      "step": 16598
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8723989382271515,
+      "learning_rate": 1.248064548791328e-06,
+      "loss": 0.2018,
+      "step": 16599
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1434520514035937,
+      "learning_rate": 1.2472679086372662e-06,
+      "loss": 0.1568,
+      "step": 16600
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3246782154021164,
+      "learning_rate": 1.2464715059015553e-06,
+      "loss": 0.1793,
+      "step": 16601
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.537808161865273,
+      "learning_rate": 1.2456753406058008e-06,
+      "loss": 0.1464,
+      "step": 16602
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4533241913453714,
+      "learning_rate": 1.2448794127715947e-06,
+      "loss": 0.1396,
+      "step": 16603
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9080860744875678,
+      "learning_rate": 1.2440837224205316e-06,
+      "loss": 0.1399,
+      "step": 16604
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.470295143980769,
+      "learning_rate": 1.243288269574191e-06,
+      "loss": 0.1974,
+      "step": 16605
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.04915953083232,
+      "learning_rate": 1.242493054254149e-06,
+      "loss": 0.1341,
+      "step": 16606
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.044190308506685,
+      "learning_rate": 1.2416980764819807e-06,
+      "loss": 0.1829,
+      "step": 16607
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.867774746271164,
+      "learning_rate": 1.2409033362792444e-06,
+      "loss": 0.1518,
+      "step": 16608
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.054702853097598,
+      "learning_rate": 1.2401088336675015e-06,
+      "loss": 0.1585,
+      "step": 16609
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9045444257729103,
+      "learning_rate": 1.2393145686682995e-06,
+      "loss": 0.1741,
+      "step": 16610
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4045243451592733,
+      "learning_rate": 1.2385205413031865e-06,
+      "loss": 0.179,
+      "step": 16611
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8820536894164042,
+      "learning_rate": 1.2377267515936964e-06,
+      "loss": 0.1542,
+      "step": 16612
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.015566468698118,
+      "learning_rate": 1.2369331995613664e-06,
+      "loss": 0.1619,
+      "step": 16613
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.103506609341633,
+      "learning_rate": 1.2361398852277151e-06,
+      "loss": 0.1456,
+      "step": 16614
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9577327705290578,
+      "learning_rate": 1.2353468086142639e-06,
+      "loss": 0.1849,
+      "step": 16615
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0037833100661795,
+      "learning_rate": 1.2345539697425269e-06,
+      "loss": 0.1584,
+      "step": 16616
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8532619138403739,
+      "learning_rate": 1.2337613686340099e-06,
+      "loss": 0.16,
+      "step": 16617
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0251008349032038,
+      "learning_rate": 1.2329690053102085e-06,
+      "loss": 0.1654,
+      "step": 16618
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8976002504409775,
+      "learning_rate": 1.2321768797926203e-06,
+      "loss": 0.1532,
+      "step": 16619
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1929559472997977,
+      "learning_rate": 1.2313849921027277e-06,
+      "loss": 0.1759,
+      "step": 16620
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2062388730124642,
+      "learning_rate": 1.2305933422620143e-06,
+      "loss": 0.1706,
+      "step": 16621
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9055888863766589,
+      "learning_rate": 1.2298019302919505e-06,
+      "loss": 0.1615,
+      "step": 16622
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9225015025746164,
+      "learning_rate": 1.2290107562140053e-06,
+      "loss": 0.1593,
+      "step": 16623
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2805442002710812,
+      "learning_rate": 1.2282198200496377e-06,
+      "loss": 0.1569,
+      "step": 16624
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4657779384979428,
+      "learning_rate": 1.2274291218203027e-06,
+      "loss": 0.167,
+      "step": 16625
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0700714587634568,
+      "learning_rate": 1.22663866154745e-06,
+      "loss": 0.1695,
+      "step": 16626
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8939815710264033,
+      "learning_rate": 1.225848439252517e-06,
+      "loss": 0.1686,
+      "step": 16627
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.144969292046293,
+      "learning_rate": 1.2250584549569433e-06,
+      "loss": 0.1475,
+      "step": 16628
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9233843808729649,
+      "learning_rate": 1.2242687086821525e-06,
+      "loss": 0.1464,
+      "step": 16629
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6290684852204285,
+      "learning_rate": 1.2234792004495699e-06,
+      "loss": 0.1749,
+      "step": 16630
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.856185263973189,
+      "learning_rate": 1.2226899302806083e-06,
+      "loss": 0.1517,
+      "step": 16631
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0364409409561266,
+      "learning_rate": 1.2219008981966785e-06,
+      "loss": 0.1678,
+      "step": 16632
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.130986837273496,
+      "learning_rate": 1.221112104219182e-06,
+      "loss": 0.1577,
+      "step": 16633
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9271609905857692,
+      "learning_rate": 1.2203235483695176e-06,
+      "loss": 0.1397,
+      "step": 16634
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3835412288620175,
+      "learning_rate": 1.2195352306690711e-06,
+      "loss": 0.1585,
+      "step": 16635
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.924188954337587,
+      "learning_rate": 1.218747151139229e-06,
+      "loss": 0.1469,
+      "step": 16636
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1342662068725529,
+      "learning_rate": 1.2179593098013642e-06,
+      "loss": 0.1648,
+      "step": 16637
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.391040675182654,
+      "learning_rate": 1.2171717066768518e-06,
+      "loss": 0.1583,
+      "step": 16638
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3017551674300645,
+      "learning_rate": 1.2163843417870503e-06,
+      "loss": 0.1599,
+      "step": 16639
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2893603271246814,
+      "learning_rate": 1.2155972151533225e-06,
+      "loss": 0.1635,
+      "step": 16640
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5560227204195407,
+      "learning_rate": 1.2148103267970135e-06,
+      "loss": 0.1589,
+      "step": 16641
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3221619669349556,
+      "learning_rate": 1.2140236767394708e-06,
+      "loss": 0.1865,
+      "step": 16642
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5757234214722537,
+      "learning_rate": 1.213237265002034e-06,
+      "loss": 0.1658,
+      "step": 16643
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.029193912238693,
+      "learning_rate": 1.2124510916060307e-06,
+      "loss": 0.1572,
+      "step": 16644
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3440173887333735,
+      "learning_rate": 1.21166515657279e-06,
+      "loss": 0.1482,
+      "step": 16645
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.427243578192571,
+      "learning_rate": 1.2108794599236262e-06,
+      "loss": 0.1679,
+      "step": 16646
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2071076339340914,
+      "learning_rate": 1.2100940016798558e-06,
+      "loss": 0.1848,
+      "step": 16647
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1556199995839302,
+      "learning_rate": 1.2093087818627801e-06,
+      "loss": 0.1781,
+      "step": 16648
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8756934847230738,
+      "learning_rate": 1.2085238004937017e-06,
+      "loss": 0.1563,
+      "step": 16649
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0378254794811703,
+      "learning_rate": 1.2077390575939097e-06,
+      "loss": 0.1684,
+      "step": 16650
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5403138077505762,
+      "learning_rate": 1.2069545531846926e-06,
+      "loss": 0.1441,
+      "step": 16651
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.753894289360156,
+      "learning_rate": 1.2061702872873304e-06,
+      "loss": 0.1792,
+      "step": 16652
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9978037641503398,
+      "learning_rate": 1.205386259923097e-06,
+      "loss": 0.1488,
+      "step": 16653
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0916070622992498,
+      "learning_rate": 1.2046024711132564e-06,
+      "loss": 0.1893,
+      "step": 16654
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0336387552043056,
+      "learning_rate": 1.2038189208790718e-06,
+      "loss": 0.1595,
+      "step": 16655
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9740026133928388,
+      "learning_rate": 1.203035609241795e-06,
+      "loss": 0.1524,
+      "step": 16656
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2294148310206503,
+      "learning_rate": 1.2022525362226755e-06,
+      "loss": 0.1754,
+      "step": 16657
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9963912849910682,
+      "learning_rate": 1.201469701842951e-06,
+      "loss": 0.1686,
+      "step": 16658
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4602436474568625,
+      "learning_rate": 1.2006871061238578e-06,
+      "loss": 0.1516,
+      "step": 16659
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2083045646730413,
+      "learning_rate": 1.1999047490866255e-06,
+      "loss": 0.1754,
+      "step": 16660
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9319805806800955,
+      "learning_rate": 1.1991226307524727e-06,
+      "loss": 0.1651,
+      "step": 16661
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.057164534609886,
+      "learning_rate": 1.198340751142617e-06,
+      "loss": 0.1414,
+      "step": 16662
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2254343247363502,
+      "learning_rate": 1.1975591102782635e-06,
+      "loss": 0.1509,
+      "step": 16663
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8658829631174367,
+      "learning_rate": 1.1967777081806187e-06,
+      "loss": 0.1448,
+      "step": 16664
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.960613665093742,
+      "learning_rate": 1.1959965448708731e-06,
+      "loss": 0.1701,
+      "step": 16665
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.748125493424765,
+      "learning_rate": 1.1952156203702215e-06,
+      "loss": 0.1563,
+      "step": 16666
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9075724124031925,
+      "learning_rate": 1.1944349346998407e-06,
+      "loss": 0.1496,
+      "step": 16667
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9928967707047157,
+      "learning_rate": 1.1936544878809097e-06,
+      "loss": 0.1748,
+      "step": 16668
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0413533784476476,
+      "learning_rate": 1.1928742799345982e-06,
+      "loss": 0.1767,
+      "step": 16669
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9206089006391661,
+      "learning_rate": 1.1920943108820714e-06,
+      "loss": 0.1706,
+      "step": 16670
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2401044893499937,
+      "learning_rate": 1.1913145807444815e-06,
+      "loss": 0.1882,
+      "step": 16671
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.908453897796349,
+      "learning_rate": 1.1905350895429835e-06,
+      "loss": 0.1746,
+      "step": 16672
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8886683505554861,
+      "learning_rate": 1.1897558372987172e-06,
+      "loss": 0.1677,
+      "step": 16673
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4196567280330579,
+      "learning_rate": 1.1889768240328225e-06,
+      "loss": 0.1716,
+      "step": 16674
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0678208880582205,
+      "learning_rate": 1.1881980497664282e-06,
+      "loss": 0.1358,
+      "step": 16675
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0379333061722018,
+      "learning_rate": 1.1874195145206603e-06,
+      "loss": 0.1687,
+      "step": 16676
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.126038895773892,
+      "learning_rate": 1.1866412183166343e-06,
+      "loss": 0.1762,
+      "step": 16677
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0333889541378998,
+      "learning_rate": 1.1858631611754623e-06,
+      "loss": 0.1668,
+      "step": 16678
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1298072797466057,
+      "learning_rate": 1.185085343118253e-06,
+      "loss": 0.1697,
+      "step": 16679
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7944560604697056,
+      "learning_rate": 1.1843077641660994e-06,
+      "loss": 0.1484,
+      "step": 16680
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.622632460724044,
+      "learning_rate": 1.183530424340098e-06,
+      "loss": 0.1536,
+      "step": 16681
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2091299150586754,
+      "learning_rate": 1.1827533236613287e-06,
+      "loss": 0.1691,
+      "step": 16682
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0788931153139116,
+      "learning_rate": 1.1819764621508757e-06,
+      "loss": 0.1669,
+      "step": 16683
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4777706715450771,
+      "learning_rate": 1.1811998398298074e-06,
+      "loss": 0.1733,
+      "step": 16684
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9636743457847207,
+      "learning_rate": 1.1804234567191919e-06,
+      "loss": 0.1698,
+      "step": 16685
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.93649863813618,
+      "learning_rate": 1.1796473128400888e-06,
+      "loss": 0.1561,
+      "step": 16686
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1509318525011094,
+      "learning_rate": 1.178871408213551e-06,
+      "loss": 0.1648,
+      "step": 16687
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1851623200092514,
+      "learning_rate": 1.1780957428606232e-06,
+      "loss": 0.16,
+      "step": 16688
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.303285860586854,
+      "learning_rate": 1.1773203168023496e-06,
+      "loss": 0.1832,
+      "step": 16689
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5576203670591835,
+      "learning_rate": 1.1765451300597574e-06,
+      "loss": 0.1639,
+      "step": 16690
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3140735072944771,
+      "learning_rate": 1.1757701826538792e-06,
+      "loss": 0.1562,
+      "step": 16691
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0278507652035873,
+      "learning_rate": 1.1749954746057313e-06,
+      "loss": 0.166,
+      "step": 16692
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.224398636284123,
+      "learning_rate": 1.1742210059363312e-06,
+      "loss": 0.1546,
+      "step": 16693
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.26970783716945,
+      "learning_rate": 1.1734467766666835e-06,
+      "loss": 0.1856,
+      "step": 16694
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0206282492563024,
+      "learning_rate": 1.1726727868177902e-06,
+      "loss": 0.1519,
+      "step": 16695
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0267392565262938,
+      "learning_rate": 1.1718990364106476e-06,
+      "loss": 0.1599,
+      "step": 16696
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0081117309857632,
+      "learning_rate": 1.1711255254662413e-06,
+      "loss": 0.168,
+      "step": 16697
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3485046268234842,
+      "learning_rate": 1.1703522540055545e-06,
+      "loss": 0.1509,
+      "step": 16698
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1515961083542794,
+      "learning_rate": 1.1695792220495605e-06,
+      "loss": 0.1563,
+      "step": 16699
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1567499848884912,
+      "learning_rate": 1.1688064296192313e-06,
+      "loss": 0.1724,
+      "step": 16700
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7545024826691027,
+      "learning_rate": 1.1680338767355237e-06,
+      "loss": 0.1678,
+      "step": 16701
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.988731933262323,
+      "learning_rate": 1.1672615634193961e-06,
+      "loss": 0.1494,
+      "step": 16702
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7795116724077813,
+      "learning_rate": 1.1664894896917966e-06,
+      "loss": 0.1538,
+      "step": 16703
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.127248147568288,
+      "learning_rate": 1.1657176555736716e-06,
+      "loss": 0.182,
+      "step": 16704
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0378052019218376,
+      "learning_rate": 1.164946061085952e-06,
+      "loss": 0.1718,
+      "step": 16705
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1251788094051784,
+      "learning_rate": 1.1641747062495723e-06,
+      "loss": 0.1738,
+      "step": 16706
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0306360633053053,
+      "learning_rate": 1.163403591085449e-06,
+      "loss": 0.1581,
+      "step": 16707
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.934544355171937,
+      "learning_rate": 1.1626327156145055e-06,
+      "loss": 0.1486,
+      "step": 16708
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7447446102909068,
+      "learning_rate": 1.1618620798576474e-06,
+      "loss": 0.1702,
+      "step": 16709
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.399895022519527,
+      "learning_rate": 1.16109168383578e-06,
+      "loss": 0.1679,
+      "step": 16710
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2903232844612942,
+      "learning_rate": 1.1603215275697988e-06,
+      "loss": 0.1856,
+      "step": 16711
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.024990843444502,
+      "learning_rate": 1.159551611080596e-06,
+      "loss": 0.1592,
+      "step": 16712
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9194915982988668,
+      "learning_rate": 1.1587819343890561e-06,
+      "loss": 0.1671,
+      "step": 16713
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1353129904878652,
+      "learning_rate": 1.1580124975160534e-06,
+      "loss": 0.1731,
+      "step": 16714
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0024954784880729,
+      "learning_rate": 1.1572433004824635e-06,
+      "loss": 0.1533,
+      "step": 16715
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0136144334892372,
+      "learning_rate": 1.1564743433091463e-06,
+      "loss": 0.1551,
+      "step": 16716
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0058020332612845,
+      "learning_rate": 1.1557056260169653e-06,
+      "loss": 0.177,
+      "step": 16717
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7741571728019853,
+      "learning_rate": 1.1549371486267646e-06,
+      "loss": 0.153,
+      "step": 16718
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.121485078716398,
+      "learning_rate": 1.1541689111593969e-06,
+      "loss": 0.1623,
+      "step": 16719
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0291767208676794,
+      "learning_rate": 1.153400913635695e-06,
+      "loss": 0.1635,
+      "step": 16720
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1490953421687071,
+      "learning_rate": 1.1526331560764926e-06,
+      "loss": 0.1611,
+      "step": 16721
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.15884448329287,
+      "learning_rate": 1.151865638502615e-06,
+      "loss": 0.1759,
+      "step": 16722
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0724547886666227,
+      "learning_rate": 1.1510983609348847e-06,
+      "loss": 0.1575,
+      "step": 16723
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9482165988645026,
+      "learning_rate": 1.1503313233941082e-06,
+      "loss": 0.1598,
+      "step": 16724
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9189930992394797,
+      "learning_rate": 1.1495645259010969e-06,
+      "loss": 0.16,
+      "step": 16725
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6922036695384618,
+      "learning_rate": 1.148797968476646e-06,
+      "loss": 0.1426,
+      "step": 16726
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3946996712571365,
+      "learning_rate": 1.1480316511415513e-06,
+      "loss": 0.1821,
+      "step": 16727
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9618638738746309,
+      "learning_rate": 1.1472655739165961e-06,
+      "loss": 0.1481,
+      "step": 16728
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9534958513144793,
+      "learning_rate": 1.1464997368225629e-06,
+      "loss": 0.1691,
+      "step": 16729
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0621077501576741,
+      "learning_rate": 1.1457341398802269e-06,
+      "loss": 0.167,
+      "step": 16730
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1239092189624629,
+      "learning_rate": 1.1449687831103495e-06,
+      "loss": 0.1778,
+      "step": 16731
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1674994384296928,
+      "learning_rate": 1.1442036665336953e-06,
+      "loss": 0.1806,
+      "step": 16732
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9455081777025037,
+      "learning_rate": 1.1434387901710164e-06,
+      "loss": 0.1679,
+      "step": 16733
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0255477263215511,
+      "learning_rate": 1.142674154043062e-06,
+      "loss": 0.1815,
+      "step": 16734
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9171922536179344,
+      "learning_rate": 1.1419097581705686e-06,
+      "loss": 0.1512,
+      "step": 16735
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0202324495736854,
+      "learning_rate": 1.1411456025742763e-06,
+      "loss": 0.1703,
+      "step": 16736
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8282685116494622,
+      "learning_rate": 1.1403816872749074e-06,
+      "loss": 0.1707,
+      "step": 16737
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3741495687109866,
+      "learning_rate": 1.1396180122931854e-06,
+      "loss": 0.1663,
+      "step": 16738
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.944588868826035,
+      "learning_rate": 1.1388545776498262e-06,
+      "loss": 0.1568,
+      "step": 16739
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9649547498178261,
+      "learning_rate": 1.1380913833655383e-06,
+      "loss": 0.1646,
+      "step": 16740
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0557333070394006,
+      "learning_rate": 1.13732842946102e-06,
+      "loss": 0.1941,
+      "step": 16741
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2316283911922998,
+      "learning_rate": 1.136565715956971e-06,
+      "loss": 0.1602,
+      "step": 16742
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1441727807938522,
+      "learning_rate": 1.1358032428740763e-06,
+      "loss": 0.1568,
+      "step": 16743
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5479886581033788,
+      "learning_rate": 1.13504101023302e-06,
+      "loss": 0.1602,
+      "step": 16744
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.934032607491036,
+      "learning_rate": 1.134279018054475e-06,
+      "loss": 0.1683,
+      "step": 16745
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0609235864393567,
+      "learning_rate": 1.1335172663591155e-06,
+      "loss": 0.161,
+      "step": 16746
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6371308461071334,
+      "learning_rate": 1.1327557551675983e-06,
+      "loss": 0.1723,
+      "step": 16747
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1378631228418552,
+      "learning_rate": 1.1319944845005815e-06,
+      "loss": 0.1641,
+      "step": 16748
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0222296313110266,
+      "learning_rate": 1.1312334543787185e-06,
+      "loss": 0.1712,
+      "step": 16749
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9525488302209721,
+      "learning_rate": 1.130472664822646e-06,
+      "loss": 0.164,
+      "step": 16750
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.048597087332812,
+      "learning_rate": 1.1297121158530056e-06,
+      "loss": 0.1345,
+      "step": 16751
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9475471414124667,
+      "learning_rate": 1.1289518074904227e-06,
+      "loss": 0.1838,
+      "step": 16752
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9362631929645722,
+      "learning_rate": 1.1281917397555253e-06,
+      "loss": 0.1598,
+      "step": 16753
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9116553796982717,
+      "learning_rate": 1.127431912668926e-06,
+      "loss": 0.1609,
+      "step": 16754
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9891970094191186,
+      "learning_rate": 1.126672326251238e-06,
+      "loss": 0.1706,
+      "step": 16755
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4497468118338226,
+      "learning_rate": 1.125912980523064e-06,
+      "loss": 0.1674,
+      "step": 16756
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0089683525593092,
+      "learning_rate": 1.1251538755050029e-06,
+      "loss": 0.1587,
+      "step": 16757
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8990294325244518,
+      "learning_rate": 1.1243950112176428e-06,
+      "loss": 0.1645,
+      "step": 16758
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9831129016045187,
+      "learning_rate": 1.1236363876815705e-06,
+      "loss": 0.1588,
+      "step": 16759
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0042388056429106,
+      "learning_rate": 1.1228780049173616e-06,
+      "loss": 0.1534,
+      "step": 16760
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0436397492912155,
+      "learning_rate": 1.1221198629455898e-06,
+      "loss": 0.1647,
+      "step": 16761
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9136767477797114,
+      "learning_rate": 1.1213619617868154e-06,
+      "loss": 0.1388,
+      "step": 16762
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8794550315521854,
+      "learning_rate": 1.120604301461602e-06,
+      "loss": 0.1854,
+      "step": 16763
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1904734038906932,
+      "learning_rate": 1.1198468819904962e-06,
+      "loss": 0.1649,
+      "step": 16764
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0325953177890068,
+      "learning_rate": 1.1190897033940461e-06,
+      "loss": 0.166,
+      "step": 16765
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0705206526153064,
+      "learning_rate": 1.11833276569279e-06,
+      "loss": 0.1667,
+      "step": 16766
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.949411680260893,
+      "learning_rate": 1.117576068907258e-06,
+      "loss": 0.166,
+      "step": 16767
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9224277427372204,
+      "learning_rate": 1.116819613057979e-06,
+      "loss": 0.1555,
+      "step": 16768
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0286982699631861,
+      "learning_rate": 1.1160633981654679e-06,
+      "loss": 0.163,
+      "step": 16769
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9503726540756631,
+      "learning_rate": 1.1153074242502404e-06,
+      "loss": 0.1821,
+      "step": 16770
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0626090541232485,
+      "learning_rate": 1.1145516913327991e-06,
+      "loss": 0.1817,
+      "step": 16771
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8112805845435304,
+      "learning_rate": 1.1137961994336467e-06,
+      "loss": 0.1458,
+      "step": 16772
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6616339502407302,
+      "learning_rate": 1.1130409485732718e-06,
+      "loss": 0.1719,
+      "step": 16773
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4096929572061465,
+      "learning_rate": 1.112285938772164e-06,
+      "loss": 0.1431,
+      "step": 16774
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1876330357447296,
+      "learning_rate": 1.1115311700508026e-06,
+      "loss": 0.1736,
+      "step": 16775
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8458628499040595,
+      "learning_rate": 1.1107766424296606e-06,
+      "loss": 0.1515,
+      "step": 16776
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9829250335898947,
+      "learning_rate": 1.1100223559292035e-06,
+      "loss": 0.1588,
+      "step": 16777
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9993965169500876,
+      "learning_rate": 1.1092683105698943e-06,
+      "loss": 0.1751,
+      "step": 16778
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0923849752085002,
+      "learning_rate": 1.1085145063721814e-06,
+      "loss": 0.1757,
+      "step": 16779
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5467147394843548,
+      "learning_rate": 1.1077609433565173e-06,
+      "loss": 0.1703,
+      "step": 16780
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8122282454607503,
+      "learning_rate": 1.1070076215433367e-06,
+      "loss": 0.1697,
+      "step": 16781
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.49125635176288,
+      "learning_rate": 1.1062545409530778e-06,
+      "loss": 0.1824,
+      "step": 16782
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0610533686743882,
+      "learning_rate": 1.1055017016061687e-06,
+      "loss": 0.1772,
+      "step": 16783
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9301093497530798,
+      "learning_rate": 1.1047491035230262e-06,
+      "loss": 0.1504,
+      "step": 16784
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.229501628580653,
+      "learning_rate": 1.1039967467240687e-06,
+      "loss": 0.1473,
+      "step": 16785
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2435714542781264,
+      "learning_rate": 1.1032446312296995e-06,
+      "loss": 0.154,
+      "step": 16786
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9557053053008547,
+      "learning_rate": 1.102492757060325e-06,
+      "loss": 0.167,
+      "step": 16787
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9880025827983654,
+      "learning_rate": 1.1017411242363341e-06,
+      "loss": 0.147,
+      "step": 16788
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9448609802832272,
+      "learning_rate": 1.1009897327781204e-06,
+      "loss": 0.1805,
+      "step": 16789
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1482386908142481,
+      "learning_rate": 1.1002385827060602e-06,
+      "loss": 0.1621,
+      "step": 16790
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6367255102843754,
+      "learning_rate": 1.0994876740405314e-06,
+      "loss": 0.1771,
+      "step": 16791
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8989530912050467,
+      "learning_rate": 1.0987370068019021e-06,
+      "loss": 0.1603,
+      "step": 16792
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1553896505928736,
+      "learning_rate": 1.0979865810105371e-06,
+      "loss": 0.1405,
+      "step": 16793
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0036497212360274,
+      "learning_rate": 1.0972363966867861e-06,
+      "loss": 0.1559,
+      "step": 16794
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6289270363632566,
+      "learning_rate": 1.0964864538510022e-06,
+      "loss": 0.1923,
+      "step": 16795
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9544763364485803,
+      "learning_rate": 1.095736752523525e-06,
+      "loss": 0.1522,
+      "step": 16796
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0078909004640204,
+      "learning_rate": 1.094987292724693e-06,
+      "loss": 0.165,
+      "step": 16797
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.453663246336247,
+      "learning_rate": 1.0942380744748315e-06,
+      "loss": 0.1784,
+      "step": 16798
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6144793012492105,
+      "learning_rate": 1.0934890977942646e-06,
+      "loss": 0.1675,
+      "step": 16799
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9762171377498543,
+      "learning_rate": 1.0927403627033129e-06,
+      "loss": 0.1633,
+      "step": 16800
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2662144545477028,
+      "learning_rate": 1.0919918692222785e-06,
+      "loss": 0.1654,
+      "step": 16801
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0651887956028305,
+      "learning_rate": 1.091243617371469e-06,
+      "loss": 0.1502,
+      "step": 16802
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8561373873884675,
+      "learning_rate": 1.0904956071711792e-06,
+      "loss": 0.1681,
+      "step": 16803
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7538486873095505,
+      "learning_rate": 1.0897478386417003e-06,
+      "loss": 0.1783,
+      "step": 16804
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.043431297504966,
+      "learning_rate": 1.0890003118033132e-06,
+      "loss": 0.1787,
+      "step": 16805
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8472181882913274,
+      "learning_rate": 1.088253026676297e-06,
+      "loss": 0.1628,
+      "step": 16806
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9808913031348376,
+      "learning_rate": 1.08750598328092e-06,
+      "loss": 0.1607,
+      "step": 16807
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0819924372854248,
+      "learning_rate": 1.0867591816374456e-06,
+      "loss": 0.1702,
+      "step": 16808
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9482495395134791,
+      "learning_rate": 1.0860126217661326e-06,
+      "loss": 0.1651,
+      "step": 16809
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2523002163974752,
+      "learning_rate": 1.0852663036872324e-06,
+      "loss": 0.1727,
+      "step": 16810
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8939924425722398,
+      "learning_rate": 1.0845202274209842e-06,
+      "loss": 0.1637,
+      "step": 16811
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9276742350222884,
+      "learning_rate": 1.0837743929876321e-06,
+      "loss": 0.162,
+      "step": 16812
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9346752169512902,
+      "learning_rate": 1.0830288004073997e-06,
+      "loss": 0.1653,
+      "step": 16813
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8831929316357998,
+      "learning_rate": 1.0822834497005174e-06,
+      "loss": 0.1514,
+      "step": 16814
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8418644534814382,
+      "learning_rate": 1.0815383408871983e-06,
+      "loss": 0.1696,
+      "step": 16815
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9462951060958208,
+      "learning_rate": 1.080793473987657e-06,
+      "loss": 0.1689,
+      "step": 16816
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1273647831520743,
+      "learning_rate": 1.080048849022095e-06,
+      "loss": 0.1578,
+      "step": 16817
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.064506698976965,
+      "learning_rate": 1.079304466010712e-06,
+      "loss": 0.152,
+      "step": 16818
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9350482498821943,
+      "learning_rate": 1.0785603249737008e-06,
+      "loss": 0.1572,
+      "step": 16819
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8666707172937299,
+      "learning_rate": 1.0778164259312418e-06,
+      "loss": 0.1439,
+      "step": 16820
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2224856485223698,
+      "learning_rate": 1.0770727689035198e-06,
+      "loss": 0.163,
+      "step": 16821
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5428579120628447,
+      "learning_rate": 1.0763293539107e-06,
+      "loss": 0.1616,
+      "step": 16822
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1434133419606083,
+      "learning_rate": 1.0755861809729518e-06,
+      "loss": 0.1518,
+      "step": 16823
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8105189335585022,
+      "learning_rate": 1.0748432501104322e-06,
+      "loss": 0.1448,
+      "step": 16824
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9990081041082408,
+      "learning_rate": 1.074100561343292e-06,
+      "loss": 0.171,
+      "step": 16825
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9316776439039638,
+      "learning_rate": 1.0733581146916793e-06,
+      "loss": 0.1706,
+      "step": 16826
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9572566077695434,
+      "learning_rate": 1.0726159101757327e-06,
+      "loss": 0.1766,
+      "step": 16827
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2773673420161995,
+      "learning_rate": 1.0718739478155827e-06,
+      "loss": 0.1915,
+      "step": 16828
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9132114296361398,
+      "learning_rate": 1.0711322276313586e-06,
+      "loss": 0.1636,
+      "step": 16829
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8430574453562424,
+      "learning_rate": 1.0703907496431743e-06,
+      "loss": 0.1673,
+      "step": 16830
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9082397791162895,
+      "learning_rate": 1.0696495138711472e-06,
+      "loss": 0.1566,
+      "step": 16831
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.021608843054946,
+      "learning_rate": 1.06890852033538e-06,
+      "loss": 0.1549,
+      "step": 16832
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.528181675833744,
+      "learning_rate": 1.0681677690559743e-06,
+      "loss": 0.1617,
+      "step": 16833
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0214552074791603,
+      "learning_rate": 1.0674272600530223e-06,
+      "loss": 0.1514,
+      "step": 16834
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0212046421885956,
+      "learning_rate": 1.0666869933466085e-06,
+      "loss": 0.1613,
+      "step": 16835
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9574791998002623,
+      "learning_rate": 1.065946968956818e-06,
+      "loss": 0.1637,
+      "step": 16836
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0040720899152797,
+      "learning_rate": 1.0652071869037172e-06,
+      "loss": 0.1667,
+      "step": 16837
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9335152830552658,
+      "learning_rate": 1.0644676472073789e-06,
+      "loss": 0.1617,
+      "step": 16838
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1742512919227117,
+      "learning_rate": 1.0637283498878592e-06,
+      "loss": 0.1684,
+      "step": 16839
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1858526578379234,
+      "learning_rate": 1.0629892949652133e-06,
+      "loss": 0.1756,
+      "step": 16840
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1309226874477336,
+      "learning_rate": 1.0622504824594859e-06,
+      "loss": 0.1786,
+      "step": 16841
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9353874388661649,
+      "learning_rate": 1.0615119123907214e-06,
+      "loss": 0.1552,
+      "step": 16842
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8959094708241735,
+      "learning_rate": 1.060773584778949e-06,
+      "loss": 0.1619,
+      "step": 16843
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.089295782757551,
+      "learning_rate": 1.0600354996441986e-06,
+      "loss": 0.168,
+      "step": 16844
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9320442391189215,
+      "learning_rate": 1.0592976570064894e-06,
+      "loss": 0.1626,
+      "step": 16845
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1291690300178296,
+      "learning_rate": 1.058560056885838e-06,
+      "loss": 0.1674,
+      "step": 16846
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0781302845432605,
+      "learning_rate": 1.0578226993022488e-06,
+      "loss": 0.1682,
+      "step": 16847
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8344528027416395,
+      "learning_rate": 1.0570855842757255e-06,
+      "loss": 0.1715,
+      "step": 16848
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9485026795896946,
+      "learning_rate": 1.0563487118262583e-06,
+      "loss": 0.1544,
+      "step": 16849
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2366380866579516,
+      "learning_rate": 1.0556120819738403e-06,
+      "loss": 0.1746,
+      "step": 16850
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8921220522014538,
+      "learning_rate": 1.0548756947384475e-06,
+      "loss": 0.1773,
+      "step": 16851
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2807101360981994,
+      "learning_rate": 1.0541395501400564e-06,
+      "loss": 0.1778,
+      "step": 16852
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9657410535613355,
+      "learning_rate": 1.0534036481986375e-06,
+      "loss": 0.1783,
+      "step": 16853
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1306778682370806,
+      "learning_rate": 1.0526679889341484e-06,
+      "loss": 0.176,
+      "step": 16854
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8940590385512903,
+      "learning_rate": 1.0519325723665463e-06,
+      "loss": 0.1595,
+      "step": 16855
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.067914678261924,
+      "learning_rate": 1.0511973985157775e-06,
+      "loss": 0.1704,
+      "step": 16856
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.20909021403613,
+      "learning_rate": 1.0504624674017872e-06,
+      "loss": 0.1612,
+      "step": 16857
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0323079356985376,
+      "learning_rate": 1.0497277790445048e-06,
+      "loss": 0.1611,
+      "step": 16858
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8631782867591103,
+      "learning_rate": 1.0489933334638648e-06,
+      "loss": 0.1794,
+      "step": 16859
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1566319344478255,
+      "learning_rate": 1.0482591306797829e-06,
+      "loss": 0.1809,
+      "step": 16860
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1401885609706566,
+      "learning_rate": 1.0475251707121791e-06,
+      "loss": 0.1381,
+      "step": 16861
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5511236998777376,
+      "learning_rate": 1.0467914535809599e-06,
+      "loss": 0.152,
+      "step": 16862
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5753143634265765,
+      "learning_rate": 1.04605797930603e-06,
+      "loss": 0.1685,
+      "step": 16863
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1623307582591522,
+      "learning_rate": 1.0453247479072814e-06,
+      "loss": 0.1507,
+      "step": 16864
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1270531847545042,
+      "learning_rate": 1.0445917594046073e-06,
+      "loss": 0.1511,
+      "step": 16865
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0339182269036757,
+      "learning_rate": 1.043859013817885e-06,
+      "loss": 0.165,
+      "step": 16866
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0908660060529582,
+      "learning_rate": 1.0431265111669952e-06,
+      "loss": 0.152,
+      "step": 16867
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0022956563358218,
+      "learning_rate": 1.0423942514718043e-06,
+      "loss": 0.1529,
+      "step": 16868
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8736893626239377,
+      "learning_rate": 1.0416622347521732e-06,
+      "loss": 0.1714,
+      "step": 16869
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6477719177196377,
+      "learning_rate": 1.0409304610279603e-06,
+      "loss": 0.174,
+      "step": 16870
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3499421597610715,
+      "learning_rate": 1.0401989303190141e-06,
+      "loss": 0.1519,
+      "step": 16871
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1109980178568921,
+      "learning_rate": 1.039467642645181e-06,
+      "loss": 0.1718,
+      "step": 16872
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.226219785772608,
+      "learning_rate": 1.038736598026291e-06,
+      "loss": 0.1719,
+      "step": 16873
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.439364984206557,
+      "learning_rate": 1.03800579648218e-06,
+      "loss": 0.1446,
+      "step": 16874
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1988578257912608,
+      "learning_rate": 1.0372752380326645e-06,
+      "loss": 0.1809,
+      "step": 16875
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5934155252133868,
+      "learning_rate": 1.0365449226975677e-06,
+      "loss": 0.1803,
+      "step": 16876
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1377548167853007,
+      "learning_rate": 1.0358148504966935e-06,
+      "loss": 0.1548,
+      "step": 16877
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.436773878572512,
+      "learning_rate": 1.0350850214498486e-06,
+      "loss": 0.1562,
+      "step": 16878
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9801581415679828,
+      "learning_rate": 1.0343554355768282e-06,
+      "loss": 0.171,
+      "step": 16879
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7754791213662016,
+      "learning_rate": 1.0336260928974252e-06,
+      "loss": 0.1658,
+      "step": 16880
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0383938709850917,
+      "learning_rate": 1.0328969934314181e-06,
+      "loss": 0.1607,
+      "step": 16881
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8080092997766583,
+      "learning_rate": 1.0321681371985892e-06,
+      "loss": 0.1422,
+      "step": 16882
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0266233977427306,
+      "learning_rate": 1.0314395242187037e-06,
+      "loss": 0.1649,
+      "step": 16883
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4384666085729678,
+      "learning_rate": 1.0307111545115301e-06,
+      "loss": 0.1672,
+      "step": 16884
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1492817195259588,
+      "learning_rate": 1.0299830280968205e-06,
+      "loss": 0.1709,
+      "step": 16885
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0985167211243825,
+      "learning_rate": 1.02925514499433e-06,
+      "loss": 0.1668,
+      "step": 16886
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.032134037777604,
+      "learning_rate": 1.0285275052237987e-06,
+      "loss": 0.1665,
+      "step": 16887
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.078501377404676,
+      "learning_rate": 1.027800108804966e-06,
+      "loss": 0.1602,
+      "step": 16888
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9876329553773102,
+      "learning_rate": 1.027072955757563e-06,
+      "loss": 0.1541,
+      "step": 16889
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0754178281906779,
+      "learning_rate": 1.026346046101312e-06,
+      "loss": 0.1579,
+      "step": 16890
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2273783014245425,
+      "learning_rate": 1.0256193798559322e-06,
+      "loss": 0.169,
+      "step": 16891
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2721252163758574,
+      "learning_rate": 1.0248929570411327e-06,
+      "loss": 0.1998,
+      "step": 16892
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9013353020840622,
+      "learning_rate": 1.0241667776766196e-06,
+      "loss": 0.1588,
+      "step": 16893
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9338838445718988,
+      "learning_rate": 1.0234408417820884e-06,
+      "loss": 0.151,
+      "step": 16894
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1278581521392987,
+      "learning_rate": 1.0227151493772324e-06,
+      "loss": 0.1667,
+      "step": 16895
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0304483129985431,
+      "learning_rate": 1.0219897004817337e-06,
+      "loss": 0.1555,
+      "step": 16896
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1925611182107805,
+      "learning_rate": 1.0212644951152718e-06,
+      "loss": 0.1734,
+      "step": 16897
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.005882606197199,
+      "learning_rate": 1.0205395332975165e-06,
+      "loss": 0.1622,
+      "step": 16898
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0572915878224836,
+      "learning_rate": 1.019814815048137e-06,
+      "loss": 0.1744,
+      "step": 16899
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9945705083558409,
+      "learning_rate": 1.0190903403867847e-06,
+      "loss": 0.1624,
+      "step": 16900
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3335322820199638,
+      "learning_rate": 1.0183661093331165e-06,
+      "loss": 0.1597,
+      "step": 16901
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8729913239957138,
+      "learning_rate": 1.0176421219067734e-06,
+      "loss": 0.1777,
+      "step": 16902
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4870799694049563,
+      "learning_rate": 1.0169183781273962e-06,
+      "loss": 0.1648,
+      "step": 16903
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0104638473649,
+      "learning_rate": 1.0161948780146136e-06,
+      "loss": 0.1772,
+      "step": 16904
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.696486495966895,
+      "learning_rate": 1.0154716215880523e-06,
+      "loss": 0.1871,
+      "step": 16905
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9349237651600266,
+      "learning_rate": 1.014748608867333e-06,
+      "loss": 0.1545,
+      "step": 16906
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1359011223613704,
+      "learning_rate": 1.0140258398720625e-06,
+      "loss": 0.1712,
+      "step": 16907
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0986364477647335,
+      "learning_rate": 1.0133033146218518e-06,
+      "loss": 0.1805,
+      "step": 16908
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7640555565366878,
+      "learning_rate": 1.012581033136294e-06,
+      "loss": 0.1664,
+      "step": 16909
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2273472753204244,
+      "learning_rate": 1.0118589954349845e-06,
+      "loss": 0.1764,
+      "step": 16910
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0131939116632758,
+      "learning_rate": 1.0111372015375054e-06,
+      "loss": 0.1775,
+      "step": 16911
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7986491351427615,
+      "learning_rate": 1.01041565146344e-06,
+      "loss": 0.1536,
+      "step": 16912
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1023363192896058,
+      "learning_rate": 1.009694345232356e-06,
+      "loss": 0.1632,
+      "step": 16913
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0947374161202323,
+      "learning_rate": 1.00897328286382e-06,
+      "loss": 0.1689,
+      "step": 16914
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.305555176945722,
+      "learning_rate": 1.0082524643773916e-06,
+      "loss": 0.1832,
+      "step": 16915
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9597581032536183,
+      "learning_rate": 1.0075318897926255e-06,
+      "loss": 0.1506,
+      "step": 16916
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8767785279661219,
+      "learning_rate": 1.0068115591290628e-06,
+      "loss": 0.1808,
+      "step": 16917
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9193785028161882,
+      "learning_rate": 1.0060914724062454e-06,
+      "loss": 0.1582,
+      "step": 16918
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8982585434629353,
+      "learning_rate": 1.0053716296437034e-06,
+      "loss": 0.1622,
+      "step": 16919
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9873719152167707,
+      "learning_rate": 1.0046520308609664e-06,
+      "loss": 0.1576,
+      "step": 16920
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9879782842011524,
+      "learning_rate": 1.0039326760775492e-06,
+      "loss": 0.1362,
+      "step": 16921
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9697752776839753,
+      "learning_rate": 1.003213565312966e-06,
+      "loss": 0.1545,
+      "step": 16922
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8439307478501283,
+      "learning_rate": 1.0024946985867244e-06,
+      "loss": 0.1654,
+      "step": 16923
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9783270450751373,
+      "learning_rate": 1.0017760759183203e-06,
+      "loss": 0.1753,
+      "step": 16924
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0366479590941498,
+      "learning_rate": 1.0010576973272512e-06,
+      "loss": 0.1519,
+      "step": 16925
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4252173197466278,
+      "learning_rate": 1.0003395628329982e-06,
+      "loss": 0.1745,
+      "step": 16926
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0046565519981416,
+      "learning_rate": 9.996216724550445e-07,
+      "loss": 0.1617,
+      "step": 16927
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1973466911484718,
+      "learning_rate": 9.98904026212859e-07,
+      "loss": 0.1726,
+      "step": 16928
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.023743851749609,
+      "learning_rate": 9.981866241259131e-07,
+      "loss": 0.1789,
+      "step": 16929
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9877633466971224,
+      "learning_rate": 9.974694662136609e-07,
+      "loss": 0.1487,
+      "step": 16930
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3325695332137415,
+      "learning_rate": 9.967525524955579e-07,
+      "loss": 0.1973,
+      "step": 16931
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.949333856302942,
+      "learning_rate": 9.9603588299105e-07,
+      "loss": 0.1755,
+      "step": 16932
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9485511719358928,
+      "learning_rate": 9.9531945771958e-07,
+      "loss": 0.165,
+      "step": 16933
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9261506185655848,
+      "learning_rate": 9.946032767005752e-07,
+      "loss": 0.1637,
+      "step": 16934
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9878937517566453,
+      "learning_rate": 9.938873399534688e-07,
+      "loss": 0.1424,
+      "step": 16935
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9556000137729997,
+      "learning_rate": 9.931716474976738e-07,
+      "loss": 0.1391,
+      "step": 16936
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9511096528875889,
+      "learning_rate": 9.924561993526082e-07,
+      "loss": 0.1454,
+      "step": 16937
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5058829954646482,
+      "learning_rate": 9.917409955376778e-07,
+      "loss": 0.1548,
+      "step": 16938
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9453923029256704,
+      "learning_rate": 9.910260360722802e-07,
+      "loss": 0.1553,
+      "step": 16939
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.388034607806563,
+      "learning_rate": 9.903113209758098e-07,
+      "loss": 0.1567,
+      "step": 16940
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1517223486250001,
+      "learning_rate": 9.895968502676533e-07,
+      "loss": 0.1749,
+      "step": 16941
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8428053004420055,
+      "learning_rate": 9.888826239671934e-07,
+      "loss": 0.1547,
+      "step": 16942
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8433502857727055,
+      "learning_rate": 9.881686420937986e-07,
+      "loss": 0.1471,
+      "step": 16943
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.2771510094579086,
+      "learning_rate": 9.874549046668413e-07,
+      "loss": 0.1807,
+      "step": 16944
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9337510319746428,
+      "learning_rate": 9.867414117056763e-07,
+      "loss": 0.1473,
+      "step": 16945
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2713121575294868,
+      "learning_rate": 9.860281632296609e-07,
+      "loss": 0.1457,
+      "step": 16946
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0183252588457072,
+      "learning_rate": 9.8531515925814e-07,
+      "loss": 0.1712,
+      "step": 16947
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4441436451206293,
+      "learning_rate": 9.846023998104536e-07,
+      "loss": 0.1787,
+      "step": 16948
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8774364051310735,
+      "learning_rate": 9.838898849059364e-07,
+      "loss": 0.148,
+      "step": 16949
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8797088309314717,
+      "learning_rate": 9.831776145639182e-07,
+      "loss": 0.1535,
+      "step": 16950
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9116377986628547,
+      "learning_rate": 9.824655888037138e-07,
+      "loss": 0.1485,
+      "step": 16951
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.921263524194139,
+      "learning_rate": 9.817538076446409e-07,
+      "loss": 0.1454,
+      "step": 16952
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8002452664261146,
+      "learning_rate": 9.810422711060042e-07,
+      "loss": 0.1457,
+      "step": 16953
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0264341071482261,
+      "learning_rate": 9.80330979207108e-07,
+      "loss": 0.1914,
+      "step": 16954
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9512470084198945,
+      "learning_rate": 9.796199319672416e-07,
+      "loss": 0.1529,
+      "step": 16955
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9112980601957212,
+      "learning_rate": 9.78909129405694e-07,
+      "loss": 0.1685,
+      "step": 16956
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.146254875669652,
+      "learning_rate": 9.78198571541744e-07,
+      "loss": 0.1802,
+      "step": 16957
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.978711356522037,
+      "learning_rate": 9.774882583946688e-07,
+      "loss": 0.1465,
+      "step": 16958
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8759311125604019,
+      "learning_rate": 9.767781899837348e-07,
+      "loss": 0.1703,
+      "step": 16959
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8155043599843381,
+      "learning_rate": 9.760683663282012e-07,
+      "loss": 0.1608,
+      "step": 16960
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3248677378427802,
+      "learning_rate": 9.753587874473235e-07,
+      "loss": 0.176,
+      "step": 16961
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.061681244714024,
+      "learning_rate": 9.746494533603478e-07,
+      "loss": 0.1809,
+      "step": 16962
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8151729487186564,
+      "learning_rate": 9.739403640865164e-07,
+      "loss": 0.1664,
+      "step": 16963
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1969609035416964,
+      "learning_rate": 9.732315196450615e-07,
+      "loss": 0.1886,
+      "step": 16964
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3341288883277864,
+      "learning_rate": 9.725229200552123e-07,
+      "loss": 0.1751,
+      "step": 16965
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9657370156268987,
+      "learning_rate": 9.718145653361878e-07,
+      "loss": 0.197,
+      "step": 16966
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0373624882601507,
+      "learning_rate": 9.711064555072026e-07,
+      "loss": 0.1779,
+      "step": 16967
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0226918276226975,
+      "learning_rate": 9.703985905874646e-07,
+      "loss": 0.1604,
+      "step": 16968
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6206548522077415,
+      "learning_rate": 9.696909705961776e-07,
+      "loss": 0.17,
+      "step": 16969
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9537709638528886,
+      "learning_rate": 9.689835955525307e-07,
+      "loss": 0.1632,
+      "step": 16970
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2901100670976893,
+      "learning_rate": 9.682764654757149e-07,
+      "loss": 0.1727,
+      "step": 16971
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.843111449613659,
+      "learning_rate": 9.675695803849094e-07,
+      "loss": 0.1562,
+      "step": 16972
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0565847073406056,
+      "learning_rate": 9.668629402992902e-07,
+      "loss": 0.1769,
+      "step": 16973
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8774607355627476,
+      "learning_rate": 9.661565452380228e-07,
+      "loss": 0.1604,
+      "step": 16974
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8361969432025993,
+      "learning_rate": 9.654503952202687e-07,
+      "loss": 0.1505,
+      "step": 16975
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1873928931233717,
+      "learning_rate": 9.647444902651847e-07,
+      "loss": 0.1659,
+      "step": 16976
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6912378470493363,
+      "learning_rate": 9.640388303919156e-07,
+      "loss": 0.164,
+      "step": 16977
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9739879821560055,
+      "learning_rate": 9.63333415619605e-07,
+      "loss": 0.1601,
+      "step": 16978
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9162667771255657,
+      "learning_rate": 9.626282459673842e-07,
+      "loss": 0.169,
+      "step": 16979
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0533004190917632,
+      "learning_rate": 9.619233214543833e-07,
+      "loss": 0.1456,
+      "step": 16980
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1535147922097448,
+      "learning_rate": 9.612186420997227e-07,
+      "loss": 0.1706,
+      "step": 16981
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2602299328060675,
+      "learning_rate": 9.605142079225183e-07,
+      "loss": 0.1586,
+      "step": 16982
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.674843203986836,
+      "learning_rate": 9.598100189418736e-07,
+      "loss": 0.1344,
+      "step": 16983
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0166396943321496,
+      "learning_rate": 9.591060751768943e-07,
+      "loss": 0.1781,
+      "step": 16984
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0828299161586112,
+      "learning_rate": 9.584023766466721e-07,
+      "loss": 0.1953,
+      "step": 16985
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.089162076325917,
+      "learning_rate": 9.576989233702993e-07,
+      "loss": 0.1856,
+      "step": 16986
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1043827985285775,
+      "learning_rate": 9.569957153668507e-07,
+      "loss": 0.1617,
+      "step": 16987
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0517660021678024,
+      "learning_rate": 9.562927526554066e-07,
+      "loss": 0.1583,
+      "step": 16988
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5950084990409343,
+      "learning_rate": 9.555900352550308e-07,
+      "loss": 0.1689,
+      "step": 16989
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9618689152709855,
+      "learning_rate": 9.548875631847875e-07,
+      "loss": 0.1714,
+      "step": 16990
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9033812724850739,
+      "learning_rate": 9.541853364637299e-07,
+      "loss": 0.1571,
+      "step": 16991
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9739470181448753,
+      "learning_rate": 9.534833551109035e-07,
+      "loss": 0.1388,
+      "step": 16992
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9077329611081956,
+      "learning_rate": 9.527816191453531e-07,
+      "loss": 0.1683,
+      "step": 16993
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8965000939025348,
+      "learning_rate": 9.520801285861126e-07,
+      "loss": 0.1708,
+      "step": 16994
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8336839169695421,
+      "learning_rate": 9.513788834522108e-07,
+      "loss": 0.1565,
+      "step": 16995
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5352836241948775,
+      "learning_rate": 9.506778837626652e-07,
+      "loss": 0.1605,
+      "step": 16996
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9124484802811895,
+      "learning_rate": 9.499771295364957e-07,
+      "loss": 0.1536,
+      "step": 16997
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.10837471926042,
+      "learning_rate": 9.492766207927062e-07,
+      "loss": 0.1658,
+      "step": 16998
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.166103641007375,
+      "learning_rate": 9.485763575503015e-07,
+      "loss": 0.1366,
+      "step": 16999
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8544681293800753,
+      "learning_rate": 9.47876339828272e-07,
+      "loss": 0.1594,
+      "step": 17000
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1404024846192349,
+      "learning_rate": 9.471765676456079e-07,
+      "loss": 0.1733,
+      "step": 17001
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1349490458873757,
+      "learning_rate": 9.464770410212909e-07,
+      "loss": 0.1495,
+      "step": 17002
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2651117346806118,
+      "learning_rate": 9.457777599742979e-07,
+      "loss": 0.1682,
+      "step": 17003
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2507880576208228,
+      "learning_rate": 9.450787245235926e-07,
+      "loss": 0.1689,
+      "step": 17004
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9300768396542201,
+      "learning_rate": 9.443799346881388e-07,
+      "loss": 0.1604,
+      "step": 17005
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0231991740589423,
+      "learning_rate": 9.436813904868902e-07,
+      "loss": 0.1793,
+      "step": 17006
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0992909713211707,
+      "learning_rate": 9.429830919387972e-07,
+      "loss": 0.1623,
+      "step": 17007
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.870414843937221,
+      "learning_rate": 9.422850390627991e-07,
+      "loss": 0.1626,
+      "step": 17008
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3107033059662958,
+      "learning_rate": 9.415872318778285e-07,
+      "loss": 0.1745,
+      "step": 17009
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7913604862726662,
+      "learning_rate": 9.40889670402817e-07,
+      "loss": 0.1709,
+      "step": 17010
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3926226337980574,
+      "learning_rate": 9.401923546566838e-07,
+      "loss": 0.1497,
+      "step": 17011
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.255256928642862,
+      "learning_rate": 9.394952846583472e-07,
+      "loss": 0.1872,
+      "step": 17012
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8369092897709353,
+      "learning_rate": 9.387984604267109e-07,
+      "loss": 0.1717,
+      "step": 17013
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8268688499431834,
+      "learning_rate": 9.381018819806797e-07,
+      "loss": 0.152,
+      "step": 17014
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7473049573367447,
+      "learning_rate": 9.374055493391455e-07,
+      "loss": 0.1608,
+      "step": 17015
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.917773236637632,
+      "learning_rate": 9.367094625209983e-07,
+      "loss": 0.1705,
+      "step": 17016
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9269986492790633,
+      "learning_rate": 9.360136215451177e-07,
+      "loss": 0.1783,
+      "step": 17017
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9102869074299589,
+      "learning_rate": 9.353180264303818e-07,
+      "loss": 0.1713,
+      "step": 17018
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.847096642866557,
+      "learning_rate": 9.346226771956523e-07,
+      "loss": 0.1719,
+      "step": 17019
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.115828035313165,
+      "learning_rate": 9.339275738597975e-07,
+      "loss": 0.1617,
+      "step": 17020
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9756168196493309,
+      "learning_rate": 9.332327164416688e-07,
+      "loss": 0.173,
+      "step": 17021
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.120436059367365,
+      "learning_rate": 9.325381049601157e-07,
+      "loss": 0.1631,
+      "step": 17022
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9422116076804968,
+      "learning_rate": 9.318437394339774e-07,
+      "loss": 0.1624,
+      "step": 17023
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.919121126930504,
+      "learning_rate": 9.311496198820913e-07,
+      "loss": 0.1802,
+      "step": 17024
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.408213788699369,
+      "learning_rate": 9.304557463232844e-07,
+      "loss": 0.1517,
+      "step": 17025
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1067485928654712,
+      "learning_rate": 9.297621187763761e-07,
+      "loss": 0.1771,
+      "step": 17026
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1005835224655929,
+      "learning_rate": 9.290687372601814e-07,
+      "loss": 0.1662,
+      "step": 17027
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5768878819440173,
+      "learning_rate": 9.283756017935108e-07,
+      "loss": 0.1446,
+      "step": 17028
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8738222182636233,
+      "learning_rate": 9.276827123951648e-07,
+      "loss": 0.1458,
+      "step": 17029
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9600910172910317,
+      "learning_rate": 9.269900690839373e-07,
+      "loss": 0.1729,
+      "step": 17030
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9455883463851124,
+      "learning_rate": 9.262976718786176e-07,
+      "loss": 0.1481,
+      "step": 17031
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4184803589090373,
+      "learning_rate": 9.256055207979841e-07,
+      "loss": 0.1701,
+      "step": 17032
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0587723639969662,
+      "learning_rate": 9.249136158608163e-07,
+      "loss": 0.1896,
+      "step": 17033
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0309704538708153,
+      "learning_rate": 9.242219570858757e-07,
+      "loss": 0.1813,
+      "step": 17034
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0474496824022568,
+      "learning_rate": 9.235305444919307e-07,
+      "loss": 0.1718,
+      "step": 17035
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.23008560442467,
+      "learning_rate": 9.228393780977296e-07,
+      "loss": 0.1671,
+      "step": 17036
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9214761008903439,
+      "learning_rate": 9.22148457922023e-07,
+      "loss": 0.1703,
+      "step": 17037
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.447550516118549,
+      "learning_rate": 9.214577839835514e-07,
+      "loss": 0.144,
+      "step": 17038
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4592960258901386,
+      "learning_rate": 9.207673563010533e-07,
+      "loss": 0.187,
+      "step": 17039
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8769187476094875,
+      "learning_rate": 9.200771748932513e-07,
+      "loss": 0.1853,
+      "step": 17040
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9573474621913002,
+      "learning_rate": 9.193872397788705e-07,
+      "loss": 0.1656,
+      "step": 17041
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6841387334089557,
+      "learning_rate": 9.186975509766216e-07,
+      "loss": 0.1465,
+      "step": 17042
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.230747654561539,
+      "learning_rate": 9.180081085052162e-07,
+      "loss": 0.1594,
+      "step": 17043
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9144311106279304,
+      "learning_rate": 9.173189123833526e-07,
+      "loss": 0.1746,
+      "step": 17044
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.969731855311959,
+      "learning_rate": 9.166299626297271e-07,
+      "loss": 0.1699,
+      "step": 17045
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0739965557393911,
+      "learning_rate": 9.159412592630279e-07,
+      "loss": 0.1706,
+      "step": 17046
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5170650586505527,
+      "learning_rate": 9.152528023019325e-07,
+      "loss": 0.1634,
+      "step": 17047
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.441823463208156,
+      "learning_rate": 9.145645917651214e-07,
+      "loss": 0.152,
+      "step": 17048
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0354265673263898,
+      "learning_rate": 9.138766276712552e-07,
+      "loss": 0.1649,
+      "step": 17049
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1997505349833961,
+      "learning_rate": 9.131889100390024e-07,
+      "loss": 0.1737,
+      "step": 17050
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.196969721625916,
+      "learning_rate": 9.125014388870101e-07,
+      "loss": 0.1546,
+      "step": 17051
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9409125676079062,
+      "learning_rate": 9.118142142339326e-07,
+      "loss": 0.1916,
+      "step": 17052
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.046336542925199,
+      "learning_rate": 9.111272360984058e-07,
+      "loss": 0.158,
+      "step": 17053
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9049607362766768,
+      "learning_rate": 9.104405044990661e-07,
+      "loss": 0.1799,
+      "step": 17054
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0200458401151316,
+      "learning_rate": 9.097540194545407e-07,
+      "loss": 0.1589,
+      "step": 17055
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8076764602199117,
+      "learning_rate": 9.090677809834525e-07,
+      "loss": 0.1961,
+      "step": 17056
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0527309977219788,
+      "learning_rate": 9.083817891044133e-07,
+      "loss": 0.1601,
+      "step": 17057
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3595022523979385,
+      "learning_rate": 9.076960438360327e-07,
+      "loss": 0.1576,
+      "step": 17058
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9395096482761963,
+      "learning_rate": 9.070105451969091e-07,
+      "loss": 0.1624,
+      "step": 17059
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3353654985682613,
+      "learning_rate": 9.063252932056399e-07,
+      "loss": 0.1646,
+      "step": 17060
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1705992125708042,
+      "learning_rate": 9.056402878808102e-07,
+      "loss": 0.1655,
+      "step": 17061
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1283675058654659,
+      "learning_rate": 9.049555292409995e-07,
+      "loss": 0.1509,
+      "step": 17062
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8843788307703369,
+      "learning_rate": 9.042710173047841e-07,
+      "loss": 0.1508,
+      "step": 17063
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9171842968857437,
+      "learning_rate": 9.035867520907304e-07,
+      "loss": 0.1591,
+      "step": 17064
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9310906259002731,
+      "learning_rate": 9.029027336174023e-07,
+      "loss": 0.1681,
+      "step": 17065
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.831040289624738,
+      "learning_rate": 9.022189619033495e-07,
+      "loss": 0.1806,
+      "step": 17066
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3134835153634918,
+      "learning_rate": 9.015354369671237e-07,
+      "loss": 0.185,
+      "step": 17067
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1991897700260155,
+      "learning_rate": 9.0085215882726e-07,
+      "loss": 0.1491,
+      "step": 17068
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0090186658831273,
+      "learning_rate": 9.001691275022984e-07,
+      "loss": 0.1621,
+      "step": 17069
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8294614748653688,
+      "learning_rate": 8.994863430107603e-07,
+      "loss": 0.154,
+      "step": 17070
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.917475210040729,
+      "learning_rate": 8.988038053711701e-07,
+      "loss": 0.1641,
+      "step": 17071
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9202157068800895,
+      "learning_rate": 8.981215146020394e-07,
+      "loss": 0.1689,
+      "step": 17072
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1480664342916895,
+      "learning_rate": 8.974394707218792e-07,
+      "loss": 0.1692,
+      "step": 17073
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1053708115759986,
+      "learning_rate": 8.967576737491856e-07,
+      "loss": 0.1842,
+      "step": 17074
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9201851719480871,
+      "learning_rate": 8.960761237024573e-07,
+      "loss": 0.1747,
+      "step": 17075
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0454642154287848,
+      "learning_rate": 8.95394820600175e-07,
+      "loss": 0.1831,
+      "step": 17076
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0702369137636214,
+      "learning_rate": 8.94713764460825e-07,
+      "loss": 0.1695,
+      "step": 17077
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0091538146196728,
+      "learning_rate": 8.940329553028782e-07,
+      "loss": 0.1739,
+      "step": 17078
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0326055197662956,
+      "learning_rate": 8.933523931447996e-07,
+      "loss": 0.1647,
+      "step": 17079
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0176709480126727,
+      "learning_rate": 8.926720780050513e-07,
+      "loss": 0.1689,
+      "step": 17080
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0575101352035323,
+      "learning_rate": 8.919920099020884e-07,
+      "loss": 0.1482,
+      "step": 17081
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8582034731028793,
+      "learning_rate": 8.913121888543575e-07,
+      "loss": 0.1623,
+      "step": 17082
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.43270078497795,
+      "learning_rate": 8.906326148802968e-07,
+      "loss": 0.1651,
+      "step": 17083
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2905694891493917,
+      "learning_rate": 8.89953287998343e-07,
+      "loss": 0.1678,
+      "step": 17084
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.019744957485963,
+      "learning_rate": 8.892742082269179e-07,
+      "loss": 0.1683,
+      "step": 17085
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.270537102765076,
+      "learning_rate": 8.885953755844467e-07,
+      "loss": 0.1568,
+      "step": 17086
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0910139306492328,
+      "learning_rate": 8.879167900893392e-07,
+      "loss": 0.1664,
+      "step": 17087
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.833649447834106,
+      "learning_rate": 8.872384517600053e-07,
+      "loss": 0.1686,
+      "step": 17088
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9117543020684723,
+      "learning_rate": 8.86560360614841e-07,
+      "loss": 0.1638,
+      "step": 17089
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3114404092976746,
+      "learning_rate": 8.858825166722418e-07,
+      "loss": 0.1694,
+      "step": 17090
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9301874381300534,
+      "learning_rate": 8.852049199505941e-07,
+      "loss": 0.1502,
+      "step": 17091
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6722753021063932,
+      "learning_rate": 8.845275704682788e-07,
+      "loss": 0.1689,
+      "step": 17092
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0828515003759154,
+      "learning_rate": 8.838504682436666e-07,
+      "loss": 0.1768,
+      "step": 17093
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9844296891643936,
+      "learning_rate": 8.831736132951274e-07,
+      "loss": 0.1713,
+      "step": 17094
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9202759734089795,
+      "learning_rate": 8.824970056410187e-07,
+      "loss": 0.1538,
+      "step": 17095
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4026351905757806,
+      "learning_rate": 8.818206452996924e-07,
+      "loss": 0.1789,
+      "step": 17096
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.615643107469265,
+      "learning_rate": 8.811445322894951e-07,
+      "loss": 0.1541,
+      "step": 17097
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1891702976515879,
+      "learning_rate": 8.804686666287688e-07,
+      "loss": 0.1564,
+      "step": 17098
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0217262338262187,
+      "learning_rate": 8.797930483358452e-07,
+      "loss": 0.167,
+      "step": 17099
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0987630781058417,
+      "learning_rate": 8.7911767742905e-07,
+      "loss": 0.1531,
+      "step": 17100
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0437865886697673,
+      "learning_rate": 8.784425539267038e-07,
+      "loss": 0.1842,
+      "step": 17101
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7743462507231501,
+      "learning_rate": 8.777676778471167e-07,
+      "loss": 0.1587,
+      "step": 17102
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8877364528856684,
+      "learning_rate": 8.770930492085983e-07,
+      "loss": 0.1584,
+      "step": 17103
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7920300896197154,
+      "learning_rate": 8.764186680294451e-07,
+      "loss": 0.179,
+      "step": 17104
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3643034663684792,
+      "learning_rate": 8.757445343279514e-07,
+      "loss": 0.1764,
+      "step": 17105
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9771049397743856,
+      "learning_rate": 8.750706481224014e-07,
+      "loss": 0.1593,
+      "step": 17106
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1156830433209912,
+      "learning_rate": 8.743970094310761e-07,
+      "loss": 0.1466,
+      "step": 17107
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2084500367718818,
+      "learning_rate": 8.737236182722464e-07,
+      "loss": 0.1795,
+      "step": 17108
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0696994804947388,
+      "learning_rate": 8.730504746641811e-07,
+      "loss": 0.1802,
+      "step": 17109
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9620892356217626,
+      "learning_rate": 8.723775786251354e-07,
+      "loss": 0.1533,
+      "step": 17110
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9459466502861161,
+      "learning_rate": 8.71704930173366e-07,
+      "loss": 0.136,
+      "step": 17111
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4660240342842592,
+      "learning_rate": 8.710325293271126e-07,
+      "loss": 0.1786,
+      "step": 17112
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.228860617049589,
+      "learning_rate": 8.703603761046209e-07,
+      "loss": 0.1744,
+      "step": 17113
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.841572818613883,
+      "learning_rate": 8.696884705241182e-07,
+      "loss": 0.1544,
+      "step": 17114
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1666695309699784,
+      "learning_rate": 8.690168126038301e-07,
+      "loss": 0.1818,
+      "step": 17115
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3421305915982424,
+      "learning_rate": 8.683454023619775e-07,
+      "loss": 0.158,
+      "step": 17116
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.894492935229557,
+      "learning_rate": 8.676742398167704e-07,
+      "loss": 0.1709,
+      "step": 17117
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0467244786440408,
+      "learning_rate": 8.670033249864174e-07,
+      "loss": 0.1636,
+      "step": 17118
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2673527693727311,
+      "learning_rate": 8.66332657889114e-07,
+      "loss": 0.1778,
+      "step": 17119
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0575893042859574,
+      "learning_rate": 8.656622385430547e-07,
+      "loss": 0.1773,
+      "step": 17120
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1402664070621547,
+      "learning_rate": 8.649920669664202e-07,
+      "loss": 0.1673,
+      "step": 17121
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.012078317229577,
+      "learning_rate": 8.643221431773952e-07,
+      "loss": 0.1697,
+      "step": 17122
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9484744378294417,
+      "learning_rate": 8.636524671941449e-07,
+      "loss": 0.1586,
+      "step": 17123
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9925048517257558,
+      "learning_rate": 8.629830390348382e-07,
+      "loss": 0.1731,
+      "step": 17124
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9828673798522906,
+      "learning_rate": 8.623138587176327e-07,
+      "loss": 0.1712,
+      "step": 17125
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7792158808509042,
+      "learning_rate": 8.616449262606819e-07,
+      "loss": 0.1464,
+      "step": 17126
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0968797454498505,
+      "learning_rate": 8.609762416821255e-07,
+      "loss": 0.171,
+      "step": 17127
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9891852371438813,
+      "learning_rate": 8.603078050001079e-07,
+      "loss": 0.1573,
+      "step": 17128
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.21659724600745,
+      "learning_rate": 8.596396162327547e-07,
+      "loss": 0.1641,
+      "step": 17129
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.141819368951561,
+      "learning_rate": 8.589716753981958e-07,
+      "loss": 0.1607,
+      "step": 17130
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.566854918335254,
+      "learning_rate": 8.583039825145456e-07,
+      "loss": 0.1803,
+      "step": 17131
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0486597977148742,
+      "learning_rate": 8.576365375999151e-07,
+      "loss": 0.174,
+      "step": 17132
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8382358364538444,
+      "learning_rate": 8.569693406724089e-07,
+      "loss": 0.1534,
+      "step": 17133
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9025809558840777,
+      "learning_rate": 8.563023917501267e-07,
+      "loss": 0.1739,
+      "step": 17134
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0805583000627847,
+      "learning_rate": 8.556356908511598e-07,
+      "loss": 0.173,
+      "step": 17135
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0793640614355466,
+      "learning_rate": 8.549692379935904e-07,
+      "loss": 0.1488,
+      "step": 17136
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.873788082427515,
+      "learning_rate": 8.543030331954971e-07,
+      "loss": 0.1416,
+      "step": 17137
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5731957662613225,
+      "learning_rate": 8.536370764749502e-07,
+      "loss": 0.1654,
+      "step": 17138
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.055450540196962,
+      "learning_rate": 8.529713678500151e-07,
+      "loss": 0.1739,
+      "step": 17139
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9008341102955426,
+      "learning_rate": 8.523059073387474e-07,
+      "loss": 0.1509,
+      "step": 17140
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8612485946862515,
+      "learning_rate": 8.516406949591982e-07,
+      "loss": 0.1641,
+      "step": 17141
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0342607155695587,
+      "learning_rate": 8.509757307294109e-07,
+      "loss": 0.173,
+      "step": 17142
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9384308186441249,
+      "learning_rate": 8.503110146674265e-07,
+      "loss": 0.1612,
+      "step": 17143
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1232769793082433,
+      "learning_rate": 8.496465467912707e-07,
+      "loss": 0.1474,
+      "step": 17144
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.414174825300634,
+      "learning_rate": 8.489823271189712e-07,
+      "loss": 0.1608,
+      "step": 17145
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.43889602224765,
+      "learning_rate": 8.483183556685404e-07,
+      "loss": 0.1896,
+      "step": 17146
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.271727693081388,
+      "learning_rate": 8.476546324579937e-07,
+      "loss": 0.165,
+      "step": 17147
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3473452346504362,
+      "learning_rate": 8.469911575053314e-07,
+      "loss": 0.1632,
+      "step": 17148
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9080943240602966,
+      "learning_rate": 8.463279308285488e-07,
+      "loss": 0.1635,
+      "step": 17149
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.002355761226064,
+      "learning_rate": 8.456649524456384e-07,
+      "loss": 0.1702,
+      "step": 17150
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1178232175726797,
+      "learning_rate": 8.450022223745836e-07,
+      "loss": 0.1625,
+      "step": 17151
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9855127077410061,
+      "learning_rate": 8.44339740633362e-07,
+      "loss": 0.1483,
+      "step": 17152
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2085619193638986,
+      "learning_rate": 8.436775072399406e-07,
+      "loss": 0.1676,
+      "step": 17153
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.106924279440494,
+      "learning_rate": 8.43015522212286e-07,
+      "loss": 0.162,
+      "step": 17154
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9675200693427021,
+      "learning_rate": 8.423537855683494e-07,
+      "loss": 0.1629,
+      "step": 17155
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.046538127580827,
+      "learning_rate": 8.416922973260865e-07,
+      "loss": 0.1471,
+      "step": 17156
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9846894068376528,
+      "learning_rate": 8.410310575034353e-07,
+      "loss": 0.1523,
+      "step": 17157
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4189671040115495,
+      "learning_rate": 8.403700661183356e-07,
+      "loss": 0.1633,
+      "step": 17158
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1215961139129562,
+      "learning_rate": 8.397093231887143e-07,
+      "loss": 0.1657,
+      "step": 17159
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9463614162161704,
+      "learning_rate": 8.390488287324938e-07,
+      "loss": 0.1509,
+      "step": 17160
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7667384852227885,
+      "learning_rate": 8.383885827675919e-07,
+      "loss": 0.1589,
+      "step": 17161
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.957756528157922,
+      "learning_rate": 8.377285853119188e-07,
+      "loss": 0.1726,
+      "step": 17162
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8567222388031468,
+      "learning_rate": 8.370688363833734e-07,
+      "loss": 0.1536,
+      "step": 17163
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5066683002465677,
+      "learning_rate": 8.364093359998549e-07,
+      "loss": 0.1671,
+      "step": 17164
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4478565414467472,
+      "learning_rate": 8.35750084179251e-07,
+      "loss": 0.1595,
+      "step": 17165
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9872017321212999,
+      "learning_rate": 8.350910809394419e-07,
+      "loss": 0.1471,
+      "step": 17166
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.831180223310793,
+      "learning_rate": 8.344323262983056e-07,
+      "loss": 0.1693,
+      "step": 17167
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1506089263597188,
+      "learning_rate": 8.337738202737089e-07,
+      "loss": 0.1629,
+      "step": 17168
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5383110955910784,
+      "learning_rate": 8.331155628835174e-07,
+      "loss": 0.1663,
+      "step": 17169
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0783555363988484,
+      "learning_rate": 8.324575541455815e-07,
+      "loss": 0.1681,
+      "step": 17170
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.396062909846658,
+      "learning_rate": 8.317997940777555e-07,
+      "loss": 0.1819,
+      "step": 17171
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8100993706187744,
+      "learning_rate": 8.311422826978743e-07,
+      "loss": 0.167,
+      "step": 17172
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1519109872457536,
+      "learning_rate": 8.304850200237801e-07,
+      "loss": 0.1864,
+      "step": 17173
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1145067988423865,
+      "learning_rate": 8.298280060732944e-07,
+      "loss": 0.1596,
+      "step": 17174
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9373264532460865,
+      "learning_rate": 8.29171240864245e-07,
+      "loss": 0.1651,
+      "step": 17175
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.015744660484614,
+      "learning_rate": 8.285147244144409e-07,
+      "loss": 0.1858,
+      "step": 17176
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2143262440929645,
+      "learning_rate": 8.278584567416936e-07,
+      "loss": 0.1677,
+      "step": 17177
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8214701062742716,
+      "learning_rate": 8.272024378638033e-07,
+      "loss": 0.1551,
+      "step": 17178
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.089253746380905,
+      "learning_rate": 8.265466677985667e-07,
+      "loss": 0.1829,
+      "step": 17179
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4385101425570883,
+      "learning_rate": 8.258911465637675e-07,
+      "loss": 0.1705,
+      "step": 17180
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9000616820305697,
+      "learning_rate": 8.252358741771915e-07,
+      "loss": 0.1539,
+      "step": 17181
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.435773434610318,
+      "learning_rate": 8.245808506566088e-07,
+      "loss": 0.1807,
+      "step": 17182
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7531987585070594,
+      "learning_rate": 8.239260760197909e-07,
+      "loss": 0.1493,
+      "step": 17183
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9821965062763874,
+      "learning_rate": 8.232715502844968e-07,
+      "loss": 0.1592,
+      "step": 17184
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.918881429900906,
+      "learning_rate": 8.226172734684779e-07,
+      "loss": 0.1879,
+      "step": 17185
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2048594743894412,
+      "learning_rate": 8.219632455894833e-07,
+      "loss": 0.1753,
+      "step": 17186
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7541521591860647,
+      "learning_rate": 8.213094666652544e-07,
+      "loss": 0.1596,
+      "step": 17187
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.348671888935361,
+      "learning_rate": 8.206559367135258e-07,
+      "loss": 0.1725,
+      "step": 17188
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2557319159842633,
+      "learning_rate": 8.200026557520224e-07,
+      "loss": 0.1568,
+      "step": 17189
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1154418262195018,
+      "learning_rate": 8.193496237984677e-07,
+      "loss": 0.1638,
+      "step": 17190
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9471020486573046,
+      "learning_rate": 8.186968408705697e-07,
+      "loss": 0.1498,
+      "step": 17191
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8734603646022715,
+      "learning_rate": 8.18044306986041e-07,
+      "loss": 0.1548,
+      "step": 17192
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0798723064063107,
+      "learning_rate": 8.173920221625776e-07,
+      "loss": 0.1455,
+      "step": 17193
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8520548621086595,
+      "learning_rate": 8.167399864178749e-07,
+      "loss": 0.1777,
+      "step": 17194
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8022428202826044,
+      "learning_rate": 8.160881997696169e-07,
+      "loss": 0.1615,
+      "step": 17195
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5703257428569193,
+      "learning_rate": 8.154366622354881e-07,
+      "loss": 0.1734,
+      "step": 17196
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8567723378618287,
+      "learning_rate": 8.147853738331569e-07,
+      "loss": 0.1329,
+      "step": 17197
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.858800319683022,
+      "learning_rate": 8.141343345802933e-07,
+      "loss": 0.1812,
+      "step": 17198
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9505104143578171,
+      "learning_rate": 8.134835444945521e-07,
+      "loss": 0.1556,
+      "step": 17199
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8372452692250099,
+      "learning_rate": 8.128330035935906e-07,
+      "loss": 0.148,
+      "step": 17200
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9504940848887924,
+      "learning_rate": 8.121827118950521e-07,
+      "loss": 0.1598,
+      "step": 17201
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.674017728736617,
+      "learning_rate": 8.115326694165759e-07,
+      "loss": 0.1491,
+      "step": 17202
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0173530063273255,
+      "learning_rate": 8.108828761757948e-07,
+      "loss": 0.1596,
+      "step": 17203
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7623952342628695,
+      "learning_rate": 8.102333321903344e-07,
+      "loss": 0.1666,
+      "step": 17204
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.194863581749474,
+      "learning_rate": 8.095840374778153e-07,
+      "loss": 0.1646,
+      "step": 17205
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9392827691446882,
+      "learning_rate": 8.089349920558465e-07,
+      "loss": 0.1578,
+      "step": 17206
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0778890805293149,
+      "learning_rate": 8.082861959420374e-07,
+      "loss": 0.17,
+      "step": 17207
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6435653004467239,
+      "learning_rate": 8.076376491539827e-07,
+      "loss": 0.1663,
+      "step": 17208
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2995772904802547,
+      "learning_rate": 8.069893517092775e-07,
+      "loss": 0.1644,
+      "step": 17209
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3175793197088113,
+      "learning_rate": 8.063413036255041e-07,
+      "loss": 0.1531,
+      "step": 17210
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0686728260509748,
+      "learning_rate": 8.05693504920243e-07,
+      "loss": 0.1778,
+      "step": 17211
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7564258312190575,
+      "learning_rate": 8.050459556110635e-07,
+      "loss": 0.1698,
+      "step": 17212
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2441905694267967,
+      "learning_rate": 8.043986557155315e-07,
+      "loss": 0.1581,
+      "step": 17213
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0520955464046744,
+      "learning_rate": 8.037516052512062e-07,
+      "loss": 0.1643,
+      "step": 17214
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3677567086585491,
+      "learning_rate": 8.031048042356393e-07,
+      "loss": 0.1551,
+      "step": 17215
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2042220749762846,
+      "learning_rate": 8.024582526863722e-07,
+      "loss": 0.1711,
+      "step": 17216
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0138279779706278,
+      "learning_rate": 8.018119506209454e-07,
+      "loss": 0.1576,
+      "step": 17217
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2803233628460835,
+      "learning_rate": 8.011658980568903e-07,
+      "loss": 0.1636,
+      "step": 17218
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8006138395157962,
+      "learning_rate": 8.005200950117275e-07,
+      "loss": 0.1514,
+      "step": 17219
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0802690996728705,
+      "learning_rate": 7.998745415029762e-07,
+      "loss": 0.1676,
+      "step": 17220
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.227853461420169,
+      "learning_rate": 7.99229237548148e-07,
+      "loss": 0.1644,
+      "step": 17221
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0425876802294116,
+      "learning_rate": 7.985841831647489e-07,
+      "loss": 0.168,
+      "step": 17222
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0061004327128427,
+      "learning_rate": 7.979393783702704e-07,
+      "loss": 0.1621,
+      "step": 17223
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6922528782775406,
+      "learning_rate": 7.972948231822087e-07,
+      "loss": 0.1624,
+      "step": 17224
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9172038906437257,
+      "learning_rate": 7.966505176180428e-07,
+      "loss": 0.1756,
+      "step": 17225
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3744976825971134,
+      "learning_rate": 7.960064616952523e-07,
+      "loss": 0.1641,
+      "step": 17226
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.268173030375666,
+      "learning_rate": 7.953626554313055e-07,
+      "loss": 0.1457,
+      "step": 17227
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4362620959002659,
+      "learning_rate": 7.947190988436681e-07,
+      "loss": 0.1489,
+      "step": 17228
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1619879743795989,
+      "learning_rate": 7.940757919497944e-07,
+      "loss": 0.1814,
+      "step": 17229
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9927394194170999,
+      "learning_rate": 7.934327347671333e-07,
+      "loss": 0.1714,
+      "step": 17230
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9038842448831277,
+      "learning_rate": 7.927899273131301e-07,
+      "loss": 0.1454,
+      "step": 17231
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3595367981253321,
+      "learning_rate": 7.921473696052206e-07,
+      "loss": 0.1528,
+      "step": 17232
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0757902365280851,
+      "learning_rate": 7.915050616608333e-07,
+      "loss": 0.1663,
+      "step": 17233
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2885867737737462,
+      "learning_rate": 7.90863003497393e-07,
+      "loss": 0.1816,
+      "step": 17234
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.374042545612596,
+      "learning_rate": 7.902211951323135e-07,
+      "loss": 0.1461,
+      "step": 17235
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2403023423821855,
+      "learning_rate": 7.895796365830021e-07,
+      "loss": 0.1737,
+      "step": 17236
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1516657453656927,
+      "learning_rate": 7.889383278668661e-07,
+      "loss": 0.1486,
+      "step": 17237
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3981759897934891,
+      "learning_rate": 7.882972690012957e-07,
+      "loss": 0.1755,
+      "step": 17238
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3367001388773048,
+      "learning_rate": 7.876564600036818e-07,
+      "loss": 0.1781,
+      "step": 17239
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9591334225113681,
+      "learning_rate": 7.870159008914069e-07,
+      "loss": 0.163,
+      "step": 17240
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1275579944257983,
+      "learning_rate": 7.863755916818483e-07,
+      "loss": 0.1838,
+      "step": 17241
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4577251812287217,
+      "learning_rate": 7.85735532392371e-07,
+      "loss": 0.164,
+      "step": 17242
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0104330875157992,
+      "learning_rate": 7.850957230403378e-07,
+      "loss": 0.1717,
+      "step": 17243
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.104711385825246,
+      "learning_rate": 7.844561636431036e-07,
+      "loss": 0.1879,
+      "step": 17244
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5995755578306345,
+      "learning_rate": 7.838168542180169e-07,
+      "loss": 0.1398,
+      "step": 17245
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1064120870453258,
+      "learning_rate": 7.83177794782417e-07,
+      "loss": 0.1694,
+      "step": 17246
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9905646769100281,
+      "learning_rate": 7.825389853536403e-07,
+      "loss": 0.1724,
+      "step": 17247
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0883238839154536,
+      "learning_rate": 7.819004259490148e-07,
+      "loss": 0.1671,
+      "step": 17248
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.589772887880888,
+      "learning_rate": 7.812621165858625e-07,
+      "loss": 0.1989,
+      "step": 17249
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9767483701679833,
+      "learning_rate": 7.806240572814927e-07,
+      "loss": 0.1762,
+      "step": 17250
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9107924632802049,
+      "learning_rate": 7.799862480532194e-07,
+      "loss": 0.1513,
+      "step": 17251
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9780873235784965,
+      "learning_rate": 7.793486889183377e-07,
+      "loss": 0.1605,
+      "step": 17252
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2925286403963623,
+      "learning_rate": 7.787113798941449e-07,
+      "loss": 0.1701,
+      "step": 17253
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2855320737032634,
+      "learning_rate": 7.780743209979269e-07,
+      "loss": 0.1574,
+      "step": 17254
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0986164566088186,
+      "learning_rate": 7.774375122469624e-07,
+      "loss": 0.1723,
+      "step": 17255
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3614995748491658,
+      "learning_rate": 7.768009536585264e-07,
+      "loss": 0.1501,
+      "step": 17256
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6770303975905876,
+      "learning_rate": 7.76164645249885e-07,
+      "loss": 0.1551,
+      "step": 17257
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.076382257356652,
+      "learning_rate": 7.755285870383011e-07,
+      "loss": 0.165,
+      "step": 17258
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5438451565334113,
+      "learning_rate": 7.748927790410221e-07,
+      "loss": 0.1292,
+      "step": 17259
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2707148714608554,
+      "learning_rate": 7.742572212753008e-07,
+      "loss": 0.1611,
+      "step": 17260
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1250090687287368,
+      "learning_rate": 7.736219137583701e-07,
+      "loss": 0.1839,
+      "step": 17261
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9268465624245334,
+      "learning_rate": 7.729868565074694e-07,
+      "loss": 0.1683,
+      "step": 17262
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8984489920500216,
+      "learning_rate": 7.723520495398185e-07,
+      "loss": 0.1573,
+      "step": 17263
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2633143592884477,
+      "learning_rate": 7.717174928726401e-07,
+      "loss": 0.1511,
+      "step": 17264
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4414316188321858,
+      "learning_rate": 7.710831865231461e-07,
+      "loss": 0.1825,
+      "step": 17265
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0270181484599707,
+      "learning_rate": 7.704491305085427e-07,
+      "loss": 0.193,
+      "step": 17266
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7770444000137979,
+      "learning_rate": 7.698153248460271e-07,
+      "loss": 0.1773,
+      "step": 17267
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3377018839713426,
+      "learning_rate": 7.691817695527936e-07,
+      "loss": 0.1898,
+      "step": 17268
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2142810653464722,
+      "learning_rate": 7.68548464646024e-07,
+      "loss": 0.1691,
+      "step": 17269
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.133137337818727,
+      "learning_rate": 7.679154101428998e-07,
+      "loss": 0.1435,
+      "step": 17270
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0938459587105016,
+      "learning_rate": 7.672826060605931e-07,
+      "loss": 0.1618,
+      "step": 17271
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.30171994693312,
+      "learning_rate": 7.666500524162646e-07,
+      "loss": 0.1507,
+      "step": 17272
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2058208735977238,
+      "learning_rate": 7.660177492270749e-07,
+      "loss": 0.1453,
+      "step": 17273
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1198741455359467,
+      "learning_rate": 7.653856965101747e-07,
+      "loss": 0.1589,
+      "step": 17274
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8746745470939669,
+      "learning_rate": 7.647538942827115e-07,
+      "loss": 0.145,
+      "step": 17275
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2037543741940957,
+      "learning_rate": 7.641223425618193e-07,
+      "loss": 0.1648,
+      "step": 17276
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8927725992603462,
+      "learning_rate": 7.634910413646313e-07,
+      "loss": 0.1376,
+      "step": 17277
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3217628032750142,
+      "learning_rate": 7.62859990708269e-07,
+      "loss": 0.1622,
+      "step": 17278
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1443610497786343,
+      "learning_rate": 7.622291906098523e-07,
+      "loss": 0.156,
+      "step": 17279
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2311048614980642,
+      "learning_rate": 7.615986410864895e-07,
+      "loss": 0.1614,
+      "step": 17280
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2165201760489044,
+      "learning_rate": 7.609683421552861e-07,
+      "loss": 0.1638,
+      "step": 17281
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5208694414991424,
+      "learning_rate": 7.603382938333382e-07,
+      "loss": 0.1596,
+      "step": 17282
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2366002878916877,
+      "learning_rate": 7.597084961377343e-07,
+      "loss": 0.1453,
+      "step": 17283
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.956702883190355,
+      "learning_rate": 7.590789490855599e-07,
+      "loss": 0.1609,
+      "step": 17284
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7430471552680176,
+      "learning_rate": 7.584496526938933e-07,
+      "loss": 0.1864,
+      "step": 17285
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1854800332218853,
+      "learning_rate": 7.578206069797989e-07,
+      "loss": 0.1511,
+      "step": 17286
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5933026546514464,
+      "learning_rate": 7.57191811960345e-07,
+      "loss": 0.1418,
+      "step": 17287
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.140500896412719,
+      "learning_rate": 7.565632676525858e-07,
+      "loss": 0.1621,
+      "step": 17288
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2614036948724896,
+      "learning_rate": 7.559349740735677e-07,
+      "loss": 0.1618,
+      "step": 17289
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0229832009884556,
+      "learning_rate": 7.55306931240335e-07,
+      "loss": 0.1566,
+      "step": 17290
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1404776623724786,
+      "learning_rate": 7.546791391699248e-07,
+      "loss": 0.151,
+      "step": 17291
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.027669539432624,
+      "learning_rate": 7.540515978793661e-07,
+      "loss": 0.154,
+      "step": 17292
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9730493748099621,
+      "learning_rate": 7.534243073856784e-07,
+      "loss": 0.1629,
+      "step": 17293
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9318541669733853,
+      "learning_rate": 7.527972677058814e-07,
+      "loss": 0.1712,
+      "step": 17294
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.128684780041153,
+      "learning_rate": 7.521704788569783e-07,
+      "loss": 0.1829,
+      "step": 17295
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0252096353600924,
+      "learning_rate": 7.515439408559744e-07,
+      "loss": 0.1749,
+      "step": 17296
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.995727140664652,
+      "learning_rate": 7.509176537198626e-07,
+      "loss": 0.1692,
+      "step": 17297
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.969441578451583,
+      "learning_rate": 7.502916174656338e-07,
+      "loss": 0.1593,
+      "step": 17298
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2329684303060555,
+      "learning_rate": 7.496658321102646e-07,
+      "loss": 0.1613,
+      "step": 17299
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9423363240103179,
+      "learning_rate": 7.490402976707323e-07,
+      "loss": 0.1572,
+      "step": 17300
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2564010048538554,
+      "learning_rate": 7.484150141640056e-07,
+      "loss": 0.173,
+      "step": 17301
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.070435552031945,
+      "learning_rate": 7.477899816070444e-07,
+      "loss": 0.1659,
+      "step": 17302
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5672121229787728,
+      "learning_rate": 7.471652000168017e-07,
+      "loss": 0.1677,
+      "step": 17303
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0885715864729035,
+      "learning_rate": 7.465406694102273e-07,
+      "loss": 0.1572,
+      "step": 17304
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.436480958250531,
+      "learning_rate": 7.459163898042599e-07,
+      "loss": 0.1626,
+      "step": 17305
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.405159042399723,
+      "learning_rate": 7.452923612158303e-07,
+      "loss": 0.1751,
+      "step": 17306
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9347108488017326,
+      "learning_rate": 7.446685836618706e-07,
+      "loss": 0.1812,
+      "step": 17307
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.16810165610296,
+      "learning_rate": 7.440450571592972e-07,
+      "loss": 0.1463,
+      "step": 17308
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.045784749400454,
+      "learning_rate": 7.434217817250233e-07,
+      "loss": 0.1556,
+      "step": 17309
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2025574622438935,
+      "learning_rate": 7.427987573759576e-07,
+      "loss": 0.1656,
+      "step": 17310
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0111063760692782,
+      "learning_rate": 7.421759841289989e-07,
+      "loss": 0.1608,
+      "step": 17311
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.522813352090518,
+      "learning_rate": 7.41553462001039e-07,
+      "loss": 0.1501,
+      "step": 17312
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2803206073977835,
+      "learning_rate": 7.409311910089645e-07,
+      "loss": 0.1971,
+      "step": 17313
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8767646651286921,
+      "learning_rate": 7.403091711696542e-07,
+      "loss": 0.1629,
+      "step": 17314
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2290321284414285,
+      "learning_rate": 7.396874024999811e-07,
+      "loss": 0.1695,
+      "step": 17315
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4577615994194053,
+      "learning_rate": 7.390658850168098e-07,
+      "loss": 0.1624,
+      "step": 17316
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.235987193669687,
+      "learning_rate": 7.384446187369987e-07,
+      "loss": 0.1489,
+      "step": 17317
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7957026665132977,
+      "learning_rate": 7.378236036774e-07,
+      "loss": 0.1522,
+      "step": 17318
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8319149713315096,
+      "learning_rate": 7.372028398548614e-07,
+      "loss": 0.137,
+      "step": 17319
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2876008102788261,
+      "learning_rate": 7.365823272862183e-07,
+      "loss": 0.1818,
+      "step": 17320
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1938219095258311,
+      "learning_rate": 7.359620659883026e-07,
+      "loss": 0.1785,
+      "step": 17321
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1088942410264782,
+      "learning_rate": 7.35342055977939e-07,
+      "loss": 0.1956,
+      "step": 17322
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.463606545831211,
+      "learning_rate": 7.347222972719459e-07,
+      "loss": 0.1475,
+      "step": 17323
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9638203971133041,
+      "learning_rate": 7.341027898871345e-07,
+      "loss": 0.1626,
+      "step": 17324
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7616047380436473,
+      "learning_rate": 7.334835338403056e-07,
+      "loss": 0.1497,
+      "step": 17325
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1910719387044761,
+      "learning_rate": 7.328645291482606e-07,
+      "loss": 0.1651,
+      "step": 17326
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.595718060050182,
+      "learning_rate": 7.322457758277879e-07,
+      "loss": 0.1504,
+      "step": 17327
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7702924587303707,
+      "learning_rate": 7.316272738956731e-07,
+      "loss": 0.1508,
+      "step": 17328
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0275799989309966,
+      "learning_rate": 7.310090233686917e-07,
+      "loss": 0.1594,
+      "step": 17329
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9451932165357124,
+      "learning_rate": 7.303910242636147e-07,
+      "loss": 0.1729,
+      "step": 17330
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8652856570187278,
+      "learning_rate": 7.297732765972033e-07,
+      "loss": 0.158,
+      "step": 17331
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7772120827819856,
+      "learning_rate": 7.29155780386217e-07,
+      "loss": 0.1576,
+      "step": 17332
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8908650335082182,
+      "learning_rate": 7.285385356474017e-07,
+      "loss": 0.1343,
+      "step": 17333
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.237847844903063,
+      "learning_rate": 7.27921542397505e-07,
+      "loss": 0.1735,
+      "step": 17334
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1559617010008287,
+      "learning_rate": 7.273048006532569e-07,
+      "loss": 0.1639,
+      "step": 17335
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1807040984900672,
+      "learning_rate": 7.266883104313916e-07,
+      "loss": 0.1661,
+      "step": 17336
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9171491441040258,
+      "learning_rate": 7.260720717486281e-07,
+      "loss": 0.1743,
+      "step": 17337
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9859557166131068,
+      "learning_rate": 7.254560846216863e-07,
+      "loss": 0.1809,
+      "step": 17338
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1130369511140106,
+      "learning_rate": 7.248403490672695e-07,
+      "loss": 0.1541,
+      "step": 17339
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8683717597560087,
+      "learning_rate": 7.242248651020845e-07,
+      "loss": 0.1517,
+      "step": 17340
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8444625384194985,
+      "learning_rate": 7.236096327428233e-07,
+      "loss": 0.1519,
+      "step": 17341
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2794386141678484,
+      "learning_rate": 7.229946520061737e-07,
+      "loss": 0.1743,
+      "step": 17342
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0310179364533691,
+      "learning_rate": 7.223799229088179e-07,
+      "loss": 0.1449,
+      "step": 17343
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7725121361002629,
+      "learning_rate": 7.217654454674305e-07,
+      "loss": 0.1587,
+      "step": 17344
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5382257187815973,
+      "learning_rate": 7.211512196986803e-07,
+      "loss": 0.1663,
+      "step": 17345
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0917289681842852,
+      "learning_rate": 7.205372456192272e-07,
+      "loss": 0.164,
+      "step": 17346
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9496468908631138,
+      "learning_rate": 7.199235232457258e-07,
+      "loss": 0.1825,
+      "step": 17347
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8778324978603581,
+      "learning_rate": 7.193100525948227e-07,
+      "loss": 0.1506,
+      "step": 17348
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.910535826795464,
+      "learning_rate": 7.18696833683159e-07,
+      "loss": 0.1748,
+      "step": 17349
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.096729609227987,
+      "learning_rate": 7.18083866527367e-07,
+      "loss": 0.1566,
+      "step": 17350
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8855534457008581,
+      "learning_rate": 7.174711511440757e-07,
+      "loss": 0.1595,
+      "step": 17351
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1438526198987724,
+      "learning_rate": 7.168586875499018e-07,
+      "loss": 0.1628,
+      "step": 17352
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9813627770034073,
+      "learning_rate": 7.162464757614606e-07,
+      "loss": 0.1505,
+      "step": 17353
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8963296520021135,
+      "learning_rate": 7.156345157953581e-07,
+      "loss": 0.1593,
+      "step": 17354
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0822470726148679,
+      "learning_rate": 7.150228076681954e-07,
+      "loss": 0.128,
+      "step": 17355
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0598933426021497,
+      "learning_rate": 7.144113513965623e-07,
+      "loss": 0.1831,
+      "step": 17356
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.100669695491159,
+      "learning_rate": 7.138001469970468e-07,
+      "loss": 0.1648,
+      "step": 17357
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.753646196389935,
+      "learning_rate": 7.131891944862269e-07,
+      "loss": 0.1572,
+      "step": 17358
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.060548591987011,
+      "learning_rate": 7.125784938806723e-07,
+      "loss": 0.1664,
+      "step": 17359
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2714760170007762,
+      "learning_rate": 7.119680451969524e-07,
+      "loss": 0.1563,
+      "step": 17360
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.391024615858319,
+      "learning_rate": 7.113578484516226e-07,
+      "loss": 0.1485,
+      "step": 17361
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0065629449928741,
+      "learning_rate": 7.107479036612375e-07,
+      "loss": 0.1482,
+      "step": 17362
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9222886347377515,
+      "learning_rate": 7.101382108423383e-07,
+      "loss": 0.1642,
+      "step": 17363
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9073932362794302,
+      "learning_rate": 7.095287700114673e-07,
+      "loss": 0.1695,
+      "step": 17364
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9807012578354181,
+      "learning_rate": 7.089195811851502e-07,
+      "loss": 0.1545,
+      "step": 17365
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1494039005467738,
+      "learning_rate": 7.083106443799171e-07,
+      "loss": 0.1607,
+      "step": 17366
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9978285693974382,
+      "learning_rate": 7.077019596122802e-07,
+      "loss": 0.1654,
+      "step": 17367
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.099167488039827,
+      "learning_rate": 7.070935268987545e-07,
+      "loss": 0.1693,
+      "step": 17368
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8713895734861333,
+      "learning_rate": 7.064853462558397e-07,
+      "loss": 0.1564,
+      "step": 17369
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1131621662933266,
+      "learning_rate": 7.05877417700035e-07,
+      "loss": 0.1686,
+      "step": 17370
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0392605432584487,
+      "learning_rate": 7.052697412478304e-07,
+      "loss": 0.161,
+      "step": 17371
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9436772944917501,
+      "learning_rate": 7.046623169157107e-07,
+      "loss": 0.1495,
+      "step": 17372
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9696917040806536,
+      "learning_rate": 7.040551447201494e-07,
+      "loss": 0.1708,
+      "step": 17373
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9909409102170923,
+      "learning_rate": 7.034482246776187e-07,
+      "loss": 0.1651,
+      "step": 17374
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8782964173375589,
+      "learning_rate": 7.028415568045799e-07,
+      "loss": 0.1585,
+      "step": 17375
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4118814762288865,
+      "learning_rate": 7.022351411174866e-07,
+      "loss": 0.1663,
+      "step": 17376
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0070090083368075,
+      "learning_rate": 7.016289776327922e-07,
+      "loss": 0.1451,
+      "step": 17377
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.237220825335278,
+      "learning_rate": 7.010230663669359e-07,
+      "loss": 0.1594,
+      "step": 17378
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4767425759880979,
+      "learning_rate": 7.004174073363546e-07,
+      "loss": 0.1521,
+      "step": 17379
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.376161358185498,
+      "learning_rate": 6.998120005574749e-07,
+      "loss": 0.1733,
+      "step": 17380
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2946303494952718,
+      "learning_rate": 6.992068460467227e-07,
+      "loss": 0.1671,
+      "step": 17381
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9239760508107825,
+      "learning_rate": 6.986019438205082e-07,
+      "loss": 0.1603,
+      "step": 17382
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7276860307771067,
+      "learning_rate": 6.979972938952428e-07,
+      "loss": 0.1722,
+      "step": 17383
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2215560557678822,
+      "learning_rate": 6.973928962873244e-07,
+      "loss": 0.1796,
+      "step": 17384
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2635692699347436,
+      "learning_rate": 6.96788751013151e-07,
+      "loss": 0.1584,
+      "step": 17385
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8953773561330749,
+      "learning_rate": 6.961848580891062e-07,
+      "loss": 0.1671,
+      "step": 17386
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9207265981103875,
+      "learning_rate": 6.955812175315735e-07,
+      "loss": 0.1659,
+      "step": 17387
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3557730967996362,
+      "learning_rate": 6.949778293569253e-07,
+      "loss": 0.1504,
+      "step": 17388
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4123195119500518,
+      "learning_rate": 6.943746935815299e-07,
+      "loss": 0.1495,
+      "step": 17389
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3687925436747972,
+      "learning_rate": 6.937718102217461e-07,
+      "loss": 0.1978,
+      "step": 17390
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2313131796932406,
+      "learning_rate": 6.931691792939288e-07,
+      "loss": 0.1657,
+      "step": 17391
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9197371602670851,
+      "learning_rate": 6.925668008144204e-07,
+      "loss": 0.1782,
+      "step": 17392
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8727111408694579,
+      "learning_rate": 6.919646747995668e-07,
+      "loss": 0.1529,
+      "step": 17393
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8399857072424712,
+      "learning_rate": 6.913628012656959e-07,
+      "loss": 0.1724,
+      "step": 17394
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2380908736083993,
+      "learning_rate": 6.907611802291325e-07,
+      "loss": 0.154,
+      "step": 17395
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0970936765726857,
+      "learning_rate": 6.901598117061992e-07,
+      "loss": 0.1483,
+      "step": 17396
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0378284593686915,
+      "learning_rate": 6.895586957132061e-07,
+      "loss": 0.1594,
+      "step": 17397
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0045317486233272,
+      "learning_rate": 6.889578322664614e-07,
+      "loss": 0.1703,
+      "step": 17398
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.096540638874496,
+      "learning_rate": 6.883572213822598e-07,
+      "loss": 0.1555,
+      "step": 17399
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6452198223024794,
+      "learning_rate": 6.87756863076896e-07,
+      "loss": 0.1605,
+      "step": 17400
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1708595471622938,
+      "learning_rate": 6.871567573666516e-07,
+      "loss": 0.1745,
+      "step": 17401
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9364607771802455,
+      "learning_rate": 6.865569042678066e-07,
+      "loss": 0.1622,
+      "step": 17402
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.069816199579359,
+      "learning_rate": 6.859573037966316e-07,
+      "loss": 0.153,
+      "step": 17403
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.110928969063031,
+      "learning_rate": 6.853579559693913e-07,
+      "loss": 0.1694,
+      "step": 17404
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.982718671441663,
+      "learning_rate": 6.847588608023414e-07,
+      "loss": 0.1646,
+      "step": 17405
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9125047628380294,
+      "learning_rate": 6.841600183117336e-07,
+      "loss": 0.1578,
+      "step": 17406
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.941456937741606,
+      "learning_rate": 6.835614285138115e-07,
+      "loss": 0.1549,
+      "step": 17407
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5330577556007376,
+      "learning_rate": 6.829630914248131e-07,
+      "loss": 0.1576,
+      "step": 17408
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6233044430724675,
+      "learning_rate": 6.823650070609666e-07,
+      "loss": 0.1449,
+      "step": 17409
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2905297846245838,
+      "learning_rate": 6.817671754384958e-07,
+      "loss": 0.167,
+      "step": 17410
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.994848823740401,
+      "learning_rate": 6.811695965736176e-07,
+      "loss": 0.1641,
+      "step": 17411
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2405990289193678,
+      "learning_rate": 6.805722704825379e-07,
+      "loss": 0.1834,
+      "step": 17412
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.263778459068319,
+      "learning_rate": 6.799751971814628e-07,
+      "loss": 0.1523,
+      "step": 17413
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.227892496199799,
+      "learning_rate": 6.793783766865858e-07,
+      "loss": 0.1575,
+      "step": 17414
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9766224628091549,
+      "learning_rate": 6.787818090140985e-07,
+      "loss": 0.1624,
+      "step": 17415
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0826552449351832,
+      "learning_rate": 6.781854941801802e-07,
+      "loss": 0.1557,
+      "step": 17416
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.076768625593469,
+      "learning_rate": 6.775894322010079e-07,
+      "loss": 0.1573,
+      "step": 17417
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.107756832754881,
+      "learning_rate": 6.769936230927477e-07,
+      "loss": 0.1561,
+      "step": 17418
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9123922774250832,
+      "learning_rate": 6.763980668715631e-07,
+      "loss": 0.1576,
+      "step": 17419
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4473130817984259,
+      "learning_rate": 6.758027635536057e-07,
+      "loss": 0.1613,
+      "step": 17420
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.038560087888518,
+      "learning_rate": 6.752077131550272e-07,
+      "loss": 0.1369,
+      "step": 17421
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1176387403979628,
+      "learning_rate": 6.746129156919645e-07,
+      "loss": 0.164,
+      "step": 17422
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4993249242433018,
+      "learning_rate": 6.740183711805537e-07,
+      "loss": 0.1457,
+      "step": 17423
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.496667523309315,
+      "learning_rate": 6.734240796369207e-07,
+      "loss": 0.159,
+      "step": 17424
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.389756354612381,
+      "learning_rate": 6.728300410771871e-07,
+      "loss": 0.1447,
+      "step": 17425
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2795009098863095,
+      "learning_rate": 6.722362555174644e-07,
+      "loss": 0.1725,
+      "step": 17426
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9403435892659728,
+      "learning_rate": 6.71642722973862e-07,
+      "loss": 0.1661,
+      "step": 17427
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8253890419597139,
+      "learning_rate": 6.710494434624781e-07,
+      "loss": 0.1522,
+      "step": 17428
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9326051873116562,
+      "learning_rate": 6.704564169994022e-07,
+      "loss": 0.1802,
+      "step": 17429
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.464770167732049,
+      "learning_rate": 6.698636436007256e-07,
+      "loss": 0.1509,
+      "step": 17430
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1107217098498596,
+      "learning_rate": 6.692711232825222e-07,
+      "loss": 0.1652,
+      "step": 17431
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.305902964651348,
+      "learning_rate": 6.686788560608671e-07,
+      "loss": 0.1673,
+      "step": 17432
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8709144814823717,
+      "learning_rate": 6.680868419518249e-07,
+      "loss": 0.1666,
+      "step": 17433
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1971517608356697,
+      "learning_rate": 6.674950809714553e-07,
+      "loss": 0.1649,
+      "step": 17434
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9360655756463602,
+      "learning_rate": 6.669035731358075e-07,
+      "loss": 0.1852,
+      "step": 17435
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0354548469117784,
+      "learning_rate": 6.663123184609299e-07,
+      "loss": 0.1611,
+      "step": 17436
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.266579707399008,
+      "learning_rate": 6.657213169628551e-07,
+      "loss": 0.1612,
+      "step": 17437
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0852045706792273,
+      "learning_rate": 6.651305686576182e-07,
+      "loss": 0.1614,
+      "step": 17438
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4032412323848755,
+      "learning_rate": 6.645400735612417e-07,
+      "loss": 0.1786,
+      "step": 17439
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.065819611607627,
+      "learning_rate": 6.639498316897419e-07,
+      "loss": 0.165,
+      "step": 17440
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8775842417319027,
+      "learning_rate": 6.633598430591304e-07,
+      "loss": 0.1523,
+      "step": 17441
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9855080549366656,
+      "learning_rate": 6.627701076854121e-07,
+      "loss": 0.1678,
+      "step": 17442
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8520611272160736,
+      "learning_rate": 6.621806255845797e-07,
+      "loss": 0.177,
+      "step": 17443
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3422182280118413,
+      "learning_rate": 6.615913967726273e-07,
+      "loss": 0.162,
+      "step": 17444
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8702945134702448,
+      "learning_rate": 6.610024212655364e-07,
+      "loss": 0.1519,
+      "step": 17445
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8964156650542517,
+      "learning_rate": 6.604136990792797e-07,
+      "loss": 0.1481,
+      "step": 17446
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5792301251622436,
+      "learning_rate": 6.598252302298313e-07,
+      "loss": 0.1649,
+      "step": 17447
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0581326679188312,
+      "learning_rate": 6.592370147331495e-07,
+      "loss": 0.159,
+      "step": 17448
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5677395291460763,
+      "learning_rate": 6.586490526051903e-07,
+      "loss": 0.1634,
+      "step": 17449
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9677135043618276,
+      "learning_rate": 6.580613438619044e-07,
+      "loss": 0.1741,
+      "step": 17450
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1615411184151843,
+      "learning_rate": 6.574738885192322e-07,
+      "loss": 0.1679,
+      "step": 17451
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0354720596249762,
+      "learning_rate": 6.568866865931078e-07,
+      "loss": 0.1655,
+      "step": 17452
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9829772905302709,
+      "learning_rate": 6.562997380994618e-07,
+      "loss": 0.1731,
+      "step": 17453
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1348821137608869,
+      "learning_rate": 6.557130430542114e-07,
+      "loss": 0.1869,
+      "step": 17454
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9861542333307887,
+      "learning_rate": 6.551266014732738e-07,
+      "loss": 0.1699,
+      "step": 17455
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2911984007830952,
+      "learning_rate": 6.54540413372553e-07,
+      "loss": 0.1657,
+      "step": 17456
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.093145298497331,
+      "learning_rate": 6.53954478767953e-07,
+      "loss": 0.1652,
+      "step": 17457
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9231078419410501,
+      "learning_rate": 6.533687976753644e-07,
+      "loss": 0.1535,
+      "step": 17458
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.001758271155291,
+      "learning_rate": 6.527833701106745e-07,
+      "loss": 0.1653,
+      "step": 17459
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9807167557360661,
+      "learning_rate": 6.521981960897639e-07,
+      "loss": 0.1545,
+      "step": 17460
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0752620461582922,
+      "learning_rate": 6.516132756285065e-07,
+      "loss": 0.1577,
+      "step": 17461
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9511644088197974,
+      "learning_rate": 6.510286087427664e-07,
+      "loss": 0.1579,
+      "step": 17462
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.01542497369087,
+      "learning_rate": 6.504441954484042e-07,
+      "loss": 0.1461,
+      "step": 17463
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0220895756406778,
+      "learning_rate": 6.498600357612717e-07,
+      "loss": 0.1428,
+      "step": 17464
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1266382645524304,
+      "learning_rate": 6.492761296972117e-07,
+      "loss": 0.1868,
+      "step": 17465
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2207458765720829,
+      "learning_rate": 6.486924772720648e-07,
+      "loss": 0.1826,
+      "step": 17466
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2781352593174329,
+      "learning_rate": 6.481090785016631e-07,
+      "loss": 0.1654,
+      "step": 17467
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1189948022298877,
+      "learning_rate": 6.475259334018314e-07,
+      "loss": 0.1608,
+      "step": 17468
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0855749092863607,
+      "learning_rate": 6.46943041988387e-07,
+      "loss": 0.153,
+      "step": 17469
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9274200063910341,
+      "learning_rate": 6.463604042771409e-07,
+      "loss": 0.1548,
+      "step": 17470
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9209645234721234,
+      "learning_rate": 6.457780202838959e-07,
+      "loss": 0.1528,
+      "step": 17471
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9047802876618245,
+      "learning_rate": 6.451958900244526e-07,
+      "loss": 0.1596,
+      "step": 17472
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0981685953170108,
+      "learning_rate": 6.446140135145973e-07,
+      "loss": 0.152,
+      "step": 17473
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8935378090645346,
+      "learning_rate": 6.440323907701173e-07,
+      "loss": 0.1533,
+      "step": 17474
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8424974250053628,
+      "learning_rate": 6.434510218067846e-07,
+      "loss": 0.1673,
+      "step": 17475
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8613847064248749,
+      "learning_rate": 6.428699066403721e-07,
+      "loss": 0.1462,
+      "step": 17476
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0212232620116857,
+      "learning_rate": 6.422890452866415e-07,
+      "loss": 0.1787,
+      "step": 17477
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.7089778175383827,
+      "learning_rate": 6.417084377613514e-07,
+      "loss": 0.1809,
+      "step": 17478
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0972190976416634,
+      "learning_rate": 6.411280840802459e-07,
+      "loss": 0.145,
+      "step": 17479
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1042160964760777,
+      "learning_rate": 6.405479842590723e-07,
+      "loss": 0.1591,
+      "step": 17480
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9220425796674508,
+      "learning_rate": 6.399681383135625e-07,
+      "loss": 0.1582,
+      "step": 17481
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9683037480178739,
+      "learning_rate": 6.39388546259444e-07,
+      "loss": 0.1504,
+      "step": 17482
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.115613663388435,
+      "learning_rate": 6.388092081124398e-07,
+      "loss": 0.1617,
+      "step": 17483
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7121992562713897,
+      "learning_rate": 6.382301238882649e-07,
+      "loss": 0.1673,
+      "step": 17484
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0209392183063348,
+      "learning_rate": 6.37651293602628e-07,
+      "loss": 0.1584,
+      "step": 17485
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.334734571271829,
+      "learning_rate": 6.370727172712276e-07,
+      "loss": 0.1617,
+      "step": 17486
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9918643664287256,
+      "learning_rate": 6.364943949097591e-07,
+      "loss": 0.1551,
+      "step": 17487
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9000122737801736,
+      "learning_rate": 6.359163265339085e-07,
+      "loss": 0.1652,
+      "step": 17488
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0864971415623828,
+      "learning_rate": 6.353385121593569e-07,
+      "loss": 0.1427,
+      "step": 17489
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.266860418067606,
+      "learning_rate": 6.347609518017761e-07,
+      "loss": 0.1563,
+      "step": 17490
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0140534448588183,
+      "learning_rate": 6.341836454768358e-07,
+      "loss": 0.1619,
+      "step": 17491
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3410852805223372,
+      "learning_rate": 6.3360659320019e-07,
+      "loss": 0.1645,
+      "step": 17492
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.255999216791831,
+      "learning_rate": 6.330297949874952e-07,
+      "loss": 0.1625,
+      "step": 17493
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2359667320876915,
+      "learning_rate": 6.324532508543967e-07,
+      "loss": 0.1781,
+      "step": 17494
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0118891462587718,
+      "learning_rate": 6.318769608165332e-07,
+      "loss": 0.1705,
+      "step": 17495
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8681988186278751,
+      "learning_rate": 6.313009248895352e-07,
+      "loss": 0.1352,
+      "step": 17496
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8505491563844542,
+      "learning_rate": 6.307251430890315e-07,
+      "loss": 0.1637,
+      "step": 17497
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9572439560590739,
+      "learning_rate": 6.301496154306363e-07,
+      "loss": 0.1578,
+      "step": 17498
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1260369067949432,
+      "learning_rate": 6.295743419299605e-07,
+      "loss": 0.1516,
+      "step": 17499
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9257905386818395,
+      "learning_rate": 6.289993226026114e-07,
+      "loss": 0.1556,
+      "step": 17500
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5032840341708493,
+      "learning_rate": 6.284245574641834e-07,
+      "loss": 0.1533,
+      "step": 17501
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8893792018006633,
+      "learning_rate": 6.278500465302684e-07,
+      "loss": 0.1584,
+      "step": 17502
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2981337750490072,
+      "learning_rate": 6.272757898164506e-07,
+      "loss": 0.1617,
+      "step": 17503
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0296535628324643,
+      "learning_rate": 6.267017873383085e-07,
+      "loss": 0.1752,
+      "step": 17504
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9195443136945383,
+      "learning_rate": 6.261280391114077e-07,
+      "loss": 0.1646,
+      "step": 17505
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9677007307761413,
+      "learning_rate": 6.255545451513146e-07,
+      "loss": 0.1379,
+      "step": 17506
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4911280417441737,
+      "learning_rate": 6.24981305473582e-07,
+      "loss": 0.1789,
+      "step": 17507
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4267656167569154,
+      "learning_rate": 6.244083200937634e-07,
+      "loss": 0.1667,
+      "step": 17508
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8906435122285429,
+      "learning_rate": 6.238355890273973e-07,
+      "loss": 0.1511,
+      "step": 17509
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.120302944538742,
+      "learning_rate": 6.232631122900201e-07,
+      "loss": 0.1746,
+      "step": 17510
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0361573831639836,
+      "learning_rate": 6.226908898971596e-07,
+      "loss": 0.1612,
+      "step": 17511
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9827383407813749,
+      "learning_rate": 6.221189218643409e-07,
+      "loss": 0.1701,
+      "step": 17512
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2483588098983063,
+      "learning_rate": 6.21547208207075e-07,
+      "loss": 0.1732,
+      "step": 17513
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4254846113606228,
+      "learning_rate": 6.209757489408719e-07,
+      "loss": 0.1491,
+      "step": 17514
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.438762605815942,
+      "learning_rate": 6.2040454408123e-07,
+      "loss": 0.1507,
+      "step": 17515
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0460399471502997,
+      "learning_rate": 6.198335936436451e-07,
+      "loss": 0.1535,
+      "step": 17516
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.253299340529797,
+      "learning_rate": 6.192628976436044e-07,
+      "loss": 0.1462,
+      "step": 17517
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 15.88619169160222,
+      "learning_rate": 6.186924560965856e-07,
+      "loss": 0.1825,
+      "step": 17518
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3292036213000302,
+      "learning_rate": 6.181222690180644e-07,
+      "loss": 0.1443,
+      "step": 17519
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.03504688979119,
+      "learning_rate": 6.175523364235059e-07,
+      "loss": 0.1683,
+      "step": 17520
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.253321521401051,
+      "learning_rate": 6.169826583283722e-07,
+      "loss": 0.1688,
+      "step": 17521
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0547134303571801,
+      "learning_rate": 6.164132347481122e-07,
+      "loss": 0.1816,
+      "step": 17522
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9580859608411995,
+      "learning_rate": 6.158440656981746e-07,
+      "loss": 0.1651,
+      "step": 17523
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.509856539412932,
+      "learning_rate": 6.152751511939947e-07,
+      "loss": 0.158,
+      "step": 17524
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3455717781618253,
+      "learning_rate": 6.147064912510093e-07,
+      "loss": 0.1746,
+      "step": 17525
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.227922347930265,
+      "learning_rate": 6.14138085884638e-07,
+      "loss": 0.1398,
+      "step": 17526
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1176263793989942,
+      "learning_rate": 6.135699351103031e-07,
+      "loss": 0.1848,
+      "step": 17527
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8362091794328037,
+      "learning_rate": 6.130020389434121e-07,
+      "loss": 0.155,
+      "step": 17528
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9411800601075747,
+      "learning_rate": 6.124343973993707e-07,
+      "loss": 0.1598,
+      "step": 17529
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0472801531129503,
+      "learning_rate": 6.118670104935765e-07,
+      "loss": 0.1622,
+      "step": 17530
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2564447978249575,
+      "learning_rate": 6.112998782414215e-07,
+      "loss": 0.1668,
+      "step": 17531
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8298318531080735,
+      "learning_rate": 6.107330006582878e-07,
+      "loss": 0.1592,
+      "step": 17532
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.296337598576234,
+      "learning_rate": 6.101663777595501e-07,
+      "loss": 0.1737,
+      "step": 17533
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9101423442493164,
+      "learning_rate": 6.096000095605814e-07,
+      "loss": 0.1602,
+      "step": 17534
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9812618886749092,
+      "learning_rate": 6.090338960767417e-07,
+      "loss": 0.168,
+      "step": 17535
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2866841462265377,
+      "learning_rate": 6.084680373233875e-07,
+      "loss": 0.1703,
+      "step": 17536
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9056990924642179,
+      "learning_rate": 6.079024333158679e-07,
+      "loss": 0.1768,
+      "step": 17537
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0204475080312303,
+      "learning_rate": 6.073370840695269e-07,
+      "loss": 0.176,
+      "step": 17538
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3025219249136724,
+      "learning_rate": 6.067719895996971e-07,
+      "loss": 0.1602,
+      "step": 17539
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2543799126764876,
+      "learning_rate": 6.062071499217081e-07,
+      "loss": 0.162,
+      "step": 17540
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2796647743558915,
+      "learning_rate": 6.056425650508801e-07,
+      "loss": 0.1593,
+      "step": 17541
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9443775080108795,
+      "learning_rate": 6.050782350025297e-07,
+      "loss": 0.172,
+      "step": 17542
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2029689130455121,
+      "learning_rate": 6.045141597919613e-07,
+      "loss": 0.1822,
+      "step": 17543
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.939996658457354,
+      "learning_rate": 6.039503394344782e-07,
+      "loss": 0.1472,
+      "step": 17544
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1660803223243563,
+      "learning_rate": 6.033867739453703e-07,
+      "loss": 0.1518,
+      "step": 17545
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8682073832006625,
+      "learning_rate": 6.028234633399277e-07,
+      "loss": 0.1602,
+      "step": 17546
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5234733333259054,
+      "learning_rate": 6.022604076334304e-07,
+      "loss": 0.1513,
+      "step": 17547
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9627404893959367,
+      "learning_rate": 6.016976068411506e-07,
+      "loss": 0.1587,
+      "step": 17548
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0389853213896696,
+      "learning_rate": 6.011350609783529e-07,
+      "loss": 0.1841,
+      "step": 17549
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.319045427716174,
+      "learning_rate": 6.005727700602992e-07,
+      "loss": 0.1784,
+      "step": 17550
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9033564901415023,
+      "learning_rate": 6.000107341022399e-07,
+      "loss": 0.1841,
+      "step": 17551
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0968342902108015,
+      "learning_rate": 5.994489531194192e-07,
+      "loss": 0.1501,
+      "step": 17552
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2034136324685407,
+      "learning_rate": 5.988874271270773e-07,
+      "loss": 0.1458,
+      "step": 17553
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7376260021799919,
+      "learning_rate": 5.983261561404441e-07,
+      "loss": 0.1727,
+      "step": 17554
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1798377881298199,
+      "learning_rate": 5.977651401747442e-07,
+      "loss": 0.1375,
+      "step": 17555
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7601953087649302,
+      "learning_rate": 5.972043792451964e-07,
+      "loss": 0.1665,
+      "step": 17556
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0292216285996567,
+      "learning_rate": 5.966438733670121e-07,
+      "loss": 0.1547,
+      "step": 17557
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0016902697828487,
+      "learning_rate": 5.960836225553923e-07,
+      "loss": 0.1586,
+      "step": 17558
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8752761730699214,
+      "learning_rate": 5.955236268255372e-07,
+      "loss": 0.1724,
+      "step": 17559
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.607044011795207,
+      "learning_rate": 5.949638861926333e-07,
+      "loss": 0.1407,
+      "step": 17560
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1743967307776313,
+      "learning_rate": 5.944044006718674e-07,
+      "loss": 0.1352,
+      "step": 17561
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0984485539787459,
+      "learning_rate": 5.938451702784109e-07,
+      "loss": 0.1715,
+      "step": 17562
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7723001616527106,
+      "learning_rate": 5.93286195027436e-07,
+      "loss": 0.1623,
+      "step": 17563
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3667467559842605,
+      "learning_rate": 5.927274749341039e-07,
+      "loss": 0.1737,
+      "step": 17564
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.078327802462095,
+      "learning_rate": 5.921690100135713e-07,
+      "loss": 0.1619,
+      "step": 17565
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1087345773722312,
+      "learning_rate": 5.916108002809851e-07,
+      "loss": 0.1758,
+      "step": 17566
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.931356983992569,
+      "learning_rate": 5.910528457514886e-07,
+      "loss": 0.1589,
+      "step": 17567
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9761997564311885,
+      "learning_rate": 5.904951464402154e-07,
+      "loss": 0.1704,
+      "step": 17568
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8754256176667727,
+      "learning_rate": 5.89937702362291e-07,
+      "loss": 0.171,
+      "step": 17569
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9700307877911704,
+      "learning_rate": 5.89380513532839e-07,
+      "loss": 0.1542,
+      "step": 17570
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8100403713688008,
+      "learning_rate": 5.888235799669705e-07,
+      "loss": 0.1737,
+      "step": 17571
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8726291967736584,
+      "learning_rate": 5.882669016797948e-07,
+      "loss": 0.1742,
+      "step": 17572
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1240848985822962,
+      "learning_rate": 5.877104786864107e-07,
+      "loss": 0.1621,
+      "step": 17573
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0396680948721067,
+      "learning_rate": 5.871543110019128e-07,
+      "loss": 0.1665,
+      "step": 17574
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1313393019502156,
+      "learning_rate": 5.865983986413848e-07,
+      "loss": 0.1529,
+      "step": 17575
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0259427094500448,
+      "learning_rate": 5.860427416199077e-07,
+      "loss": 0.1855,
+      "step": 17576
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0316922210292752,
+      "learning_rate": 5.854873399525518e-07,
+      "loss": 0.1643,
+      "step": 17577
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8627959272048978,
+      "learning_rate": 5.84932193654385e-07,
+      "loss": 0.1958,
+      "step": 17578
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8468770809456929,
+      "learning_rate": 5.843773027404631e-07,
+      "loss": 0.1748,
+      "step": 17579
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9074617601974535,
+      "learning_rate": 5.838226672258374e-07,
+      "loss": 0.1806,
+      "step": 17580
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8759627401266152,
+      "learning_rate": 5.832682871255546e-07,
+      "loss": 0.1743,
+      "step": 17581
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1549145279627093,
+      "learning_rate": 5.827141624546528e-07,
+      "loss": 0.1754,
+      "step": 17582
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9652020221208394,
+      "learning_rate": 5.8216029322816e-07,
+      "loss": 0.1497,
+      "step": 17583
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2358436435062201,
+      "learning_rate": 5.816066794611019e-07,
+      "loss": 0.1467,
+      "step": 17584
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2438610313138372,
+      "learning_rate": 5.810533211684954e-07,
+      "loss": 0.1646,
+      "step": 17585
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8587535130383744,
+      "learning_rate": 5.805002183653474e-07,
+      "loss": 0.1407,
+      "step": 17586
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1335931999480722,
+      "learning_rate": 5.799473710666659e-07,
+      "loss": 0.1711,
+      "step": 17587
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4396105891990831,
+      "learning_rate": 5.793947792874411e-07,
+      "loss": 0.1638,
+      "step": 17588
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9790535477952904,
+      "learning_rate": 5.788424430426653e-07,
+      "loss": 0.1857,
+      "step": 17589
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8856226673905241,
+      "learning_rate": 5.782903623473202e-07,
+      "loss": 0.1668,
+      "step": 17590
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.067043539477932,
+      "learning_rate": 5.777385372163824e-07,
+      "loss": 0.174,
+      "step": 17591
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8740321505398196,
+      "learning_rate": 5.771869676648178e-07,
+      "loss": 0.15,
+      "step": 17592
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8681917829340098,
+      "learning_rate": 5.766356537075913e-07,
+      "loss": 0.1538,
+      "step": 17593
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9506182517731803,
+      "learning_rate": 5.760845953596527e-07,
+      "loss": 0.1759,
+      "step": 17594
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3177971254646998,
+      "learning_rate": 5.755337926359528e-07,
+      "loss": 0.1631,
+      "step": 17595
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.117532008903971,
+      "learning_rate": 5.749832455514292e-07,
+      "loss": 0.1544,
+      "step": 17596
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2500663676142996,
+      "learning_rate": 5.744329541210203e-07,
+      "loss": 0.1733,
+      "step": 17597
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.1405137035890365,
+      "learning_rate": 5.738829183596472e-07,
+      "loss": 0.1763,
+      "step": 17598
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.364590498600788,
+      "learning_rate": 5.733331382822327e-07,
+      "loss": 0.1627,
+      "step": 17599
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3966575165611175,
+      "learning_rate": 5.727836139036902e-07,
+      "loss": 0.1771,
+      "step": 17600
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4316548739683352,
+      "learning_rate": 5.722343452389246e-07,
+      "loss": 0.1829,
+      "step": 17601
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9225231916886446,
+      "learning_rate": 5.71685332302836e-07,
+      "loss": 0.1652,
+      "step": 17602
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9262842826877503,
+      "learning_rate": 5.711365751103126e-07,
+      "loss": 0.1615,
+      "step": 17603
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.125458347496385,
+      "learning_rate": 5.705880736762448e-07,
+      "loss": 0.1555,
+      "step": 17604
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0342644278015851,
+      "learning_rate": 5.700398280155062e-07,
+      "loss": 0.1555,
+      "step": 17605
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.657524375857265,
+      "learning_rate": 5.694918381429693e-07,
+      "loss": 0.1715,
+      "step": 17606
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7613783650392223,
+      "learning_rate": 5.689441040735e-07,
+      "loss": 0.1705,
+      "step": 17607
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1654114356377074,
+      "learning_rate": 5.683966258219553e-07,
+      "loss": 0.1794,
+      "step": 17608
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.117362526105987,
+      "learning_rate": 5.678494034031834e-07,
+      "loss": 0.1673,
+      "step": 17609
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9317194347022243,
+      "learning_rate": 5.673024368320313e-07,
+      "loss": 0.1671,
+      "step": 17610
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.772568989083743,
+      "learning_rate": 5.667557261233303e-07,
+      "loss": 0.1537,
+      "step": 17611
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8476355847340409,
+      "learning_rate": 5.662092712919165e-07,
+      "loss": 0.1822,
+      "step": 17612
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9599586056070369,
+      "learning_rate": 5.656630723526058e-07,
+      "loss": 0.1925,
+      "step": 17613
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.081935044119475,
+      "learning_rate": 5.651171293202195e-07,
+      "loss": 0.1543,
+      "step": 17614
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1782749849930128,
+      "learning_rate": 5.645714422095627e-07,
+      "loss": 0.1662,
+      "step": 17615
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.038120793501522,
+      "learning_rate": 5.640260110354379e-07,
+      "loss": 0.1791,
+      "step": 17616
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0742685175075855,
+      "learning_rate": 5.634808358126409e-07,
+      "loss": 0.169,
+      "step": 17617
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.490681460837666,
+      "learning_rate": 5.6293591655596e-07,
+      "loss": 0.1691,
+      "step": 17618
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0405156137959275,
+      "learning_rate": 5.623912532801745e-07,
+      "loss": 0.2042,
+      "step": 17619
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4385736720037017,
+      "learning_rate": 5.618468460000603e-07,
+      "loss": 0.1543,
+      "step": 17620
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3147500473045866,
+      "learning_rate": 5.613026947303846e-07,
+      "loss": 0.1723,
+      "step": 17621
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6507261497711412,
+      "learning_rate": 5.607587994859043e-07,
+      "loss": 0.173,
+      "step": 17622
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0463045500938963,
+      "learning_rate": 5.602151602813754e-07,
+      "loss": 0.1565,
+      "step": 17623
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9064519461727782,
+      "learning_rate": 5.59671777131543e-07,
+      "loss": 0.1683,
+      "step": 17624
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9830407275206406,
+      "learning_rate": 5.591286500511461e-07,
+      "loss": 0.1525,
+      "step": 17625
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4047169508316093,
+      "learning_rate": 5.585857790549176e-07,
+      "loss": 0.1647,
+      "step": 17626
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.993469300350806,
+      "learning_rate": 5.580431641575856e-07,
+      "loss": 0.1591,
+      "step": 17627
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.166303889142174,
+      "learning_rate": 5.57500805373864e-07,
+      "loss": 0.1714,
+      "step": 17628
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8781195607529766,
+      "learning_rate": 5.569587027184676e-07,
+      "loss": 0.1613,
+      "step": 17629
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.885899291646843,
+      "learning_rate": 5.564168562060989e-07,
+      "loss": 0.1503,
+      "step": 17630
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7808838110262488,
+      "learning_rate": 5.558752658514576e-07,
+      "loss": 0.1421,
+      "step": 17631
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.129720678794883,
+      "learning_rate": 5.553339316692319e-07,
+      "loss": 0.1671,
+      "step": 17632
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9170876078394514,
+      "learning_rate": 5.547928536741054e-07,
+      "loss": 0.1515,
+      "step": 17633
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.195046931226926,
+      "learning_rate": 5.542520318807575e-07,
+      "loss": 0.1425,
+      "step": 17634
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.289422396897735,
+      "learning_rate": 5.537114663038579e-07,
+      "loss": 0.1497,
+      "step": 17635
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.002269400661614,
+      "learning_rate": 5.531711569580667e-07,
+      "loss": 0.171,
+      "step": 17636
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1868860593512662,
+      "learning_rate": 5.526311038580434e-07,
+      "loss": 0.1779,
+      "step": 17637
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7414258773959721,
+      "learning_rate": 5.520913070184342e-07,
+      "loss": 0.1615,
+      "step": 17638
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4872577817974402,
+      "learning_rate": 5.515517664538816e-07,
+      "loss": 0.1672,
+      "step": 17639
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1392463829779043,
+      "learning_rate": 5.510124821790208e-07,
+      "loss": 0.1743,
+      "step": 17640
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9808252375289511,
+      "learning_rate": 5.5047345420848e-07,
+      "loss": 0.1713,
+      "step": 17641
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0052517052647605,
+      "learning_rate": 5.499346825568796e-07,
+      "loss": 0.1572,
+      "step": 17642
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9217200434394364,
+      "learning_rate": 5.493961672388359e-07,
+      "loss": 0.1684,
+      "step": 17643
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9372417416015802,
+      "learning_rate": 5.488579082689549e-07,
+      "loss": 0.186,
+      "step": 17644
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3092412178722737,
+      "learning_rate": 5.48319905661836e-07,
+      "loss": 0.1465,
+      "step": 17645
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.002333418364241,
+      "learning_rate": 5.477821594320754e-07,
+      "loss": 0.146,
+      "step": 17646
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8780082868559859,
+      "learning_rate": 5.472446695942557e-07,
+      "loss": 0.1576,
+      "step": 17647
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1158135010797137,
+      "learning_rate": 5.467074361629599e-07,
+      "loss": 0.1707,
+      "step": 17648
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2440407124510282,
+      "learning_rate": 5.461704591527573e-07,
+      "loss": 0.1658,
+      "step": 17649
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0253371799074213,
+      "learning_rate": 5.456337385782173e-07,
+      "loss": 0.1883,
+      "step": 17650
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.915436610916007,
+      "learning_rate": 5.450972744538929e-07,
+      "loss": 0.1639,
+      "step": 17651
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0029131128139648,
+      "learning_rate": 5.445610667943401e-07,
+      "loss": 0.164,
+      "step": 17652
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9462649352774675,
+      "learning_rate": 5.440251156141019e-07,
+      "loss": 0.1628,
+      "step": 17653
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3574479219373592,
+      "learning_rate": 5.434894209277186e-07,
+      "loss": 0.1578,
+      "step": 17654
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9918804611222274,
+      "learning_rate": 5.429539827497188e-07,
+      "loss": 0.1497,
+      "step": 17655
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0503724191185535,
+      "learning_rate": 5.424188010946241e-07,
+      "loss": 0.1566,
+      "step": 17656
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8451239880660211,
+      "learning_rate": 5.418838759769551e-07,
+      "loss": 0.1548,
+      "step": 17657
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.893324799854365,
+      "learning_rate": 5.413492074112192e-07,
+      "loss": 0.1468,
+      "step": 17658
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0016104659910292,
+      "learning_rate": 5.408147954119202e-07,
+      "loss": 0.1582,
+      "step": 17659
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8984243797807004,
+      "learning_rate": 5.402806399935545e-07,
+      "loss": 0.176,
+      "step": 17660
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.62575895769711,
+      "learning_rate": 5.397467411706114e-07,
+      "loss": 0.1525,
+      "step": 17661
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0367581809540989,
+      "learning_rate": 5.392130989575716e-07,
+      "loss": 0.1787,
+      "step": 17662
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0075030435562442,
+      "learning_rate": 5.386797133689125e-07,
+      "loss": 0.1581,
+      "step": 17663
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9308276120576201,
+      "learning_rate": 5.38146584419098e-07,
+      "loss": 0.1415,
+      "step": 17664
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9995172688457722,
+      "learning_rate": 5.376137121225933e-07,
+      "loss": 0.1664,
+      "step": 17665
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1181798407086685,
+      "learning_rate": 5.370810964938511e-07,
+      "loss": 0.1645,
+      "step": 17666
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8228683638441441,
+      "learning_rate": 5.365487375473189e-07,
+      "loss": 0.1484,
+      "step": 17667
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.91897186212851,
+      "learning_rate": 5.36016635297435e-07,
+      "loss": 0.1423,
+      "step": 17668
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9404184974406693,
+      "learning_rate": 5.354847897586346e-07,
+      "loss": 0.1536,
+      "step": 17669
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9341398123651984,
+      "learning_rate": 5.34953200945344e-07,
+      "loss": 0.1656,
+      "step": 17670
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9941746446631397,
+      "learning_rate": 5.344218688719849e-07,
+      "loss": 0.1838,
+      "step": 17671
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2760959638853353,
+      "learning_rate": 5.338907935529658e-07,
+      "loss": 0.1673,
+      "step": 17672
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2152025882319166,
+      "learning_rate": 5.33359975002693e-07,
+      "loss": 0.1714,
+      "step": 17673
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3978400774163173,
+      "learning_rate": 5.32829413235566e-07,
+      "loss": 0.1566,
+      "step": 17674
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.228025280490407,
+      "learning_rate": 5.322991082659745e-07,
+      "loss": 0.1587,
+      "step": 17675
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6220795682324534,
+      "learning_rate": 5.31769060108307e-07,
+      "loss": 0.1669,
+      "step": 17676
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9484679760869359,
+      "learning_rate": 5.312392687769363e-07,
+      "loss": 0.1646,
+      "step": 17677
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7112336753826645,
+      "learning_rate": 5.307097342862355e-07,
+      "loss": 0.1476,
+      "step": 17678
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9050168602368139,
+      "learning_rate": 5.301804566505675e-07,
+      "loss": 0.1739,
+      "step": 17679
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6948532263660625,
+      "learning_rate": 5.296514358842919e-07,
+      "loss": 0.1562,
+      "step": 17680
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9223356351999948,
+      "learning_rate": 5.291226720017539e-07,
+      "loss": 0.1646,
+      "step": 17681
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0710729372748458,
+      "learning_rate": 5.285941650172999e-07,
+      "loss": 0.1781,
+      "step": 17682
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0156542824190222,
+      "learning_rate": 5.280659149452627e-07,
+      "loss": 0.161,
+      "step": 17683
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0609702715785894,
+      "learning_rate": 5.275379217999732e-07,
+      "loss": 0.1584,
+      "step": 17684
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.593540452877707,
+      "learning_rate": 5.270101855957521e-07,
+      "loss": 0.1498,
+      "step": 17685
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9987113372708541,
+      "learning_rate": 5.264827063469146e-07,
+      "loss": 0.1667,
+      "step": 17686
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9799744766968221,
+      "learning_rate": 5.259554840677683e-07,
+      "loss": 0.1543,
+      "step": 17687
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0793003127543954,
+      "learning_rate": 5.25428518772616e-07,
+      "loss": 0.1772,
+      "step": 17688
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1104727507892227,
+      "learning_rate": 5.249018104757498e-07,
+      "loss": 0.1648,
+      "step": 17689
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8866876360592496,
+      "learning_rate": 5.24375359191458e-07,
+      "loss": 0.1711,
+      "step": 17690
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9318746380239975,
+      "learning_rate": 5.238491649340194e-07,
+      "loss": 0.1576,
+      "step": 17691
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.704580316984937,
+      "learning_rate": 5.233232277177058e-07,
+      "loss": 0.1481,
+      "step": 17692
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9707591751898252,
+      "learning_rate": 5.22797547556787e-07,
+      "loss": 0.198,
+      "step": 17693
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4708636570765141,
+      "learning_rate": 5.222721244655182e-07,
+      "loss": 0.1606,
+      "step": 17694
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0986553399787764,
+      "learning_rate": 5.217469584581536e-07,
+      "loss": 0.1719,
+      "step": 17695
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.24716223435193,
+      "learning_rate": 5.212220495489384e-07,
+      "loss": 0.1562,
+      "step": 17696
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8958578576870243,
+      "learning_rate": 5.206973977521113e-07,
+      "loss": 0.1731,
+      "step": 17697
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1951818186829957,
+      "learning_rate": 5.201730030819019e-07,
+      "loss": 0.1544,
+      "step": 17698
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4365885572432668,
+      "learning_rate": 5.196488655525356e-07,
+      "loss": 0.1787,
+      "step": 17699
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2545363657750488,
+      "learning_rate": 5.191249851782287e-07,
+      "loss": 0.145,
+      "step": 17700
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2246894585487254,
+      "learning_rate": 5.186013619731934e-07,
+      "loss": 0.1638,
+      "step": 17701
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9044530743019654,
+      "learning_rate": 5.180779959516303e-07,
+      "loss": 0.171,
+      "step": 17702
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8542074221845332,
+      "learning_rate": 5.175548871277358e-07,
+      "loss": 0.1606,
+      "step": 17703
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9570768342942108,
+      "learning_rate": 5.170320355157022e-07,
+      "loss": 0.1412,
+      "step": 17704
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.377473388699456,
+      "learning_rate": 5.165094411297111e-07,
+      "loss": 0.1878,
+      "step": 17705
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9740939249314572,
+      "learning_rate": 5.15987103983936e-07,
+      "loss": 0.1739,
+      "step": 17706
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1030183081306857,
+      "learning_rate": 5.154650240925474e-07,
+      "loss": 0.1718,
+      "step": 17707
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9345049545849029,
+      "learning_rate": 5.149432014697053e-07,
+      "loss": 0.1798,
+      "step": 17708
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1709174668507403,
+      "learning_rate": 5.14421636129564e-07,
+      "loss": 0.1574,
+      "step": 17709
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8308813827049973,
+      "learning_rate": 5.13900328086272e-07,
+      "loss": 0.1487,
+      "step": 17710
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.869450981379436,
+      "learning_rate": 5.133792773539681e-07,
+      "loss": 0.1568,
+      "step": 17711
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9869137581204871,
+      "learning_rate": 5.128584839467877e-07,
+      "loss": 0.1441,
+      "step": 17712
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7785372146076209,
+      "learning_rate": 5.12337947878857e-07,
+      "loss": 0.1614,
+      "step": 17713
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9084779713098403,
+      "learning_rate": 5.118176691642962e-07,
+      "loss": 0.144,
+      "step": 17714
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7622362638835993,
+      "learning_rate": 5.11297647817216e-07,
+      "loss": 0.1559,
+      "step": 17715
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6181503739545215,
+      "learning_rate": 5.107778838517241e-07,
+      "loss": 0.1525,
+      "step": 17716
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9288820385651829,
+      "learning_rate": 5.102583772819159e-07,
+      "loss": 0.1488,
+      "step": 17717
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0344836702661204,
+      "learning_rate": 5.097391281218877e-07,
+      "loss": 0.1446,
+      "step": 17718
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.756541256971086,
+      "learning_rate": 5.092201363857197e-07,
+      "loss": 0.1528,
+      "step": 17719
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3498640132390163,
+      "learning_rate": 5.087014020874936e-07,
+      "loss": 0.1704,
+      "step": 17720
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9049343493676423,
+      "learning_rate": 5.081829252412762e-07,
+      "loss": 0.1487,
+      "step": 17721
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.017207296182938,
+      "learning_rate": 5.076647058611328e-07,
+      "loss": 0.1758,
+      "step": 17722
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0428643056624292,
+      "learning_rate": 5.071467439611211e-07,
+      "loss": 0.1455,
+      "step": 17723
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1132281043897432,
+      "learning_rate": 5.066290395552909e-07,
+      "loss": 0.1574,
+      "step": 17724
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6376443306870083,
+      "learning_rate": 5.061115926576843e-07,
+      "loss": 0.1612,
+      "step": 17725
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.537279060722791,
+      "learning_rate": 5.055944032823357e-07,
+      "loss": 0.1628,
+      "step": 17726
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.133942195763958,
+      "learning_rate": 5.050774714432772e-07,
+      "loss": 0.1587,
+      "step": 17727
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6315795559668866,
+      "learning_rate": 5.045607971545263e-07,
+      "loss": 0.1767,
+      "step": 17728
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.406991068781469,
+      "learning_rate": 5.040443804300998e-07,
+      "loss": 0.14,
+      "step": 17729
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3011977502033985,
+      "learning_rate": 5.035282212840065e-07,
+      "loss": 0.182,
+      "step": 17730
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3039874258239579,
+      "learning_rate": 5.030123197302472e-07,
+      "loss": 0.1708,
+      "step": 17731
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.74425755208782,
+      "learning_rate": 5.024966757828143e-07,
+      "loss": 0.1766,
+      "step": 17732
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9541717050863457,
+      "learning_rate": 5.019812894556975e-07,
+      "loss": 0.1522,
+      "step": 17733
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9449006111982539,
+      "learning_rate": 5.014661607628723e-07,
+      "loss": 0.1705,
+      "step": 17734
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.291618929572805,
+      "learning_rate": 5.009512897183156e-07,
+      "loss": 0.1501,
+      "step": 17735
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6137589502704652,
+      "learning_rate": 5.004366763359903e-07,
+      "loss": 0.1692,
+      "step": 17736
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.088489181727494,
+      "learning_rate": 4.999223206298587e-07,
+      "loss": 0.1463,
+      "step": 17737
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7985636651852127,
+      "learning_rate": 4.994082226138686e-07,
+      "loss": 0.116,
+      "step": 17738
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2956073162739354,
+      "learning_rate": 4.988943823019676e-07,
+      "loss": 0.1524,
+      "step": 17739
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4870177517974048,
+      "learning_rate": 4.983807997080925e-07,
+      "loss": 0.1485,
+      "step": 17740
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1141072045469964,
+      "learning_rate": 4.978674748461765e-07,
+      "loss": 0.1965,
+      "step": 17741
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7596014826608797,
+      "learning_rate": 4.973544077301418e-07,
+      "loss": 0.1702,
+      "step": 17742
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8503845471946775,
+      "learning_rate": 4.968415983739039e-07,
+      "loss": 0.1554,
+      "step": 17743
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.225632954529606,
+      "learning_rate": 4.963290467913761e-07,
+      "loss": 0.1623,
+      "step": 17744
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1538746263261441,
+      "learning_rate": 4.958167529964586e-07,
+      "loss": 0.1948,
+      "step": 17745
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0146882938179274,
+      "learning_rate": 4.953047170030489e-07,
+      "loss": 0.155,
+      "step": 17746
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0771355542823016,
+      "learning_rate": 4.94792938825035e-07,
+      "loss": 0.1491,
+      "step": 17747
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.018788055431089,
+      "learning_rate": 4.942814184763001e-07,
+      "loss": 0.1723,
+      "step": 17748
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1316146560079632,
+      "learning_rate": 4.937701559707175e-07,
+      "loss": 0.1602,
+      "step": 17749
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9280565298619311,
+      "learning_rate": 4.932591513221586e-07,
+      "loss": 0.148,
+      "step": 17750
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0565016656740143,
+      "learning_rate": 4.927484045444797e-07,
+      "loss": 0.1574,
+      "step": 17751
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1490442384990298,
+      "learning_rate": 4.922379156515389e-07,
+      "loss": 0.1484,
+      "step": 17752
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0055808135829174,
+      "learning_rate": 4.917276846571806e-07,
+      "loss": 0.1499,
+      "step": 17753
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.137822230302273,
+      "learning_rate": 4.912177115752481e-07,
+      "loss": 0.1698,
+      "step": 17754
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4222448581168465,
+      "learning_rate": 4.907079964195693e-07,
+      "loss": 0.1674,
+      "step": 17755
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0891384710066645,
+      "learning_rate": 4.901985392039743e-07,
+      "loss": 0.1892,
+      "step": 17756
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.239926969876004,
+      "learning_rate": 4.896893399422809e-07,
+      "loss": 0.1529,
+      "step": 17757
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1012776690468857,
+      "learning_rate": 4.891803986483023e-07,
+      "loss": 0.1937,
+      "step": 17758
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0518012215731845,
+      "learning_rate": 4.886717153358411e-07,
+      "loss": 0.199,
+      "step": 17759
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0574357379309232,
+      "learning_rate": 4.881632900186983e-07,
+      "loss": 0.154,
+      "step": 17760
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.175176478735189,
+      "learning_rate": 4.876551227106629e-07,
+      "loss": 0.1801,
+      "step": 17761
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8789036612254449,
+      "learning_rate": 4.871472134255195e-07,
+      "loss": 0.1609,
+      "step": 17762
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.680680067709628,
+      "learning_rate": 4.866395621770458e-07,
+      "loss": 0.1423,
+      "step": 17763
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1085585855081044,
+      "learning_rate": 4.861321689790099e-07,
+      "loss": 0.1458,
+      "step": 17764
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6969531275425647,
+      "learning_rate": 4.856250338451763e-07,
+      "loss": 0.1805,
+      "step": 17765
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9945335175170363,
+      "learning_rate": 4.851181567893015e-07,
+      "loss": 0.1476,
+      "step": 17766
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1633841659011586,
+      "learning_rate": 4.846115378251348e-07,
+      "loss": 0.1401,
+      "step": 17767
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1946895954446297,
+      "learning_rate": 4.841051769664174e-07,
+      "loss": 0.1782,
+      "step": 17768
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0272983655384607,
+      "learning_rate": 4.835990742268848e-07,
+      "loss": 0.1578,
+      "step": 17769
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.881203217537691,
+      "learning_rate": 4.83093229620264e-07,
+      "loss": 0.1575,
+      "step": 17770
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.35515392014797,
+      "learning_rate": 4.825876431602772e-07,
+      "loss": 0.1692,
+      "step": 17771
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2188750181917818,
+      "learning_rate": 4.820823148606379e-07,
+      "loss": 0.183,
+      "step": 17772
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0844096893909176,
+      "learning_rate": 4.815772447350541e-07,
+      "loss": 0.1687,
+      "step": 17773
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.051618523028314,
+      "learning_rate": 4.810724327972238e-07,
+      "loss": 0.1578,
+      "step": 17774
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7931119674665117,
+      "learning_rate": 4.805678790608415e-07,
+      "loss": 0.1636,
+      "step": 17775
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8151408484958761,
+      "learning_rate": 4.80063583539594e-07,
+      "loss": 0.1484,
+      "step": 17776
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0937623325399302,
+      "learning_rate": 4.795595462471592e-07,
+      "loss": 0.1779,
+      "step": 17777
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0631272079242073,
+      "learning_rate": 4.790557671972107e-07,
+      "loss": 0.1397,
+      "step": 17778
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0143499979429285,
+      "learning_rate": 4.785522464034109e-07,
+      "loss": 0.1782,
+      "step": 17779
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.063689312103594,
+      "learning_rate": 4.7804898387942e-07,
+      "loss": 0.1645,
+      "step": 17780
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0920688128771128,
+      "learning_rate": 4.77545979638887e-07,
+      "loss": 0.1579,
+      "step": 17781
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1352514848095434,
+      "learning_rate": 4.770432336954567e-07,
+      "loss": 0.1594,
+      "step": 17782
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8693421343476601,
+      "learning_rate": 4.765407460627669e-07,
+      "loss": 0.1666,
+      "step": 17783
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.512091212186362,
+      "learning_rate": 4.76038516754449e-07,
+      "loss": 0.1636,
+      "step": 17784
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9559664234061493,
+      "learning_rate": 4.755365457841221e-07,
+      "loss": 0.1621,
+      "step": 17785
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.40871843285452,
+      "learning_rate": 4.750348331654064e-07,
+      "loss": 0.1668,
+      "step": 17786
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2538414667969688,
+      "learning_rate": 4.7453337891190776e-07,
+      "loss": 0.1684,
+      "step": 17787
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2285387934323766,
+      "learning_rate": 4.7403218303722963e-07,
+      "loss": 0.1673,
+      "step": 17788
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2252516668439057,
+      "learning_rate": 4.7353124555496566e-07,
+      "loss": 0.1568,
+      "step": 17789
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7644849909791742,
+      "learning_rate": 4.7303056647870605e-07,
+      "loss": 0.1833,
+      "step": 17790
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9280686009723658,
+      "learning_rate": 4.725301458220288e-07,
+      "loss": 0.1575,
+      "step": 17791
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9485796496560795,
+      "learning_rate": 4.7202998359850984e-07,
+      "loss": 0.1594,
+      "step": 17792
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2267938129539502,
+      "learning_rate": 4.7153007982171594e-07,
+      "loss": 0.1909,
+      "step": 17793
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7754106303778403,
+      "learning_rate": 4.7103043450520744e-07,
+      "loss": 0.1457,
+      "step": 17794
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.026104662099303,
+      "learning_rate": 4.705310476625369e-07,
+      "loss": 0.1678,
+      "step": 17795
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1562353558815135,
+      "learning_rate": 4.700319193072489e-07,
+      "loss": 0.1869,
+      "step": 17796
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6137894136029478,
+      "learning_rate": 4.695330494528838e-07,
+      "loss": 0.1491,
+      "step": 17797
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.09089695525694,
+      "learning_rate": 4.69034438112973e-07,
+      "loss": 0.1839,
+      "step": 17798
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.333152341251913,
+      "learning_rate": 4.685360853010401e-07,
+      "loss": 0.1637,
+      "step": 17799
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9602187524299798,
+      "learning_rate": 4.6803799103060544e-07,
+      "loss": 0.1653,
+      "step": 17800
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1799898341032788,
+      "learning_rate": 4.6754015531517926e-07,
+      "loss": 0.1626,
+      "step": 17801
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.665753566346427,
+      "learning_rate": 4.6704257816826306e-07,
+      "loss": 0.1961,
+      "step": 17802
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.640526741340988,
+      "learning_rate": 4.6654525960335704e-07,
+      "loss": 0.1682,
+      "step": 17803
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8810679261644867,
+      "learning_rate": 4.660481996339483e-07,
+      "loss": 0.1621,
+      "step": 17804
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9417092579675105,
+      "learning_rate": 4.655513982735216e-07,
+      "loss": 0.1822,
+      "step": 17805
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1450114665041524,
+      "learning_rate": 4.6505485553555054e-07,
+      "loss": 0.1495,
+      "step": 17806
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2342421265088601,
+      "learning_rate": 4.6455857143350657e-07,
+      "loss": 0.1718,
+      "step": 17807
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4242395673614698,
+      "learning_rate": 4.6406254598084786e-07,
+      "loss": 0.1682,
+      "step": 17808
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9017379280780949,
+      "learning_rate": 4.635667791910314e-07,
+      "loss": 0.1617,
+      "step": 17809
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0869143627570828,
+      "learning_rate": 4.630712710775054e-07,
+      "loss": 0.1639,
+      "step": 17810
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9145479259162557,
+      "learning_rate": 4.625760216537112e-07,
+      "loss": 0.1559,
+      "step": 17811
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0659886892373898,
+      "learning_rate": 4.620810309330803e-07,
+      "loss": 0.1474,
+      "step": 17812
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0288166878075367,
+      "learning_rate": 4.615862989290387e-07,
+      "loss": 0.1814,
+      "step": 17813
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1905630758669472,
+      "learning_rate": 4.61091825655009e-07,
+      "loss": 0.1932,
+      "step": 17814
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0430728818674375,
+      "learning_rate": 4.605976111244015e-07,
+      "loss": 0.1629,
+      "step": 17815
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0384609491251084,
+      "learning_rate": 4.601036553506233e-07,
+      "loss": 0.174,
+      "step": 17816
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4392457512023191,
+      "learning_rate": 4.5960995834707146e-07,
+      "loss": 0.1346,
+      "step": 17817
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9923559260449925,
+      "learning_rate": 4.591165201271386e-07,
+      "loss": 0.1774,
+      "step": 17818
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0656259011880576,
+      "learning_rate": 4.5862334070420843e-07,
+      "loss": 0.1652,
+      "step": 17819
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4443650791473384,
+      "learning_rate": 4.581304200916603e-07,
+      "loss": 0.1698,
+      "step": 17820
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9518129480656117,
+      "learning_rate": 4.576377583028624e-07,
+      "loss": 0.1504,
+      "step": 17821
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1957606269589063,
+      "learning_rate": 4.571453553511807e-07,
+      "loss": 0.176,
+      "step": 17822
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0892795482220845,
+      "learning_rate": 4.5665321124996774e-07,
+      "loss": 0.1595,
+      "step": 17823
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1566948468637954,
+      "learning_rate": 4.5616132601257857e-07,
+      "loss": 0.1695,
+      "step": 17824
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9480355518150048,
+      "learning_rate": 4.556696996523502e-07,
+      "loss": 0.1724,
+      "step": 17825
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0548757699756477,
+      "learning_rate": 4.5517833218261974e-07,
+      "loss": 0.1673,
+      "step": 17826
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8952574575856083,
+      "learning_rate": 4.546872236167166e-07,
+      "loss": 0.1576,
+      "step": 17827
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0190701158443207,
+      "learning_rate": 4.5419637396796337e-07,
+      "loss": 0.1807,
+      "step": 17828
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8335552710274327,
+      "learning_rate": 4.5370578324967054e-07,
+      "loss": 0.1783,
+      "step": 17829
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3764248852007366,
+      "learning_rate": 4.532154514751497e-07,
+      "loss": 0.1893,
+      "step": 17830
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3915203240070135,
+      "learning_rate": 4.527253786576991e-07,
+      "loss": 0.1592,
+      "step": 17831
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2861389689670186,
+      "learning_rate": 4.5223556481060913e-07,
+      "loss": 0.1528,
+      "step": 17832
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4373456377797067,
+      "learning_rate": 4.5174600994717154e-07,
+      "loss": 0.1645,
+      "step": 17833
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.348075943625048,
+      "learning_rate": 4.5125671408066006e-07,
+      "loss": 0.1546,
+      "step": 17834
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1271648016172928,
+      "learning_rate": 4.5076767722435075e-07,
+      "loss": 0.1644,
+      "step": 17835
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8177597987428121,
+      "learning_rate": 4.502788993915075e-07,
+      "loss": 0.1643,
+      "step": 17836
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4899303323960316,
+      "learning_rate": 4.4979038059538847e-07,
+      "loss": 0.1568,
+      "step": 17837
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8795186133211397,
+      "learning_rate": 4.493021208492443e-07,
+      "loss": 0.1421,
+      "step": 17838
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1307292120596275,
+      "learning_rate": 4.48814120166321e-07,
+      "loss": 0.1725,
+      "step": 17839
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0942327975567745,
+      "learning_rate": 4.483263785598524e-07,
+      "loss": 0.1607,
+      "step": 17840
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.310134964018146,
+      "learning_rate": 4.478388960430724e-07,
+      "loss": 0.1533,
+      "step": 17841
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0896854956976845,
+      "learning_rate": 4.4735167262919934e-07,
+      "loss": 0.1672,
+      "step": 17842
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4009160493602761,
+      "learning_rate": 4.468647083314537e-07,
+      "loss": 0.172,
+      "step": 17843
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0883627609353188,
+      "learning_rate": 4.4637800316304157e-07,
+      "loss": 0.1673,
+      "step": 17844
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9322656297713793,
+      "learning_rate": 4.458915571371647e-07,
+      "loss": 0.1345,
+      "step": 17845
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3953361415792829,
+      "learning_rate": 4.4540537026702026e-07,
+      "loss": 0.1563,
+      "step": 17846
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.582791955807601,
+      "learning_rate": 4.449194425657943e-07,
+      "loss": 0.1769,
+      "step": 17847
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2389198531379395,
+      "learning_rate": 4.4443377404666976e-07,
+      "loss": 0.1726,
+      "step": 17848
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8344787438413215,
+      "learning_rate": 4.439483647228171e-07,
+      "loss": 0.1823,
+      "step": 17849
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9170059286638855,
+      "learning_rate": 4.4346321460740583e-07,
+      "loss": 0.1698,
+      "step": 17850
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0316912778937046,
+      "learning_rate": 4.429783237135932e-07,
+      "loss": 0.17,
+      "step": 17851
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1489235166833727,
+      "learning_rate": 4.424936920545331e-07,
+      "loss": 0.1702,
+      "step": 17852
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0784830220040096,
+      "learning_rate": 4.420093196433717e-07,
+      "loss": 0.1726,
+      "step": 17853
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.780789153470712,
+      "learning_rate": 4.415252064932485e-07,
+      "loss": 0.1517,
+      "step": 17854
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.112357375081568,
+      "learning_rate": 4.4104135261729296e-07,
+      "loss": 0.1584,
+      "step": 17855
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1552800042958693,
+      "learning_rate": 4.4055775802863246e-07,
+      "loss": 0.1663,
+      "step": 17856
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.408008062223532,
+      "learning_rate": 4.400744227403797e-07,
+      "loss": 0.1646,
+      "step": 17857
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9212932843285123,
+      "learning_rate": 4.3959134676565097e-07,
+      "loss": 0.1524,
+      "step": 17858
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9661145345045649,
+      "learning_rate": 4.391085301175457e-07,
+      "loss": 0.1622,
+      "step": 17859
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.189610441208609,
+      "learning_rate": 4.386259728091613e-07,
+      "loss": 0.166,
+      "step": 17860
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2524582875011316,
+      "learning_rate": 4.381436748535872e-07,
+      "loss": 0.1598,
+      "step": 17861
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1669624890146575,
+      "learning_rate": 4.376616362639063e-07,
+      "loss": 0.1482,
+      "step": 17862
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2651938387594743,
+      "learning_rate": 4.371798570531927e-07,
+      "loss": 0.1598,
+      "step": 17863
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.159737621227063,
+      "learning_rate": 4.3669833723451795e-07,
+      "loss": 0.1696,
+      "step": 17864
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.210751056835837,
+      "learning_rate": 4.3621707682094063e-07,
+      "loss": 0.1556,
+      "step": 17865
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0647749145288499,
+      "learning_rate": 4.3573607582551356e-07,
+      "loss": 0.1692,
+      "step": 17866
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9487333310478399,
+      "learning_rate": 4.3525533426128643e-07,
+      "loss": 0.1684,
+      "step": 17867
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0380106070047885,
+      "learning_rate": 4.3477485214129864e-07,
+      "loss": 0.1641,
+      "step": 17868
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.052899447067935,
+      "learning_rate": 4.3429462947858327e-07,
+      "loss": 0.181,
+      "step": 17869
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4736590717566007,
+      "learning_rate": 4.338146662861664e-07,
+      "loss": 0.1609,
+      "step": 17870
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.797747112413201,
+      "learning_rate": 4.333349625770655e-07,
+      "loss": 0.1661,
+      "step": 17871
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2137042272213838,
+      "learning_rate": 4.3285551836429465e-07,
+      "loss": 0.1608,
+      "step": 17872
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8488474336115467,
+      "learning_rate": 4.3237633366085997e-07,
+      "loss": 0.1448,
+      "step": 17873
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0186497884210581,
+      "learning_rate": 4.3189740847975556e-07,
+      "loss": 0.1562,
+      "step": 17874
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0794700833681627,
+      "learning_rate": 4.3141874283397665e-07,
+      "loss": 0.1541,
+      "step": 17875
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3353539984921128,
+      "learning_rate": 4.309403367365028e-07,
+      "loss": 0.1698,
+      "step": 17876
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9808455563771287,
+      "learning_rate": 4.3046219020031366e-07,
+      "loss": 0.1791,
+      "step": 17877
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.460488370062648,
+      "learning_rate": 4.299843032383777e-07,
+      "loss": 0.1983,
+      "step": 17878
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.190477051437423,
+      "learning_rate": 4.295066758636579e-07,
+      "loss": 0.1534,
+      "step": 17879
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.117376348606542,
+      "learning_rate": 4.2902930808910946e-07,
+      "loss": 0.1607,
+      "step": 17880
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1286525714094169,
+      "learning_rate": 4.2855219992768313e-07,
+      "loss": 0.1752,
+      "step": 17881
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0822666747841385,
+      "learning_rate": 4.280753513923197e-07,
+      "loss": 0.1518,
+      "step": 17882
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1124604964562284,
+      "learning_rate": 4.27598762495951e-07,
+      "loss": 0.1501,
+      "step": 17883
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2331043057271551,
+      "learning_rate": 4.271224332515078e-07,
+      "loss": 0.1651,
+      "step": 17884
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1070061068054977,
+      "learning_rate": 4.266463636719087e-07,
+      "loss": 0.1477,
+      "step": 17885
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9162346294809607,
+      "learning_rate": 4.261705537700678e-07,
+      "loss": 0.1644,
+      "step": 17886
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7168818418985248,
+      "learning_rate": 4.2569500355889027e-07,
+      "loss": 0.1631,
+      "step": 17887
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8780179759569882,
+      "learning_rate": 4.2521971305127695e-07,
+      "loss": 0.1571,
+      "step": 17888
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0836510792985983,
+      "learning_rate": 4.2474468226011976e-07,
+      "loss": 0.1682,
+      "step": 17889
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2082434934187478,
+      "learning_rate": 4.242699111983051e-07,
+      "loss": 0.1571,
+      "step": 17890
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8432251155722694,
+      "learning_rate": 4.2379539987870924e-07,
+      "loss": 0.1522,
+      "step": 17891
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.800714309069143,
+      "learning_rate": 4.233211483142041e-07,
+      "loss": 0.155,
+      "step": 17892
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9767747713654685,
+      "learning_rate": 4.2284715651765287e-07,
+      "loss": 0.1674,
+      "step": 17893
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6629670704546073,
+      "learning_rate": 4.223734245019151e-07,
+      "loss": 0.1639,
+      "step": 17894
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2069269896015864,
+      "learning_rate": 4.2189995227983726e-07,
+      "loss": 0.1765,
+      "step": 17895
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.466101959194786,
+      "learning_rate": 4.2142673986426685e-07,
+      "loss": 0.1842,
+      "step": 17896
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0350455168520254,
+      "learning_rate": 4.2095378726803473e-07,
+      "loss": 0.1613,
+      "step": 17897
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7932209596104125,
+      "learning_rate": 4.204810945039717e-07,
+      "loss": 0.1574,
+      "step": 17898
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9474340681428581,
+      "learning_rate": 4.2000866158490084e-07,
+      "loss": 0.16,
+      "step": 17899
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1472259303799852,
+      "learning_rate": 4.195364885236375e-07,
+      "loss": 0.1704,
+      "step": 17900
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0099861991323156,
+      "learning_rate": 4.1906457533298694e-07,
+      "loss": 0.171,
+      "step": 17901
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1539727871270642,
+      "learning_rate": 4.1859292202575007e-07,
+      "loss": 0.1546,
+      "step": 17902
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8311713315990183,
+      "learning_rate": 4.1812152861472333e-07,
+      "loss": 0.1852,
+      "step": 17903
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0243626136619337,
+      "learning_rate": 4.176503951126898e-07,
+      "loss": 0.156,
+      "step": 17904
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9057591631418205,
+      "learning_rate": 4.1717952153243034e-07,
+      "loss": 0.1389,
+      "step": 17905
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.426122726890483,
+      "learning_rate": 4.16708907886717e-07,
+      "loss": 0.1613,
+      "step": 17906
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0026262405109105,
+      "learning_rate": 4.162385541883185e-07,
+      "loss": 0.1697,
+      "step": 17907
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1279028816464756,
+      "learning_rate": 4.157684604499879e-07,
+      "loss": 0.1595,
+      "step": 17908
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.137347084964881,
+      "learning_rate": 4.152986266844805e-07,
+      "loss": 0.1796,
+      "step": 17909
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3469612177043435,
+      "learning_rate": 4.1482905290453846e-07,
+      "loss": 0.1645,
+      "step": 17910
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1427791293368184,
+      "learning_rate": 4.143597391229015e-07,
+      "loss": 0.1642,
+      "step": 17911
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9551977976133195,
+      "learning_rate": 4.1389068535229615e-07,
+      "loss": 0.1625,
+      "step": 17912
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.279745687276524,
+      "learning_rate": 4.1342189160544775e-07,
+      "loss": 0.165,
+      "step": 17913
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9968432466039817,
+      "learning_rate": 4.1295335789507174e-07,
+      "loss": 0.1557,
+      "step": 17914
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8494539809967789,
+      "learning_rate": 4.124850842338779e-07,
+      "loss": 0.1633,
+      "step": 17915
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8921772831807426,
+      "learning_rate": 4.120170706345661e-07,
+      "loss": 0.1419,
+      "step": 17916
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2943009106600474,
+      "learning_rate": 4.1154931710983504e-07,
+      "loss": 0.1617,
+      "step": 17917
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2395827561193538,
+      "learning_rate": 4.1108182367237014e-07,
+      "loss": 0.1823,
+      "step": 17918
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8016794634208525,
+      "learning_rate": 4.106145903348513e-07,
+      "loss": 0.1587,
+      "step": 17919
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1507975908971602,
+      "learning_rate": 4.10147617109955e-07,
+      "loss": 0.1681,
+      "step": 17920
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8995866701271464,
+      "learning_rate": 4.0968090401034444e-07,
+      "loss": 0.1742,
+      "step": 17921
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8769457890812751,
+      "learning_rate": 4.092144510486806e-07,
+      "loss": 0.1565,
+      "step": 17922
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1400777198132501,
+      "learning_rate": 4.0874825823761676e-07,
+      "loss": 0.1481,
+      "step": 17923
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7596705469198951,
+      "learning_rate": 4.0828232558979943e-07,
+      "loss": 0.1577,
+      "step": 17924
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.412245385430175,
+      "learning_rate": 4.078166531178651e-07,
+      "loss": 0.1475,
+      "step": 17925
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9016443987867834,
+      "learning_rate": 4.07351240834446e-07,
+      "loss": 0.1589,
+      "step": 17926
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2437259871228254,
+      "learning_rate": 4.0688608875216527e-07,
+      "loss": 0.1618,
+      "step": 17927
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.763305394131877,
+      "learning_rate": 4.064211968836429e-07,
+      "loss": 0.1706,
+      "step": 17928
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.776341037606745,
+      "learning_rate": 4.059565652414865e-07,
+      "loss": 0.1604,
+      "step": 17929
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1781010449575935,
+      "learning_rate": 4.0549219383830054e-07,
+      "loss": 0.1407,
+      "step": 17930
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9927757811539168,
+      "learning_rate": 4.0502808268668034e-07,
+      "loss": 0.1776,
+      "step": 17931
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9940650804664498,
+      "learning_rate": 4.045642317992149e-07,
+      "loss": 0.1609,
+      "step": 17932
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9952558127873156,
+      "learning_rate": 4.0410064118848624e-07,
+      "loss": 0.1515,
+      "step": 17933
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3742803805870323,
+      "learning_rate": 4.03637310867071e-07,
+      "loss": 0.1851,
+      "step": 17934
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1261120911519835,
+      "learning_rate": 4.031742408475359e-07,
+      "loss": 0.1669,
+      "step": 17935
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.581412240413239,
+      "learning_rate": 4.027114311424407e-07,
+      "loss": 0.1574,
+      "step": 17936
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.590825805117261,
+      "learning_rate": 4.0224888176434105e-07,
+      "loss": 0.1638,
+      "step": 17937
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9072850744524407,
+      "learning_rate": 4.0178659272578026e-07,
+      "loss": 0.1396,
+      "step": 17938
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 44.79995566266486,
+      "learning_rate": 4.0132456403930263e-07,
+      "loss": 0.1802,
+      "step": 17939
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7677360420573207,
+      "learning_rate": 4.0086279571743715e-07,
+      "loss": 0.1641,
+      "step": 17940
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8974806317168946,
+      "learning_rate": 4.004012877727104e-07,
+      "loss": 0.1546,
+      "step": 17941
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9247563788673555,
+      "learning_rate": 3.999400402176401e-07,
+      "loss": 0.1758,
+      "step": 17942
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2695249530965618,
+      "learning_rate": 3.9947905306474077e-07,
+      "loss": 0.1596,
+      "step": 17943
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.874647864267738,
+      "learning_rate": 3.990183263265124e-07,
+      "loss": 0.15,
+      "step": 17944
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1813860921210668,
+      "learning_rate": 3.985578600154549e-07,
+      "loss": 0.1676,
+      "step": 17945
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9040988900659936,
+      "learning_rate": 3.9809765414405734e-07,
+      "loss": 0.1454,
+      "step": 17946
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0329931692731547,
+      "learning_rate": 3.976377087248051e-07,
+      "loss": 0.1761,
+      "step": 17947
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.883762203784139,
+      "learning_rate": 3.9717802377017057e-07,
+      "loss": 0.1717,
+      "step": 17948
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.64805341126127,
+      "learning_rate": 3.967185992926237e-07,
+      "loss": 0.1613,
+      "step": 17949
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.778869015406839,
+      "learning_rate": 3.9625943530462787e-07,
+      "loss": 0.1616,
+      "step": 17950
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8044296441418508,
+      "learning_rate": 3.9580053181863866e-07,
+      "loss": 0.1718,
+      "step": 17951
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1637532712791299,
+      "learning_rate": 3.953418888471017e-07,
+      "loss": 0.1708,
+      "step": 17952
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2436944226653273,
+      "learning_rate": 3.948835064024581e-07,
+      "loss": 0.1576,
+      "step": 17953
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.16526928391713,
+      "learning_rate": 3.944253844971435e-07,
+      "loss": 0.1609,
+      "step": 17954
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1371984003091298,
+      "learning_rate": 3.939675231435802e-07,
+      "loss": 0.1489,
+      "step": 17955
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0432445408536668,
+      "learning_rate": 3.935099223541927e-07,
+      "loss": 0.1471,
+      "step": 17956
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.939698839652885,
+      "learning_rate": 3.9305258214138995e-07,
+      "loss": 0.1406,
+      "step": 17957
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0156235305520485,
+      "learning_rate": 3.9259550251757763e-07,
+      "loss": 0.1752,
+      "step": 17958
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.56236736930776,
+      "learning_rate": 3.921386834951557e-07,
+      "loss": 0.1594,
+      "step": 17959
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9256258931900568,
+      "learning_rate": 3.9168212508651547e-07,
+      "loss": 0.158,
+      "step": 17960
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2545803271356366,
+      "learning_rate": 3.9122582730403924e-07,
+      "loss": 0.1604,
+      "step": 17961
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2502358925172403,
+      "learning_rate": 3.907697901601071e-07,
+      "loss": 0.1283,
+      "step": 17962
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0882395503948148,
+      "learning_rate": 3.9031401366708467e-07,
+      "loss": 0.1499,
+      "step": 17963
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4270571392860416,
+      "learning_rate": 3.8985849783733873e-07,
+      "loss": 0.1528,
+      "step": 17964
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9151584518670104,
+      "learning_rate": 3.8940324268322285e-07,
+      "loss": 0.1507,
+      "step": 17965
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9494614142440932,
+      "learning_rate": 3.889482482170881e-07,
+      "loss": 0.1779,
+      "step": 17966
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.311422455794517,
+      "learning_rate": 3.884935144512747e-07,
+      "loss": 0.1668,
+      "step": 17967
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9151959255779871,
+      "learning_rate": 3.880390413981161e-07,
+      "loss": 0.1457,
+      "step": 17968
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1434615925984102,
+      "learning_rate": 3.8758482906994245e-07,
+      "loss": 0.1798,
+      "step": 17969
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8407924412287494,
+      "learning_rate": 3.8713087747907385e-07,
+      "loss": 0.1365,
+      "step": 17970
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0285526948328574,
+      "learning_rate": 3.866771866378227e-07,
+      "loss": 0.156,
+      "step": 17971
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.141798461154904,
+      "learning_rate": 3.862237565584959e-07,
+      "loss": 0.182,
+      "step": 17972
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9341811901107416,
+      "learning_rate": 3.8577058725339235e-07,
+      "loss": 0.157,
+      "step": 17973
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.227656033497009,
+      "learning_rate": 3.8531767873480453e-07,
+      "loss": 0.162,
+      "step": 17974
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.395502251060028,
+      "learning_rate": 3.8486503101501705e-07,
+      "loss": 0.1852,
+      "step": 17975
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1583305556455232,
+      "learning_rate": 3.84412644106309e-07,
+      "loss": 0.1536,
+      "step": 17976
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3079551443868314,
+      "learning_rate": 3.839605180209527e-07,
+      "loss": 0.1573,
+      "step": 17977
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8749974615321174,
+      "learning_rate": 3.835086527712084e-07,
+      "loss": 0.1618,
+      "step": 17978
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.412753912406829,
+      "learning_rate": 3.830570483693374e-07,
+      "loss": 0.1508,
+      "step": 17979
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9967886468592525,
+      "learning_rate": 3.8260570482758554e-07,
+      "loss": 0.1565,
+      "step": 17980
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3170670252562915,
+      "learning_rate": 3.8215462215819733e-07,
+      "loss": 0.1727,
+      "step": 17981
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8589125439803201,
+      "learning_rate": 3.817038003734075e-07,
+      "loss": 0.1716,
+      "step": 17982
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.77991289215004,
+      "learning_rate": 3.8125323948544734e-07,
+      "loss": 0.1717,
+      "step": 17983
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4852015113231571,
+      "learning_rate": 3.808029395065349e-07,
+      "loss": 0.1668,
+      "step": 17984
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3957447592536913,
+      "learning_rate": 3.803529004488848e-07,
+      "loss": 0.1621,
+      "step": 17985
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0856403131783523,
+      "learning_rate": 3.7990312232470627e-07,
+      "loss": 0.1625,
+      "step": 17986
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8855645448579809,
+      "learning_rate": 3.7945360514620056e-07,
+      "loss": 0.1342,
+      "step": 17987
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0073013102385229,
+      "learning_rate": 3.7900434892555903e-07,
+      "loss": 0.1579,
+      "step": 17988
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0871019309171697,
+      "learning_rate": 3.785553536749664e-07,
+      "loss": 0.1583,
+      "step": 17989
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0891511728916528,
+      "learning_rate": 3.781066194066052e-07,
+      "loss": 0.1642,
+      "step": 17990
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.8153204240961576,
+      "learning_rate": 3.776581461326434e-07,
+      "loss": 0.1855,
+      "step": 17991
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0753731363296326,
+      "learning_rate": 3.772099338652491e-07,
+      "loss": 0.1686,
+      "step": 17992
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1277177371834488,
+      "learning_rate": 3.7676198261657803e-07,
+      "loss": 0.1571,
+      "step": 17993
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8328991730043827,
+      "learning_rate": 3.763142923987817e-07,
+      "loss": 0.1635,
+      "step": 17994
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1623322953496837,
+      "learning_rate": 3.7586686322400257e-07,
+      "loss": 0.1753,
+      "step": 17995
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2256700386655874,
+      "learning_rate": 3.7541969510438094e-07,
+      "loss": 0.1708,
+      "step": 17996
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9689492752414466,
+      "learning_rate": 3.749727880520415e-07,
+      "loss": 0.1555,
+      "step": 17997
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2295248073265919,
+      "learning_rate": 3.7452614207911133e-07,
+      "loss": 0.1701,
+      "step": 17998
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5945159084207308,
+      "learning_rate": 3.740797571977006e-07,
+      "loss": 0.1777,
+      "step": 17999
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.097649874716848,
+      "learning_rate": 3.7363363341992197e-07,
+      "loss": 0.1689,
+      "step": 18000
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8499978036235123,
+      "learning_rate": 3.731877707578735e-07,
+      "loss": 0.1546,
+      "step": 18001
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8260381407155,
+      "learning_rate": 3.72742169223651e-07,
+      "loss": 0.1513,
+      "step": 18002
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7594243057445846,
+      "learning_rate": 3.722968288293405e-07,
+      "loss": 0.137,
+      "step": 18003
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9524979700874181,
+      "learning_rate": 3.718517495870233e-07,
+      "loss": 0.173,
+      "step": 18004
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0215817900375197,
+      "learning_rate": 3.714069315087709e-07,
+      "loss": 0.1522,
+      "step": 18005
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2227474046025721,
+      "learning_rate": 3.709623746066482e-07,
+      "loss": 0.1764,
+      "step": 18006
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4786570951548887,
+      "learning_rate": 3.7051807889271653e-07,
+      "loss": 0.1805,
+      "step": 18007
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9627071613835644,
+      "learning_rate": 3.7007404437902515e-07,
+      "loss": 0.1833,
+      "step": 18008
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9180815857025116,
+      "learning_rate": 3.6963027107761896e-07,
+      "loss": 0.1715,
+      "step": 18009
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.144345240438523,
+      "learning_rate": 3.6918675900053605e-07,
+      "loss": 0.1531,
+      "step": 18010
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4240528022562584,
+      "learning_rate": 3.6874350815980565e-07,
+      "loss": 0.1887,
+      "step": 18011
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9857446024845689,
+      "learning_rate": 3.683005185674504e-07,
+      "loss": 0.1835,
+      "step": 18012
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8481856015382779,
+      "learning_rate": 3.678577902354907e-07,
+      "loss": 0.1537,
+      "step": 18013
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9481275747482545,
+      "learning_rate": 3.674153231759303e-07,
+      "loss": 0.159,
+      "step": 18014
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0549232250539309,
+      "learning_rate": 3.66973117400774e-07,
+      "loss": 0.1911,
+      "step": 18015
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1714198995947855,
+      "learning_rate": 3.665311729220156e-07,
+      "loss": 0.1681,
+      "step": 18016
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.204461169805127,
+      "learning_rate": 3.6608948975164424e-07,
+      "loss": 0.1846,
+      "step": 18017
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0571081817276509,
+      "learning_rate": 3.6564806790163833e-07,
+      "loss": 0.1925,
+      "step": 18018
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9719685244560988,
+      "learning_rate": 3.6520690738397256e-07,
+      "loss": 0.1646,
+      "step": 18019
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9468764412733461,
+      "learning_rate": 3.64766008210613e-07,
+      "loss": 0.1658,
+      "step": 18020
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7563243675377808,
+      "learning_rate": 3.643253703935223e-07,
+      "loss": 0.1605,
+      "step": 18021
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9060599741346881,
+      "learning_rate": 3.6388499394464874e-07,
+      "loss": 0.1464,
+      "step": 18022
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2158613459959573,
+      "learning_rate": 3.6344487887593926e-07,
+      "loss": 0.17,
+      "step": 18023
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.971532594962186,
+      "learning_rate": 3.630050251993311e-07,
+      "loss": 0.16,
+      "step": 18024
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1481698919830072,
+      "learning_rate": 3.6256543292675584e-07,
+      "loss": 0.1777,
+      "step": 18025
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.486437555313577,
+      "learning_rate": 3.6212610207013943e-07,
+      "loss": 0.1588,
+      "step": 18026
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1882199195943515,
+      "learning_rate": 3.616870326413946e-07,
+      "loss": 0.1843,
+      "step": 18027
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8975928779369106,
+      "learning_rate": 3.61248224652434e-07,
+      "loss": 0.1493,
+      "step": 18028
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0875976408022372,
+      "learning_rate": 3.6080967811515933e-07,
+      "loss": 0.16,
+      "step": 18029
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9381668102488647,
+      "learning_rate": 3.603713930414676e-07,
+      "loss": 0.1578,
+      "step": 18030
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8959578955261757,
+      "learning_rate": 3.59933369443245e-07,
+      "loss": 0.158,
+      "step": 18031
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7477054698311623,
+      "learning_rate": 3.594956073323763e-07,
+      "loss": 0.1612,
+      "step": 18032
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0899173714504182,
+      "learning_rate": 3.5905810672073107e-07,
+      "loss": 0.1803,
+      "step": 18033
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5364206457954415,
+      "learning_rate": 3.586208676201819e-07,
+      "loss": 0.1786,
+      "step": 18034
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.070889298110779,
+      "learning_rate": 3.581838900425838e-07,
+      "loss": 0.1661,
+      "step": 18035
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8732342679329591,
+      "learning_rate": 3.5774717399979396e-07,
+      "loss": 0.1585,
+      "step": 18036
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9400795374448295,
+      "learning_rate": 3.5731071950365625e-07,
+      "loss": 0.1693,
+      "step": 18037
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1803725967238192,
+      "learning_rate": 3.5687452656600896e-07,
+      "loss": 0.1711,
+      "step": 18038
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.16940189288093,
+      "learning_rate": 3.5643859519868594e-07,
+      "loss": 0.1701,
+      "step": 18039
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8894016048155894,
+      "learning_rate": 3.5600292541351e-07,
+      "loss": 0.1652,
+      "step": 18040
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.994107188440856,
+      "learning_rate": 3.5556751722230056e-07,
+      "loss": 0.143,
+      "step": 18041
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8943863074222658,
+      "learning_rate": 3.551323706368659e-07,
+      "loss": 0.1604,
+      "step": 18042
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8372530214153084,
+      "learning_rate": 3.546974856690111e-07,
+      "loss": 0.1682,
+      "step": 18043
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4334941554546106,
+      "learning_rate": 3.542628623305311e-07,
+      "loss": 0.1692,
+      "step": 18044
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8602732086285463,
+      "learning_rate": 3.538285006332154e-07,
+      "loss": 0.1651,
+      "step": 18045
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.27004473331578,
+      "learning_rate": 3.5339440058884565e-07,
+      "loss": 0.1518,
+      "step": 18046
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2627309938058453,
+      "learning_rate": 3.529605622092003e-07,
+      "loss": 0.1388,
+      "step": 18047
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0584636145518418,
+      "learning_rate": 3.52526985506042e-07,
+      "loss": 0.1625,
+      "step": 18048
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8819800807661746,
+      "learning_rate": 3.5209367049113596e-07,
+      "loss": 0.1535,
+      "step": 18049
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1435665267192858,
+      "learning_rate": 3.516606171762338e-07,
+      "loss": 0.1647,
+      "step": 18050
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.00416429595634,
+      "learning_rate": 3.5122782557308163e-07,
+      "loss": 0.1704,
+      "step": 18051
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.040981168819708,
+      "learning_rate": 3.507952956934202e-07,
+      "loss": 0.1615,
+      "step": 18052
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1564463804169962,
+      "learning_rate": 3.503630275489811e-07,
+      "loss": 0.1751,
+      "step": 18053
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0239001745339626,
+      "learning_rate": 3.4993102115148947e-07,
+      "loss": 0.1444,
+      "step": 18054
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6025551048585556,
+      "learning_rate": 3.494992765126637e-07,
+      "loss": 0.1668,
+      "step": 18055
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0701646834638574,
+      "learning_rate": 3.490677936442155e-07,
+      "loss": 0.1418,
+      "step": 18056
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3906487577365694,
+      "learning_rate": 3.4863657255785e-07,
+      "loss": 0.1819,
+      "step": 18057
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2564578605249104,
+      "learning_rate": 3.482056132652623e-07,
+      "loss": 0.1609,
+      "step": 18058
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2013413310258307,
+      "learning_rate": 3.477749157781407e-07,
+      "loss": 0.1592,
+      "step": 18059
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8924948411927334,
+      "learning_rate": 3.4734448010817043e-07,
+      "loss": 0.1743,
+      "step": 18060
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0088436986399607,
+      "learning_rate": 3.469143062670266e-07,
+      "loss": 0.1438,
+      "step": 18061
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1062930156517436,
+      "learning_rate": 3.464843942663776e-07,
+      "loss": 0.1498,
+      "step": 18062
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2243603416203355,
+      "learning_rate": 3.4605474411788407e-07,
+      "loss": 0.1835,
+      "step": 18063
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1303646982159368,
+      "learning_rate": 3.456253558332001e-07,
+      "loss": 0.1652,
+      "step": 18064
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.644905305055982,
+      "learning_rate": 3.451962294239741e-07,
+      "loss": 0.1728,
+      "step": 18065
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4259200726169508,
+      "learning_rate": 3.4476736490184683e-07,
+      "loss": 0.1637,
+      "step": 18066
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7865406720386555,
+      "learning_rate": 3.443387622784489e-07,
+      "loss": 0.1584,
+      "step": 18067
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7506526478559534,
+      "learning_rate": 3.439104215654088e-07,
+      "loss": 0.1562,
+      "step": 18068
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9513526709479354,
+      "learning_rate": 3.434823427743428e-07,
+      "loss": 0.1656,
+      "step": 18069
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9825643516548362,
+      "learning_rate": 3.430545259168638e-07,
+      "loss": 0.1626,
+      "step": 18070
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8435459556303768,
+      "learning_rate": 3.42626971004576e-07,
+      "loss": 0.1509,
+      "step": 18071
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1602326205822766,
+      "learning_rate": 3.421996780490766e-07,
+      "loss": 0.1743,
+      "step": 18072
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.240986025282152,
+      "learning_rate": 3.4177264706195754e-07,
+      "loss": 0.1743,
+      "step": 18073
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9798339208614196,
+      "learning_rate": 3.413458780548007e-07,
+      "loss": 0.1678,
+      "step": 18074
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1122500459593923,
+      "learning_rate": 3.409193710391834e-07,
+      "loss": 0.1846,
+      "step": 18075
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2219867673175016,
+      "learning_rate": 3.4049312602667197e-07,
+      "loss": 0.1601,
+      "step": 18076
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1584272896050063,
+      "learning_rate": 3.400671430288316e-07,
+      "loss": 0.1418,
+      "step": 18077
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.934316331610779,
+      "learning_rate": 3.396414220572142e-07,
+      "loss": 0.1549,
+      "step": 18078
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3242111151036162,
+      "learning_rate": 3.3921596312336935e-07,
+      "loss": 0.1588,
+      "step": 18079
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9434150029025636,
+      "learning_rate": 3.3879076623883677e-07,
+      "loss": 0.1556,
+      "step": 18080
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9255895715291566,
+      "learning_rate": 3.3836583141515054e-07,
+      "loss": 0.1705,
+      "step": 18081
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3659204139409478,
+      "learning_rate": 3.379411586638359e-07,
+      "loss": 0.1644,
+      "step": 18082
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2360570568932796,
+      "learning_rate": 3.3751674799641475e-07,
+      "loss": 0.1591,
+      "step": 18083
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1750311349980966,
+      "learning_rate": 3.3709259942439677e-07,
+      "loss": 0.1692,
+      "step": 18084
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2854669834238228,
+      "learning_rate": 3.3666871295928826e-07,
+      "loss": 0.168,
+      "step": 18085
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2254877804588578,
+      "learning_rate": 3.3624508861258564e-07,
+      "loss": 0.1592,
+      "step": 18086
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4678012533558435,
+      "learning_rate": 3.3582172639578304e-07,
+      "loss": 0.1746,
+      "step": 18087
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9622118655252306,
+      "learning_rate": 3.35398626320359e-07,
+      "loss": 0.1668,
+      "step": 18088
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.467378103630613,
+      "learning_rate": 3.3497578839779554e-07,
+      "loss": 0.166,
+      "step": 18089
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0194618001770677,
+      "learning_rate": 3.345532126395579e-07,
+      "loss": 0.1638,
+      "step": 18090
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.170019513023722,
+      "learning_rate": 3.3413089905711127e-07,
+      "loss": 0.1381,
+      "step": 18091
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.936399762071647,
+      "learning_rate": 3.337088476619099e-07,
+      "loss": 0.1745,
+      "step": 18092
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9416114993211048,
+      "learning_rate": 3.332870584654013e-07,
+      "loss": 0.1697,
+      "step": 18093
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.979094340683368,
+      "learning_rate": 3.3286553147902855e-07,
+      "loss": 0.1802,
+      "step": 18094
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9993114947029553,
+      "learning_rate": 3.3244426671422246e-07,
+      "loss": 0.1716,
+      "step": 18095
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9762340459586974,
+      "learning_rate": 3.320232641824139e-07,
+      "loss": 0.1665,
+      "step": 18096
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0886486279912961,
+      "learning_rate": 3.3160252389501824e-07,
+      "loss": 0.1652,
+      "step": 18097
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2737911651638636,
+      "learning_rate": 3.311820458634507e-07,
+      "loss": 0.1583,
+      "step": 18098
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9989512775252093,
+      "learning_rate": 3.3076183009911667e-07,
+      "loss": 0.1443,
+      "step": 18099
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9043083457326804,
+      "learning_rate": 3.3034187661341476e-07,
+      "loss": 0.1616,
+      "step": 18100
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.511213481907978,
+      "learning_rate": 3.299221854177337e-07,
+      "loss": 0.1606,
+      "step": 18101
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8975862189129877,
+      "learning_rate": 3.29502756523461e-07,
+      "loss": 0.1452,
+      "step": 18102
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9606207256115495,
+      "learning_rate": 3.29083589941972e-07,
+      "loss": 0.1704,
+      "step": 18103
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9537946488192327,
+      "learning_rate": 3.286646856846376e-07,
+      "loss": 0.1623,
+      "step": 18104
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.274674491839072,
+      "learning_rate": 3.2824604376281874e-07,
+      "loss": 0.166,
+      "step": 18105
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.082756311497904,
+      "learning_rate": 3.2782766418787306e-07,
+      "loss": 0.1535,
+      "step": 18106
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2348526149655745,
+      "learning_rate": 3.274095469711469e-07,
+      "loss": 0.1637,
+      "step": 18107
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1677907803256713,
+      "learning_rate": 3.2699169212398354e-07,
+      "loss": 0.1689,
+      "step": 18108
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.032761047187609,
+      "learning_rate": 3.2657409965771715e-07,
+      "loss": 0.1581,
+      "step": 18109
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5161936216848935,
+      "learning_rate": 3.2615676958367424e-07,
+      "loss": 0.1738,
+      "step": 18110
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8366912376904033,
+      "learning_rate": 3.2573970191317693e-07,
+      "loss": 0.1794,
+      "step": 18111
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8885962028795105,
+      "learning_rate": 3.2532289665753503e-07,
+      "loss": 0.1598,
+      "step": 18112
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7663929172818378,
+      "learning_rate": 3.2490635382805726e-07,
+      "loss": 0.156,
+      "step": 18113
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9646336089864638,
+      "learning_rate": 3.244900734360401e-07,
+      "loss": 0.169,
+      "step": 18114
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0913680494127547,
+      "learning_rate": 3.2407405549277683e-07,
+      "loss": 0.1544,
+      "step": 18115
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.339377476577004,
+      "learning_rate": 3.2365830000954945e-07,
+      "loss": 0.1701,
+      "step": 18116
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1835551634972878,
+      "learning_rate": 3.23242806997639e-07,
+      "loss": 0.1798,
+      "step": 18117
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7438641443938219,
+      "learning_rate": 3.2282757646831306e-07,
+      "loss": 0.1447,
+      "step": 18118
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2541257573027336,
+      "learning_rate": 3.224126084328361e-07,
+      "loss": 0.1617,
+      "step": 18119
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.298017080325946,
+      "learning_rate": 3.219979029024634e-07,
+      "loss": 0.1719,
+      "step": 18120
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0833901419058496,
+      "learning_rate": 3.21583459888446e-07,
+      "loss": 0.1504,
+      "step": 18121
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1311696910451334,
+      "learning_rate": 3.2116927940202157e-07,
+      "loss": 0.1725,
+      "step": 18122
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4654054266531409,
+      "learning_rate": 3.2075536145442897e-07,
+      "loss": 0.1606,
+      "step": 18123
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.540578843845565,
+      "learning_rate": 3.203417060568925e-07,
+      "loss": 0.1611,
+      "step": 18124
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5544623075810544,
+      "learning_rate": 3.1992831322063324e-07,
+      "loss": 0.1667,
+      "step": 18125
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9578152551564797,
+      "learning_rate": 3.195151829568666e-07,
+      "loss": 0.1612,
+      "step": 18126
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9665434629439523,
+      "learning_rate": 3.191023152767969e-07,
+      "loss": 0.1497,
+      "step": 18127
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.009141454543222,
+      "learning_rate": 3.1868971019162533e-07,
+      "loss": 0.1609,
+      "step": 18128
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0150188637483344,
+      "learning_rate": 3.1827736771253946e-07,
+      "loss": 0.1751,
+      "step": 18129
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8852232068809844,
+      "learning_rate": 3.178652878507293e-07,
+      "loss": 0.1538,
+      "step": 18130
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8192750933961805,
+      "learning_rate": 3.174534706173682e-07,
+      "loss": 0.1602,
+      "step": 18131
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8364749466835223,
+      "learning_rate": 3.170419160236293e-07,
+      "loss": 0.1617,
+      "step": 18132
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0486731726322975,
+      "learning_rate": 3.166306240806749e-07,
+      "loss": 0.1673,
+      "step": 18133
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.994510073688784,
+      "learning_rate": 3.162195947996616e-07,
+      "loss": 0.1523,
+      "step": 18134
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2142922058584487,
+      "learning_rate": 3.158088281917393e-07,
+      "loss": 0.1691,
+      "step": 18135
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0238164627534345,
+      "learning_rate": 3.153983242680503e-07,
+      "loss": 0.1507,
+      "step": 18136
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0896154357419239,
+      "learning_rate": 3.149880830397267e-07,
+      "loss": 0.1846,
+      "step": 18137
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0416929788195814,
+      "learning_rate": 3.1457810451790083e-07,
+      "loss": 0.1621,
+      "step": 18138
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2883004776773355,
+      "learning_rate": 3.1416838871368925e-07,
+      "loss": 0.19,
+      "step": 18139
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8616914377273562,
+      "learning_rate": 3.137589356382076e-07,
+      "loss": 0.1522,
+      "step": 18140
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8543744553081801,
+      "learning_rate": 3.1334974530256134e-07,
+      "loss": 0.1656,
+      "step": 18141
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1054493297414734,
+      "learning_rate": 3.1294081771785057e-07,
+      "loss": 0.1475,
+      "step": 18142
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8807211474880493,
+      "learning_rate": 3.125321528951675e-07,
+      "loss": 0.1262,
+      "step": 18143
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9575123181423777,
+      "learning_rate": 3.1212375084559767e-07,
+      "loss": 0.1529,
+      "step": 18144
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1843135991699991,
+      "learning_rate": 3.117156115802178e-07,
+      "loss": 0.1592,
+      "step": 18145
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5669667485440848,
+      "learning_rate": 3.113077351100979e-07,
+      "loss": 0.1494,
+      "step": 18146
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9169278577698446,
+      "learning_rate": 3.1090012144630476e-07,
+      "loss": 0.1438,
+      "step": 18147
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.554679031235613,
+      "learning_rate": 3.1049277059989167e-07,
+      "loss": 0.1633,
+      "step": 18148
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2194988964301239,
+      "learning_rate": 3.1008568258191095e-07,
+      "loss": 0.1622,
+      "step": 18149
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0482947007736183,
+      "learning_rate": 3.0967885740340266e-07,
+      "loss": 0.1358,
+      "step": 18150
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8393879878929953,
+      "learning_rate": 3.0927229507540126e-07,
+      "loss": 0.1526,
+      "step": 18151
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1180630860302485,
+      "learning_rate": 3.088659956089368e-07,
+      "loss": 0.1662,
+      "step": 18152
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0517176769625036,
+      "learning_rate": 3.0845995901503167e-07,
+      "loss": 0.1766,
+      "step": 18153
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3653458310572448,
+      "learning_rate": 3.080541853046948e-07,
+      "loss": 0.1944,
+      "step": 18154
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0415239876959776,
+      "learning_rate": 3.076486744889373e-07,
+      "loss": 0.1497,
+      "step": 18155
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.24217910381336,
+      "learning_rate": 3.0724342657875604e-07,
+      "loss": 0.1783,
+      "step": 18156
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.07365429308084,
+      "learning_rate": 3.068384415851455e-07,
+      "loss": 0.1538,
+      "step": 18157
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9870483996894945,
+      "learning_rate": 3.0643371951908806e-07,
+      "loss": 0.1896,
+      "step": 18158
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9650764946172202,
+      "learning_rate": 3.0602926039156487e-07,
+      "loss": 0.1665,
+      "step": 18159
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9600897084297326,
+      "learning_rate": 3.05625064213545e-07,
+      "loss": 0.1598,
+      "step": 18160
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8412084312803206,
+      "learning_rate": 3.0522113099599184e-07,
+      "loss": 0.1736,
+      "step": 18161
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.050746648506869,
+      "learning_rate": 3.048174607498644e-07,
+      "loss": 0.186,
+      "step": 18162
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1393671291834178,
+      "learning_rate": 3.044140534861106e-07,
+      "loss": 0.1768,
+      "step": 18163
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2581635137906724,
+      "learning_rate": 3.04010909215674e-07,
+      "loss": 0.1781,
+      "step": 18164
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.069811077440688,
+      "learning_rate": 3.0360802794948687e-07,
+      "loss": 0.1713,
+      "step": 18165
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8545719708007923,
+      "learning_rate": 3.032054096984816e-07,
+      "loss": 0.1502,
+      "step": 18166
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5327808645021874,
+      "learning_rate": 3.0280305447357607e-07,
+      "loss": 0.1389,
+      "step": 18167
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9530728622555193,
+      "learning_rate": 3.0240096228568606e-07,
+      "loss": 0.1396,
+      "step": 18168
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2315279397059917,
+      "learning_rate": 3.0199913314571726e-07,
+      "loss": 0.1459,
+      "step": 18169
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8872656692354135,
+      "learning_rate": 3.0159756706456987e-07,
+      "loss": 0.1484,
+      "step": 18170
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8181420046073249,
+      "learning_rate": 3.011962640531363e-07,
+      "loss": 0.1527,
+      "step": 18171
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3518825408066595,
+      "learning_rate": 3.007952241223022e-07,
+      "loss": 0.1806,
+      "step": 18172
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4158408673441485,
+      "learning_rate": 3.0039444728294563e-07,
+      "loss": 0.148,
+      "step": 18173
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1370683227512903,
+      "learning_rate": 2.999939335459379e-07,
+      "loss": 0.1633,
+      "step": 18174
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1368399110544725,
+      "learning_rate": 2.995936829221413e-07,
+      "loss": 0.1402,
+      "step": 18175
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9983565029532296,
+      "learning_rate": 2.9919369542241504e-07,
+      "loss": 0.1586,
+      "step": 18176
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8587644416400668,
+      "learning_rate": 2.9879397105760597e-07,
+      "loss": 0.1584,
+      "step": 18177
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1468541444837406,
+      "learning_rate": 2.9839450983855876e-07,
+      "loss": 0.1645,
+      "step": 18178
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8486483622789254,
+      "learning_rate": 2.979953117761103e-07,
+      "loss": 0.2073,
+      "step": 18179
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8125393213889613,
+      "learning_rate": 2.975963768810852e-07,
+      "loss": 0.1432,
+      "step": 18180
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3930677782012002,
+      "learning_rate": 2.971977051643071e-07,
+      "loss": 0.1656,
+      "step": 18181
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7007796046851476,
+      "learning_rate": 2.9679929663658957e-07,
+      "loss": 0.1799,
+      "step": 18182
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9477805648312042,
+      "learning_rate": 2.9640115130873835e-07,
+      "loss": 0.1565,
+      "step": 18183
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1168235302246006,
+      "learning_rate": 2.9600326919155486e-07,
+      "loss": 0.168,
+      "step": 18184
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.859315154971044,
+      "learning_rate": 2.956056502958304e-07,
+      "loss": 0.1585,
+      "step": 18185
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5752916829869035,
+      "learning_rate": 2.952082946323498e-07,
+      "loss": 0.1666,
+      "step": 18186
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1845490612936995,
+      "learning_rate": 2.948112022118932e-07,
+      "loss": 0.1917,
+      "step": 18187
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7967161879916749,
+      "learning_rate": 2.944143730452298e-07,
+      "loss": 0.1686,
+      "step": 18188
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0093817119350061,
+      "learning_rate": 2.9401780714312657e-07,
+      "loss": 0.168,
+      "step": 18189
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8868337153966859,
+      "learning_rate": 2.936215045163371e-07,
+      "loss": 0.1441,
+      "step": 18190
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0742994689211378,
+      "learning_rate": 2.932254651756139e-07,
+      "loss": 0.1519,
+      "step": 18191
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9396325141785344,
+      "learning_rate": 2.928296891316973e-07,
+      "loss": 0.1725,
+      "step": 18192
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1127639369893016,
+      "learning_rate": 2.9243417639532424e-07,
+      "loss": 0.1516,
+      "step": 18193
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0071558709655952,
+      "learning_rate": 2.920389269772217e-07,
+      "loss": 0.1427,
+      "step": 18194
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3870759237473882,
+      "learning_rate": 2.916439408881111e-07,
+      "loss": 0.1573,
+      "step": 18195
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2161075701965,
+      "learning_rate": 2.912492181387072e-07,
+      "loss": 0.1739,
+      "step": 18196
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1769855123428175,
+      "learning_rate": 2.9085475873971815e-07,
+      "loss": 0.1627,
+      "step": 18197
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.077500567192318,
+      "learning_rate": 2.9046056270184197e-07,
+      "loss": 0.1619,
+      "step": 18198
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.067921049897875,
+      "learning_rate": 2.9006663003576904e-07,
+      "loss": 0.1553,
+      "step": 18199
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0080994104165948,
+      "learning_rate": 2.896729607521898e-07,
+      "loss": 0.1739,
+      "step": 18200
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.031766215965229,
+      "learning_rate": 2.892795548617788e-07,
+      "loss": 0.1758,
+      "step": 18201
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2709143832706076,
+      "learning_rate": 2.8888641237520886e-07,
+      "loss": 0.1829,
+      "step": 18202
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7373246872216361,
+      "learning_rate": 2.8849353330314247e-07,
+      "loss": 0.1482,
+      "step": 18203
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6141345512194505,
+      "learning_rate": 2.881009176562377e-07,
+      "loss": 0.1835,
+      "step": 18204
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8910948700271394,
+      "learning_rate": 2.8770856544514393e-07,
+      "loss": 0.1536,
+      "step": 18205
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0048942029409267,
+      "learning_rate": 2.8731647668050477e-07,
+      "loss": 0.1706,
+      "step": 18206
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2715211457408504,
+      "learning_rate": 2.86924651372954e-07,
+      "loss": 0.1536,
+      "step": 18207
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.185487138272821,
+      "learning_rate": 2.865330895331209e-07,
+      "loss": 0.1677,
+      "step": 18208
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8032552185841167,
+      "learning_rate": 2.861417911716269e-07,
+      "loss": 0.1383,
+      "step": 18209
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.989140575419518,
+      "learning_rate": 2.8575075629908465e-07,
+      "loss": 0.1528,
+      "step": 18210
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3954966524477264,
+      "learning_rate": 2.853599849261024e-07,
+      "loss": 0.1658,
+      "step": 18211
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9393271717300197,
+      "learning_rate": 2.849694770632794e-07,
+      "loss": 0.1639,
+      "step": 18212
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8570807725947537,
+      "learning_rate": 2.8457923272120715e-07,
+      "loss": 0.157,
+      "step": 18213
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9916469584513358,
+      "learning_rate": 2.8418925191047163e-07,
+      "loss": 0.1492,
+      "step": 18214
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.875743607965698,
+      "learning_rate": 2.8379953464165334e-07,
+      "loss": 0.1633,
+      "step": 18215
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0349400008204857,
+      "learning_rate": 2.8341008092531927e-07,
+      "loss": 0.1557,
+      "step": 18216
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6943081379529428,
+      "learning_rate": 2.8302089077203776e-07,
+      "loss": 0.1753,
+      "step": 18217
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3400755030544176,
+      "learning_rate": 2.826319641923614e-07,
+      "loss": 0.1639,
+      "step": 18218
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.458439540499437,
+      "learning_rate": 2.8224330119684286e-07,
+      "loss": 0.1778,
+      "step": 18219
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0458337430912972,
+      "learning_rate": 2.818549017960237e-07,
+      "loss": 0.1697,
+      "step": 18220
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1383391293406424,
+      "learning_rate": 2.8146676600043777e-07,
+      "loss": 0.1956,
+      "step": 18221
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1962578304457712,
+      "learning_rate": 2.810788938206155e-07,
+      "loss": 0.1767,
+      "step": 18222
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9702710290737139,
+      "learning_rate": 2.8069128526707845e-07,
+      "loss": 0.178,
+      "step": 18223
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9376196152192133,
+      "learning_rate": 2.8030394035033827e-07,
+      "loss": 0.1586,
+      "step": 18224
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.05253853210062,
+      "learning_rate": 2.7991685908090316e-07,
+      "loss": 0.16,
+      "step": 18225
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.939644911654303,
+      "learning_rate": 2.7953004146927145e-07,
+      "loss": 0.1365,
+      "step": 18226
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1506454940271946,
+      "learning_rate": 2.791434875259369e-07,
+      "loss": 0.1536,
+      "step": 18227
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8693278008577767,
+      "learning_rate": 2.787571972613845e-07,
+      "loss": 0.1691,
+      "step": 18228
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8956275554521052,
+      "learning_rate": 2.7837117068609254e-07,
+      "loss": 0.1611,
+      "step": 18229
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3026610082054975,
+      "learning_rate": 2.779854078105304e-07,
+      "loss": 0.1835,
+      "step": 18230
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8988603613018683,
+      "learning_rate": 2.7759990864516306e-07,
+      "loss": 0.1412,
+      "step": 18231
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4338602081572938,
+      "learning_rate": 2.772146732004488e-07,
+      "loss": 0.1496,
+      "step": 18232
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7382306793567448,
+      "learning_rate": 2.768297014868337e-07,
+      "loss": 0.1581,
+      "step": 18233
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9871482240366329,
+      "learning_rate": 2.7644499351476396e-07,
+      "loss": 0.151,
+      "step": 18234
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4205508635886646,
+      "learning_rate": 2.760605492946722e-07,
+      "loss": 0.1774,
+      "step": 18235
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9918044669951058,
+      "learning_rate": 2.756763688369879e-07,
+      "loss": 0.176,
+      "step": 18236
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.07450755147161,
+      "learning_rate": 2.7529245215213053e-07,
+      "loss": 0.1628,
+      "step": 18237
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0434581153695943,
+      "learning_rate": 2.7490879925051397e-07,
+      "loss": 0.1525,
+      "step": 18238
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9332330854715319,
+      "learning_rate": 2.745254101425465e-07,
+      "loss": 0.1678,
+      "step": 18239
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3009425937901884,
+      "learning_rate": 2.741422848386266e-07,
+      "loss": 0.1606,
+      "step": 18240
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2168959844464464,
+      "learning_rate": 2.737594233491458e-07,
+      "loss": 0.1757,
+      "step": 18241
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.052663313889229,
+      "learning_rate": 2.733768256844915e-07,
+      "loss": 0.1414,
+      "step": 18242
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9262570446313309,
+      "learning_rate": 2.729944918550387e-07,
+      "loss": 0.1429,
+      "step": 18243
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.810952260793739,
+      "learning_rate": 2.726124218711612e-07,
+      "loss": 0.1595,
+      "step": 18244
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0981094901401551,
+      "learning_rate": 2.7223061574321975e-07,
+      "loss": 0.184,
+      "step": 18245
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9443083295769201,
+      "learning_rate": 2.7184907348157377e-07,
+      "loss": 0.1632,
+      "step": 18246
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.075805429275343,
+      "learning_rate": 2.714677950965694e-07,
+      "loss": 0.151,
+      "step": 18247
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1320019826123289,
+      "learning_rate": 2.7108678059855064e-07,
+      "loss": 0.1539,
+      "step": 18248
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9815125581439055,
+      "learning_rate": 2.707060299978537e-07,
+      "loss": 0.1659,
+      "step": 18249
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0098818213384717,
+      "learning_rate": 2.7032554330480464e-07,
+      "loss": 0.1505,
+      "step": 18250
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1575032174458653,
+      "learning_rate": 2.699453205297253e-07,
+      "loss": 0.1633,
+      "step": 18251
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2478476951841007,
+      "learning_rate": 2.6956536168292747e-07,
+      "loss": 0.1638,
+      "step": 18252
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2882046648564152,
+      "learning_rate": 2.6918566677471946e-07,
+      "loss": 0.1565,
+      "step": 18253
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1423199526776788,
+      "learning_rate": 2.688062358153998e-07,
+      "loss": 0.1486,
+      "step": 18254
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8599154346841992,
+      "learning_rate": 2.6842706881526125e-07,
+      "loss": 0.1521,
+      "step": 18255
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0747232249223346,
+      "learning_rate": 2.680481657845868e-07,
+      "loss": 0.1405,
+      "step": 18256
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1667972927867134,
+      "learning_rate": 2.67669526733656e-07,
+      "loss": 0.1858,
+      "step": 18257
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3900194062396538,
+      "learning_rate": 2.6729115167273834e-07,
+      "loss": 0.1652,
+      "step": 18258
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0342281416929027,
+      "learning_rate": 2.66913040612099e-07,
+      "loss": 0.1841,
+      "step": 18259
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1819018631064098,
+      "learning_rate": 2.665351935619931e-07,
+      "loss": 0.1683,
+      "step": 18260
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.394688171936252,
+      "learning_rate": 2.661576105326702e-07,
+      "loss": 0.1641,
+      "step": 18261
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8566435495872279,
+      "learning_rate": 2.65780291534371e-07,
+      "loss": 0.1631,
+      "step": 18262
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.355142728772295,
+      "learning_rate": 2.654032365773318e-07,
+      "loss": 0.1421,
+      "step": 18263
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1998677089669416,
+      "learning_rate": 2.650264456717788e-07,
+      "loss": 0.1559,
+      "step": 18264
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2343661872770706,
+      "learning_rate": 2.646499188279328e-07,
+      "loss": 0.1569,
+      "step": 18265
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9675566866904394,
+      "learning_rate": 2.6427365605600883e-07,
+      "loss": 0.1459,
+      "step": 18266
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2944387885918796,
+      "learning_rate": 2.638976573662122e-07,
+      "loss": 0.1758,
+      "step": 18267
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0901993055937835,
+      "learning_rate": 2.635219227687413e-07,
+      "loss": 0.1604,
+      "step": 18268
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8275611356887129,
+      "learning_rate": 2.63146452273787e-07,
+      "loss": 0.1381,
+      "step": 18269
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1146506553802311,
+      "learning_rate": 2.6277124589153657e-07,
+      "loss": 0.1734,
+      "step": 18270
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5159778292664159,
+      "learning_rate": 2.623963036321642e-07,
+      "loss": 0.157,
+      "step": 18271
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1055152798151293,
+      "learning_rate": 2.6202162550584387e-07,
+      "loss": 0.1712,
+      "step": 18272
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.671505281711295,
+      "learning_rate": 2.6164721152273644e-07,
+      "loss": 0.1883,
+      "step": 18273
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9763923828625075,
+      "learning_rate": 2.6127306169299815e-07,
+      "loss": 0.1563,
+      "step": 18274
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1784812501395956,
+      "learning_rate": 2.608991760267776e-07,
+      "loss": 0.1732,
+      "step": 18275
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9906157660421311,
+      "learning_rate": 2.605255545342178e-07,
+      "loss": 0.1457,
+      "step": 18276
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1435715096827683,
+      "learning_rate": 2.6015219722545173e-07,
+      "loss": 0.1449,
+      "step": 18277
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0535128935316218,
+      "learning_rate": 2.5977910411060905e-07,
+      "loss": 0.1861,
+      "step": 18278
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5980771885217533,
+      "learning_rate": 2.594062751998061e-07,
+      "loss": 0.1623,
+      "step": 18279
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4537834976700899,
+      "learning_rate": 2.590337105031604e-07,
+      "loss": 0.166,
+      "step": 18280
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.231681614406639,
+      "learning_rate": 2.586614100307738e-07,
+      "loss": 0.1532,
+      "step": 18281
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0291305206451635,
+      "learning_rate": 2.582893737927472e-07,
+      "loss": 0.1823,
+      "step": 18282
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8890034877363463,
+      "learning_rate": 2.5791760179917135e-07,
+      "loss": 0.1658,
+      "step": 18283
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0296792104223675,
+      "learning_rate": 2.5754609406013044e-07,
+      "loss": 0.161,
+      "step": 18284
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8199592585300417,
+      "learning_rate": 2.5717485058570304e-07,
+      "loss": 0.1665,
+      "step": 18285
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0453994543406147,
+      "learning_rate": 2.5680387138595663e-07,
+      "loss": 0.1708,
+      "step": 18286
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.324591897357187,
+      "learning_rate": 2.5643315647095655e-07,
+      "loss": 0.1461,
+      "step": 18287
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.942078050240307,
+      "learning_rate": 2.56062705850757e-07,
+      "loss": 0.1643,
+      "step": 18288
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0278227471601282,
+      "learning_rate": 2.5569251953540763e-07,
+      "loss": 0.1624,
+      "step": 18289
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1004008321153336,
+      "learning_rate": 2.5532259753494825e-07,
+      "loss": 0.165,
+      "step": 18290
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2990139603467061,
+      "learning_rate": 2.5495293985941414e-07,
+      "loss": 0.1631,
+      "step": 18291
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.930321254713959,
+      "learning_rate": 2.5458354651883065e-07,
+      "loss": 0.1714,
+      "step": 18292
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.093198105156117,
+      "learning_rate": 2.5421441752322086e-07,
+      "loss": 0.1641,
+      "step": 18293
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9736377661780048,
+      "learning_rate": 2.538455528825945e-07,
+      "loss": 0.1577,
+      "step": 18294
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8873869753036656,
+      "learning_rate": 2.5347695260695805e-07,
+      "loss": 0.1459,
+      "step": 18295
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.216394455946613,
+      "learning_rate": 2.5310861670631015e-07,
+      "loss": 0.1644,
+      "step": 18296
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2671143732954295,
+      "learning_rate": 2.5274054519064175e-07,
+      "loss": 0.1524,
+      "step": 18297
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0028280164130396,
+      "learning_rate": 2.523727380699348e-07,
+      "loss": 0.1736,
+      "step": 18298
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3123715393982744,
+      "learning_rate": 2.5200519535417024e-07,
+      "loss": 0.157,
+      "step": 18299
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9316140446756942,
+      "learning_rate": 2.5163791705331343e-07,
+      "loss": 0.1543,
+      "step": 18300
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.046187498563863,
+      "learning_rate": 2.5127090317732973e-07,
+      "loss": 0.1673,
+      "step": 18301
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1597565818611806,
+      "learning_rate": 2.5090415373617336e-07,
+      "loss": 0.1635,
+      "step": 18302
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8820846486012719,
+      "learning_rate": 2.5053766873979204e-07,
+      "loss": 0.163,
+      "step": 18303
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9880960254385381,
+      "learning_rate": 2.5017144819812766e-07,
+      "loss": 0.1483,
+      "step": 18304
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2739535762192775,
+      "learning_rate": 2.4980549212111236e-07,
+      "loss": 0.1561,
+      "step": 18305
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.084334086475733,
+      "learning_rate": 2.494398005186749e-07,
+      "loss": 0.1584,
+      "step": 18306
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.933607113556294,
+      "learning_rate": 2.490743734007328e-07,
+      "loss": 0.147,
+      "step": 18307
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0565751035924777,
+      "learning_rate": 2.4870921077720043e-07,
+      "loss": 0.1712,
+      "step": 18308
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7900491077807519,
+      "learning_rate": 2.483443126579799e-07,
+      "loss": 0.135,
+      "step": 18309
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7874166677831361,
+      "learning_rate": 2.4797967905297093e-07,
+      "loss": 0.1366,
+      "step": 18310
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1423803695769863,
+      "learning_rate": 2.476153099720635e-07,
+      "loss": 0.1542,
+      "step": 18311
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8768095068058398,
+      "learning_rate": 2.47251205425143e-07,
+      "loss": 0.1553,
+      "step": 18312
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1056877390442028,
+      "learning_rate": 2.468873654220838e-07,
+      "loss": 0.1493,
+      "step": 18313
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7899447781250339,
+      "learning_rate": 2.465237899727557e-07,
+      "loss": 0.1755,
+      "step": 18314
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7863005549546918,
+      "learning_rate": 2.461604790870209e-07,
+      "loss": 0.1535,
+      "step": 18315
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4838306408263966,
+      "learning_rate": 2.4579743277473365e-07,
+      "loss": 0.1582,
+      "step": 18316
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9945185936066469,
+      "learning_rate": 2.454346510457417e-07,
+      "loss": 0.1703,
+      "step": 18317
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.184309812436225,
+      "learning_rate": 2.4507213390988604e-07,
+      "loss": 0.1555,
+      "step": 18318
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7978023591704787,
+      "learning_rate": 2.4470988137699993e-07,
+      "loss": 0.1503,
+      "step": 18319
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9209433934114493,
+      "learning_rate": 2.443478934569088e-07,
+      "loss": 0.1604,
+      "step": 18320
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5905991257382877,
+      "learning_rate": 2.439861701594326e-07,
+      "loss": 0.1831,
+      "step": 18321
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9910997370507246,
+      "learning_rate": 2.4362471149438237e-07,
+      "loss": 0.1515,
+      "step": 18322
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.50148417664633,
+      "learning_rate": 2.4326351747156365e-07,
+      "loss": 0.1458,
+      "step": 18323
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9682930228090741,
+      "learning_rate": 2.4290258810077183e-07,
+      "loss": 0.1564,
+      "step": 18324
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0085186474825831,
+      "learning_rate": 2.4254192339179915e-07,
+      "loss": 0.1577,
+      "step": 18325
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0797598581900214,
+      "learning_rate": 2.421815233544267e-07,
+      "loss": 0.1556,
+      "step": 18326
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8254359964739783,
+      "learning_rate": 2.418213879984321e-07,
+      "loss": 0.1522,
+      "step": 18327
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9154350038423441,
+      "learning_rate": 2.4146151733358327e-07,
+      "loss": 0.1597,
+      "step": 18328
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8372875062859295,
+      "learning_rate": 2.4110191136964333e-07,
+      "loss": 0.1628,
+      "step": 18329
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.228360493795967,
+      "learning_rate": 2.407425701163635e-07,
+      "loss": 0.1649,
+      "step": 18330
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.325196990957879,
+      "learning_rate": 2.403834935834948e-07,
+      "loss": 0.1629,
+      "step": 18331
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3452048806832089,
+      "learning_rate": 2.400246817807728e-07,
+      "loss": 0.1703,
+      "step": 18332
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9918467529509656,
+      "learning_rate": 2.396661347179341e-07,
+      "loss": 0.1876,
+      "step": 18333
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4787781081476281,
+      "learning_rate": 2.393078524047021e-07,
+      "loss": 0.1545,
+      "step": 18334
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2674619556440334,
+      "learning_rate": 2.3894983485079683e-07,
+      "loss": 0.1668,
+      "step": 18335
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0590605390474992,
+      "learning_rate": 2.385920820659271e-07,
+      "loss": 0.1529,
+      "step": 18336
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.201334926543082,
+      "learning_rate": 2.3823459405979854e-07,
+      "loss": 0.1582,
+      "step": 18337
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1333430396610051,
+      "learning_rate": 2.3787737084210893e-07,
+      "loss": 0.1554,
+      "step": 18338
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7409420607842309,
+      "learning_rate": 2.375204124225472e-07,
+      "loss": 0.1495,
+      "step": 18339
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0680731399439665,
+      "learning_rate": 2.3716371881079558e-07,
+      "loss": 0.1726,
+      "step": 18340
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0948349108272253,
+      "learning_rate": 2.3680729001652858e-07,
+      "loss": 0.1696,
+      "step": 18341
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.051007648963238,
+      "learning_rate": 2.3645112604941623e-07,
+      "loss": 0.1737,
+      "step": 18342
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0056811990684162,
+      "learning_rate": 2.3609522691911746e-07,
+      "loss": 0.1605,
+      "step": 18343
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.025955618299223,
+      "learning_rate": 2.3573959263528677e-07,
+      "loss": 0.1697,
+      "step": 18344
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0134681991322119,
+      "learning_rate": 2.3538422320757204e-07,
+      "loss": 0.1532,
+      "step": 18345
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2651014914706251,
+      "learning_rate": 2.3502911864561218e-07,
+      "loss": 0.1739,
+      "step": 18346
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0810092009920989,
+      "learning_rate": 2.346742789590384e-07,
+      "loss": 0.1534,
+      "step": 18347
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7203593062803417,
+      "learning_rate": 2.343197041574774e-07,
+      "loss": 0.1703,
+      "step": 18348
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.988197154022997,
+      "learning_rate": 2.3396539425054376e-07,
+      "loss": 0.1713,
+      "step": 18349
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1065800214268064,
+      "learning_rate": 2.3361134924785313e-07,
+      "loss": 0.1667,
+      "step": 18350
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8575847025459382,
+      "learning_rate": 2.3325756915900445e-07,
+      "loss": 0.133,
+      "step": 18351
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8488946851219393,
+      "learning_rate": 2.329040539935967e-07,
+      "loss": 0.1523,
+      "step": 18352
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8545552151939171,
+      "learning_rate": 2.3255080376121676e-07,
+      "loss": 0.1528,
+      "step": 18353
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7129768044688733,
+      "learning_rate": 2.3219781847144906e-07,
+      "loss": 0.1884,
+      "step": 18354
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2532969807361725,
+      "learning_rate": 2.3184509813386713e-07,
+      "loss": 0.1554,
+      "step": 18355
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.865328382304051,
+      "learning_rate": 2.3149264275803884e-07,
+      "loss": 0.1481,
+      "step": 18356
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.819398377750584,
+      "learning_rate": 2.3114045235352433e-07,
+      "loss": 0.1597,
+      "step": 18357
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9760997913944155,
+      "learning_rate": 2.3078852692987596e-07,
+      "loss": 0.1533,
+      "step": 18358
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8124370575612277,
+      "learning_rate": 2.3043686649664165e-07,
+      "loss": 0.149,
+      "step": 18359
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9997317244657435,
+      "learning_rate": 2.3008547106335822e-07,
+      "loss": 0.1624,
+      "step": 18360
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2057741927307049,
+      "learning_rate": 2.2973434063955802e-07,
+      "loss": 0.1463,
+      "step": 18361
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8137021469097535,
+      "learning_rate": 2.2938347523476568e-07,
+      "loss": 0.1533,
+      "step": 18362
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.0808834168689883,
+      "learning_rate": 2.2903287485849913e-07,
+      "loss": 0.1568,
+      "step": 18363
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9410298125961627,
+      "learning_rate": 2.2868253952026741e-07,
+      "loss": 0.1491,
+      "step": 18364
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0671457063576877,
+      "learning_rate": 2.2833246922957408e-07,
+      "loss": 0.1649,
+      "step": 18365
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.163398353724536,
+      "learning_rate": 2.2798266399591374e-07,
+      "loss": 0.1666,
+      "step": 18366
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.893536336859922,
+      "learning_rate": 2.2763312382877656e-07,
+      "loss": 0.1493,
+      "step": 18367
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1466722066958055,
+      "learning_rate": 2.2728384873764276e-07,
+      "loss": 0.1667,
+      "step": 18368
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1088802038655343,
+      "learning_rate": 2.2693483873198584e-07,
+      "loss": 0.1415,
+      "step": 18369
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1395079425027594,
+      "learning_rate": 2.2658609382127384e-07,
+      "loss": 0.1837,
+      "step": 18370
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0489862597909092,
+      "learning_rate": 2.2623761401496468e-07,
+      "loss": 0.1582,
+      "step": 18371
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.030787526510104,
+      "learning_rate": 2.2588939932251418e-07,
+      "loss": 0.1592,
+      "step": 18372
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9229346827535737,
+      "learning_rate": 2.2554144975336477e-07,
+      "loss": 0.1509,
+      "step": 18373
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2346282121275605,
+      "learning_rate": 2.251937653169567e-07,
+      "loss": 0.1606,
+      "step": 18374
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8715933060605532,
+      "learning_rate": 2.2484634602271794e-07,
+      "loss": 0.1683,
+      "step": 18375
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.941253110893302,
+      "learning_rate": 2.244991918800754e-07,
+      "loss": 0.1521,
+      "step": 18376
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9266730919291768,
+      "learning_rate": 2.2415230289844382e-07,
+      "loss": 0.1589,
+      "step": 18377
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9771410611139677,
+      "learning_rate": 2.238056790872345e-07,
+      "loss": 0.1724,
+      "step": 18378
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5235936206743117,
+      "learning_rate": 2.2345932045584662e-07,
+      "loss": 0.196,
+      "step": 18379
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9841273240110063,
+      "learning_rate": 2.231132270136771e-07,
+      "loss": 0.1763,
+      "step": 18380
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2765539802867825,
+      "learning_rate": 2.22767398770114e-07,
+      "loss": 0.155,
+      "step": 18381
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1186553643443082,
+      "learning_rate": 2.2242183573453756e-07,
+      "loss": 0.1868,
+      "step": 18382
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0287318642723777,
+      "learning_rate": 2.2207653791632035e-07,
+      "loss": 0.1417,
+      "step": 18383
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3699336996747302,
+      "learning_rate": 2.2173150532483035e-07,
+      "loss": 0.1648,
+      "step": 18384
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9571519442266735,
+      "learning_rate": 2.2138673796942457e-07,
+      "loss": 0.1568,
+      "step": 18385
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2746480790935992,
+      "learning_rate": 2.210422358594566e-07,
+      "loss": 0.167,
+      "step": 18386
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.106230206821423,
+      "learning_rate": 2.2069799900427014e-07,
+      "loss": 0.1719,
+      "step": 18387
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.791480398398672,
+      "learning_rate": 2.203540274132021e-07,
+      "loss": 0.149,
+      "step": 18388
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0617422537257237,
+      "learning_rate": 2.2001032109558396e-07,
+      "loss": 0.1616,
+      "step": 18389
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8104952562718482,
+      "learning_rate": 2.196668800607382e-07,
+      "loss": 0.1635,
+      "step": 18390
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9657527974823339,
+      "learning_rate": 2.1932370431798077e-07,
+      "loss": 0.1476,
+      "step": 18391
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.983063383851243,
+      "learning_rate": 2.1898079387662085e-07,
+      "loss": 0.1919,
+      "step": 18392
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6038285199676197,
+      "learning_rate": 2.186381487459588e-07,
+      "loss": 0.1672,
+      "step": 18393
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8242832192874088,
+      "learning_rate": 2.1829576893528938e-07,
+      "loss": 0.1552,
+      "step": 18394
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.072684200243416,
+      "learning_rate": 2.179536544538996e-07,
+      "loss": 0.1466,
+      "step": 18395
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9253757492899954,
+      "learning_rate": 2.1761180531106873e-07,
+      "loss": 0.1728,
+      "step": 18396
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.066376706831964,
+      "learning_rate": 2.1727022151607046e-07,
+      "loss": 0.1488,
+      "step": 18397
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.314431093429177,
+      "learning_rate": 2.1692890307817073e-07,
+      "loss": 0.1631,
+      "step": 18398
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8498934849664621,
+      "learning_rate": 2.1658785000662763e-07,
+      "loss": 0.1486,
+      "step": 18399
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0378436730938367,
+      "learning_rate": 2.1624706231068936e-07,
+      "loss": 0.1573,
+      "step": 18400
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8970051524119859,
+      "learning_rate": 2.159065399996041e-07,
+      "loss": 0.1624,
+      "step": 18401
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9922525353722269,
+      "learning_rate": 2.1556628308260552e-07,
+      "loss": 0.1818,
+      "step": 18402
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8332483233289575,
+      "learning_rate": 2.1522629156892516e-07,
+      "loss": 0.1438,
+      "step": 18403
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8701610657938882,
+      "learning_rate": 2.1488656546778342e-07,
+      "loss": 0.1745,
+      "step": 18404
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9793402914905001,
+      "learning_rate": 2.1454710478839736e-07,
+      "loss": 0.174,
+      "step": 18405
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.817119570072124,
+      "learning_rate": 2.1420790953997294e-07,
+      "loss": 0.1722,
+      "step": 18406
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.782561768407933,
+      "learning_rate": 2.138689797317106e-07,
+      "loss": 0.1741,
+      "step": 18407
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4030966551078101,
+      "learning_rate": 2.1353031537280743e-07,
+      "loss": 0.1519,
+      "step": 18408
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4607777385501066,
+      "learning_rate": 2.1319191647244497e-07,
+      "loss": 0.1799,
+      "step": 18409
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9416992136910253,
+      "learning_rate": 2.1285378303980585e-07,
+      "loss": 0.1441,
+      "step": 18410
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8708030270127827,
+      "learning_rate": 2.1251591508405945e-07,
+      "loss": 0.1561,
+      "step": 18411
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.04718883719092,
+      "learning_rate": 2.1217831261437283e-07,
+      "loss": 0.1471,
+      "step": 18412
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4080305099714423,
+      "learning_rate": 2.1184097563990204e-07,
+      "loss": 0.1692,
+      "step": 18413
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3780724995478995,
+      "learning_rate": 2.115039041697964e-07,
+      "loss": 0.1723,
+      "step": 18414
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0810492064587849,
+      "learning_rate": 2.111670982132008e-07,
+      "loss": 0.1642,
+      "step": 18415
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4676322049548796,
+      "learning_rate": 2.1083055777925244e-07,
+      "loss": 0.1595,
+      "step": 18416
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3846819242414135,
+      "learning_rate": 2.104942828770762e-07,
+      "loss": 0.172,
+      "step": 18417
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.007773102930094,
+      "learning_rate": 2.1015827351579588e-07,
+      "loss": 0.1845,
+      "step": 18418
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9008750167154151,
+      "learning_rate": 2.0982252970452532e-07,
+      "loss": 0.1526,
+      "step": 18419
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0345250458075697,
+      "learning_rate": 2.0948705145237168e-07,
+      "loss": 0.1622,
+      "step": 18420
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.982428132994956,
+      "learning_rate": 2.0915183876843436e-07,
+      "loss": 0.1604,
+      "step": 18421
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9070283705308484,
+      "learning_rate": 2.0881689166180718e-07,
+      "loss": 0.1657,
+      "step": 18422
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2758911966430224,
+      "learning_rate": 2.0848221014157398e-07,
+      "loss": 0.1613,
+      "step": 18423
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1603708286280212,
+      "learning_rate": 2.081477942168142e-07,
+      "loss": 0.1639,
+      "step": 18424
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9503295180273585,
+      "learning_rate": 2.078136438965983e-07,
+      "loss": 0.1695,
+      "step": 18425
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0634378735733225,
+      "learning_rate": 2.0747975918999018e-07,
+      "loss": 0.1596,
+      "step": 18426
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3307297893329564,
+      "learning_rate": 2.0714614010604815e-07,
+      "loss": 0.1695,
+      "step": 18427
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3323480771188736,
+      "learning_rate": 2.0681278665381833e-07,
+      "loss": 0.145,
+      "step": 18428
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1547516123571364,
+      "learning_rate": 2.0647969884234676e-07,
+      "loss": 0.162,
+      "step": 18429
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.032996754590199,
+      "learning_rate": 2.0614687668066403e-07,
+      "loss": 0.161,
+      "step": 18430
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3829844629918333,
+      "learning_rate": 2.058143201778029e-07,
+      "loss": 0.1819,
+      "step": 18431
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6345825829161515,
+      "learning_rate": 2.054820293427795e-07,
+      "loss": 0.1622,
+      "step": 18432
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.13058969646442,
+      "learning_rate": 2.0515000418460995e-07,
+      "loss": 0.1688,
+      "step": 18433
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1264091710524033,
+      "learning_rate": 2.0481824471229927e-07,
+      "loss": 0.126,
+      "step": 18434
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9524093713904969,
+      "learning_rate": 2.0448675093484805e-07,
+      "loss": 0.1369,
+      "step": 18435
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0332060689778997,
+      "learning_rate": 2.0415552286124685e-07,
+      "loss": 0.164,
+      "step": 18436
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.154708584588988,
+      "learning_rate": 2.0382456050048073e-07,
+      "loss": 0.1745,
+      "step": 18437
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9288602652094318,
+      "learning_rate": 2.034938638615247e-07,
+      "loss": 0.1668,
+      "step": 18438
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0836570479763294,
+      "learning_rate": 2.0316343295335272e-07,
+      "loss": 0.1602,
+      "step": 18439
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0766862097296297,
+      "learning_rate": 2.028332677849254e-07,
+      "loss": 0.1405,
+      "step": 18440
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.012468415541933,
+      "learning_rate": 2.025033683651989e-07,
+      "loss": 0.1652,
+      "step": 18441
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.088899367723937,
+      "learning_rate": 2.0217373470312275e-07,
+      "loss": 0.163,
+      "step": 18442
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1267366751772347,
+      "learning_rate": 2.018443668076364e-07,
+      "loss": 0.158,
+      "step": 18443
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1178922069284865,
+      "learning_rate": 2.0151526468767502e-07,
+      "loss": 0.1898,
+      "step": 18444
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9563470216819234,
+      "learning_rate": 2.0118642835216584e-07,
+      "loss": 0.1625,
+      "step": 18445
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9093202957962132,
+      "learning_rate": 2.0085785781002843e-07,
+      "loss": 0.1569,
+      "step": 18446
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0411398327926424,
+      "learning_rate": 2.005295530701745e-07,
+      "loss": 0.1465,
+      "step": 18447
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2165271777702973,
+      "learning_rate": 2.0020151414151146e-07,
+      "loss": 0.1421,
+      "step": 18448
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2697362647353578,
+      "learning_rate": 1.9987374103293433e-07,
+      "loss": 0.1722,
+      "step": 18449
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9059765341830827,
+      "learning_rate": 1.9954623375333493e-07,
+      "loss": 0.1608,
+      "step": 18450
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9790695354920086,
+      "learning_rate": 1.9921899231159836e-07,
+      "loss": 0.1538,
+      "step": 18451
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7967358277210099,
+      "learning_rate": 1.9889201671660084e-07,
+      "loss": 0.1439,
+      "step": 18452
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1676863492777445,
+      "learning_rate": 1.9856530697720976e-07,
+      "loss": 0.1732,
+      "step": 18453
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9026498634322633,
+      "learning_rate": 1.9823886310228911e-07,
+      "loss": 0.1515,
+      "step": 18454
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5898085358785705,
+      "learning_rate": 1.9791268510069184e-07,
+      "loss": 0.1791,
+      "step": 18455
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8838055911204667,
+      "learning_rate": 1.975867729812686e-07,
+      "loss": 0.1318,
+      "step": 18456
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8962589559141396,
+      "learning_rate": 1.9726112675285568e-07,
+      "loss": 0.1487,
+      "step": 18457
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1455441239812183,
+      "learning_rate": 1.9693574642428935e-07,
+      "loss": 0.1594,
+      "step": 18458
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8697554124013273,
+      "learning_rate": 1.9661063200439478e-07,
+      "loss": 0.1747,
+      "step": 18459
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8473698779305922,
+      "learning_rate": 1.9628578350198933e-07,
+      "loss": 0.1416,
+      "step": 18460
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1622171600184905,
+      "learning_rate": 1.959612009258871e-07,
+      "loss": 0.1898,
+      "step": 18461
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9635651441512039,
+      "learning_rate": 1.9563688428489103e-07,
+      "loss": 0.1467,
+      "step": 18462
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9709659456803279,
+      "learning_rate": 1.953128335877974e-07,
+      "loss": 0.1551,
+      "step": 18463
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1570414393860904,
+      "learning_rate": 1.9498904884339697e-07,
+      "loss": 0.1885,
+      "step": 18464
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0220800299049926,
+      "learning_rate": 1.9466553006047383e-07,
+      "loss": 0.1556,
+      "step": 18465
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.17271748461141,
+      "learning_rate": 1.9434227724779987e-07,
+      "loss": 0.1475,
+      "step": 18466
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9698973968273322,
+      "learning_rate": 1.9401929041414692e-07,
+      "loss": 0.2208,
+      "step": 18467
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9539471986986922,
+      "learning_rate": 1.9369656956827355e-07,
+      "loss": 0.1628,
+      "step": 18468
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0196693761635764,
+      "learning_rate": 1.9337411471893498e-07,
+      "loss": 0.1703,
+      "step": 18469
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1934751615375507,
+      "learning_rate": 1.9305192587487753e-07,
+      "loss": 0.1462,
+      "step": 18470
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.199618048089345,
+      "learning_rate": 1.927300030448409e-07,
+      "loss": 0.1973,
+      "step": 18471
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1152576551484823,
+      "learning_rate": 1.924083462375559e-07,
+      "loss": 0.1673,
+      "step": 18472
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0406551584167096,
+      "learning_rate": 1.9208695546174994e-07,
+      "loss": 0.1679,
+      "step": 18473
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1662476783946836,
+      "learning_rate": 1.917658307261383e-07,
+      "loss": 0.1899,
+      "step": 18474
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0191715895198943,
+      "learning_rate": 1.914449720394329e-07,
+      "loss": 0.1636,
+      "step": 18475
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1530279147333242,
+      "learning_rate": 1.9112437941033567e-07,
+      "loss": 0.1574,
+      "step": 18476
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1700995784032147,
+      "learning_rate": 1.908040528475441e-07,
+      "loss": 0.1662,
+      "step": 18477
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1758161723996137,
+      "learning_rate": 1.904839923597468e-07,
+      "loss": 0.1666,
+      "step": 18478
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2056737356424008,
+      "learning_rate": 1.9016419795562568e-07,
+      "loss": 0.177,
+      "step": 18479
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8871261305872965,
+      "learning_rate": 1.8984466964385384e-07,
+      "loss": 0.1552,
+      "step": 18480
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.878197883112823,
+      "learning_rate": 1.8952540743309988e-07,
+      "loss": 0.1556,
+      "step": 18481
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0543218480340648,
+      "learning_rate": 1.8920641133202356e-07,
+      "loss": 0.1494,
+      "step": 18482
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.941396341109944,
+      "learning_rate": 1.888876813492768e-07,
+      "loss": 0.1685,
+      "step": 18483
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9301788089742421,
+      "learning_rate": 1.8856921749350608e-07,
+      "loss": 0.1506,
+      "step": 18484
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1855751996628046,
+      "learning_rate": 1.8825101977334891e-07,
+      "loss": 0.1714,
+      "step": 18485
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7965655623295367,
+      "learning_rate": 1.8793308819743837e-07,
+      "loss": 0.1793,
+      "step": 18486
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0720094037637893,
+      "learning_rate": 1.8761542277439648e-07,
+      "loss": 0.1738,
+      "step": 18487
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4204108111036178,
+      "learning_rate": 1.8729802351284077e-07,
+      "loss": 0.1427,
+      "step": 18488
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0684230576153166,
+      "learning_rate": 1.869808904213799e-07,
+      "loss": 0.1785,
+      "step": 18489
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8876069325807378,
+      "learning_rate": 1.8666402350861701e-07,
+      "loss": 0.1544,
+      "step": 18490
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6348599888108668,
+      "learning_rate": 1.8634742278314632e-07,
+      "loss": 0.1696,
+      "step": 18491
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1283463049697549,
+      "learning_rate": 1.8603108825355654e-07,
+      "loss": 0.1829,
+      "step": 18492
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0989175147355235,
+      "learning_rate": 1.8571501992842634e-07,
+      "loss": 0.1541,
+      "step": 18493
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5485827407936898,
+      "learning_rate": 1.8539921781633107e-07,
+      "loss": 0.1574,
+      "step": 18494
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2035629459141746,
+      "learning_rate": 1.8508368192583838e-07,
+      "loss": 0.1599,
+      "step": 18495
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.175110774320181,
+      "learning_rate": 1.8476841226550247e-07,
+      "loss": 0.166,
+      "step": 18496
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8648008632983784,
+      "learning_rate": 1.8445340884387986e-07,
+      "loss": 0.1555,
+      "step": 18497
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.186022566237658,
+      "learning_rate": 1.841386716695115e-07,
+      "loss": 0.1621,
+      "step": 18498
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.966907537171716,
+      "learning_rate": 1.8382420075093722e-07,
+      "loss": 0.1663,
+      "step": 18499
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.35987273141365,
+      "learning_rate": 1.8350999609668462e-07,
+      "loss": 0.1931,
+      "step": 18500
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7185887664553376,
+      "learning_rate": 1.8319605771527916e-07,
+      "loss": 0.1508,
+      "step": 18501
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9654463459049943,
+      "learning_rate": 1.8288238561523397e-07,
+      "loss": 0.1851,
+      "step": 18502
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5191769294067194,
+      "learning_rate": 1.8256897980505895e-07,
+      "loss": 0.1786,
+      "step": 18503
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0547956256766307,
+      "learning_rate": 1.8225584029325394e-07,
+      "loss": 0.1451,
+      "step": 18504
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9602419545290328,
+      "learning_rate": 1.8194296708831548e-07,
+      "loss": 0.1654,
+      "step": 18505
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9905278074256648,
+      "learning_rate": 1.8163036019872682e-07,
+      "loss": 0.1699,
+      "step": 18506
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9878465962130885,
+      "learning_rate": 1.8131801963297112e-07,
+      "loss": 0.1574,
+      "step": 18507
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.202120778350107,
+      "learning_rate": 1.810059453995172e-07,
+      "loss": 0.1595,
+      "step": 18508
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5424916385881604,
+      "learning_rate": 1.8069413750683274e-07,
+      "loss": 0.1531,
+      "step": 18509
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2032151268116966,
+      "learning_rate": 1.8038259596337316e-07,
+      "loss": 0.1527,
+      "step": 18510
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9692901581171245,
+      "learning_rate": 1.8007132077759059e-07,
+      "loss": 0.1611,
+      "step": 18511
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2384022796992629,
+      "learning_rate": 1.7976031195792942e-07,
+      "loss": 0.1616,
+      "step": 18512
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9518845029745795,
+      "learning_rate": 1.794495695128229e-07,
+      "loss": 0.1513,
+      "step": 18513
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3445048561231276,
+      "learning_rate": 1.7913909345070202e-07,
+      "loss": 0.1584,
+      "step": 18514
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8910941735915785,
+      "learning_rate": 1.7882888377998787e-07,
+      "loss": 0.1346,
+      "step": 18515
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.058905833198794,
+      "learning_rate": 1.7851894050909479e-07,
+      "loss": 0.1772,
+      "step": 18516
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.889457926665445,
+      "learning_rate": 1.7820926364643054e-07,
+      "loss": 0.1433,
+      "step": 18517
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1393201867917613,
+      "learning_rate": 1.7789985320039505e-07,
+      "loss": 0.1922,
+      "step": 18518
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1461805626320478,
+      "learning_rate": 1.7759070917937937e-07,
+      "loss": 0.1691,
+      "step": 18519
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9552372498782165,
+      "learning_rate": 1.7728183159177126e-07,
+      "loss": 0.1574,
+      "step": 18520
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.317634139784253,
+      "learning_rate": 1.7697322044594846e-07,
+      "loss": 0.1621,
+      "step": 18521
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5026527540251304,
+      "learning_rate": 1.76664875750282e-07,
+      "loss": 0.1535,
+      "step": 18522
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9529695966362643,
+      "learning_rate": 1.7635679751313529e-07,
+      "loss": 0.1604,
+      "step": 18523
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.362781251116253,
+      "learning_rate": 1.7604898574286488e-07,
+      "loss": 0.1676,
+      "step": 18524
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.794298494281626,
+      "learning_rate": 1.7574144044782083e-07,
+      "loss": 0.1632,
+      "step": 18525
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3543147984644823,
+      "learning_rate": 1.754341616363464e-07,
+      "loss": 0.1539,
+      "step": 18526
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8098682812727673,
+      "learning_rate": 1.7512714931677387e-07,
+      "loss": 0.1744,
+      "step": 18527
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0965203264417904,
+      "learning_rate": 1.7482040349743323e-07,
+      "loss": 0.1719,
+      "step": 18528
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7137149759900159,
+      "learning_rate": 1.7451392418664227e-07,
+      "loss": 0.1512,
+      "step": 18529
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8312516693775386,
+      "learning_rate": 1.7420771139271765e-07,
+      "loss": 0.1578,
+      "step": 18530
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.941353238215755,
+      "learning_rate": 1.7390176512396384e-07,
+      "loss": 0.1575,
+      "step": 18531
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9570430798627169,
+      "learning_rate": 1.7359608538867868e-07,
+      "loss": 0.164,
+      "step": 18532
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2019136658490237,
+      "learning_rate": 1.732906721951555e-07,
+      "loss": 0.1681,
+      "step": 18533
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6202511232946657,
+      "learning_rate": 1.729855255516777e-07,
+      "loss": 0.1778,
+      "step": 18534
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.134760261935929,
+      "learning_rate": 1.7268064546652308e-07,
+      "loss": 0.1998,
+      "step": 18535
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8102705538864348,
+      "learning_rate": 1.7237603194795948e-07,
+      "loss": 0.1581,
+      "step": 18536
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8801215439384064,
+      "learning_rate": 1.7207168500425142e-07,
+      "loss": 0.1523,
+      "step": 18537
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8586275430989962,
+      "learning_rate": 1.7176760464365449e-07,
+      "loss": 0.1519,
+      "step": 18538
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9018279699676834,
+      "learning_rate": 1.7146379087441655e-07,
+      "loss": 0.1914,
+      "step": 18539
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5663057154553512,
+      "learning_rate": 1.711602437047788e-07,
+      "loss": 0.1854,
+      "step": 18540
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0250812414102985,
+      "learning_rate": 1.708569631429746e-07,
+      "loss": 0.1579,
+      "step": 18541
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4051658931794355,
+      "learning_rate": 1.7055394919722856e-07,
+      "loss": 0.1577,
+      "step": 18542
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7920877521485331,
+      "learning_rate": 1.7025120187576406e-07,
+      "loss": 0.1664,
+      "step": 18543
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0138473039450129,
+      "learning_rate": 1.6994872118679006e-07,
+      "loss": 0.1551,
+      "step": 18544
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8477002854411233,
+      "learning_rate": 1.6964650713851228e-07,
+      "loss": 0.1292,
+      "step": 18545
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2484301198202659,
+      "learning_rate": 1.6934455973912744e-07,
+      "loss": 0.1529,
+      "step": 18546
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.576777670975822,
+      "learning_rate": 1.690428789968268e-07,
+      "loss": 0.1777,
+      "step": 18547
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.91852055430131,
+      "learning_rate": 1.6874146491979493e-07,
+      "loss": 0.1524,
+      "step": 18548
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9500129432796155,
+      "learning_rate": 1.6844031751620414e-07,
+      "loss": 0.1788,
+      "step": 18549
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7767599888761454,
+      "learning_rate": 1.6813943679422684e-07,
+      "loss": 0.1486,
+      "step": 18550
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9680141309651069,
+      "learning_rate": 1.678388227620209e-07,
+      "loss": 0.1601,
+      "step": 18551
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0630914404291878,
+      "learning_rate": 1.6753847542774315e-07,
+      "loss": 0.1498,
+      "step": 18552
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5211018887090226,
+      "learning_rate": 1.6723839479953929e-07,
+      "loss": 0.1689,
+      "step": 18553
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.468104798877948,
+      "learning_rate": 1.669385808855495e-07,
+      "loss": 0.1652,
+      "step": 18554
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.001087755929558,
+      "learning_rate": 1.666390336939061e-07,
+      "loss": 0.1736,
+      "step": 18555
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8453852231440885,
+      "learning_rate": 1.6633975323273376e-07,
+      "loss": 0.1438,
+      "step": 18556
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0352472415273843,
+      "learning_rate": 1.6604073951015154e-07,
+      "loss": 0.1709,
+      "step": 18557
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9459021394612915,
+      "learning_rate": 1.6574199253426958e-07,
+      "loss": 0.1691,
+      "step": 18558
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8641474095406372,
+      "learning_rate": 1.6544351231319145e-07,
+      "loss": 0.1436,
+      "step": 18559
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1152650363612875,
+      "learning_rate": 1.6514529885501397e-07,
+      "loss": 0.1423,
+      "step": 18560
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2606785611258506,
+      "learning_rate": 1.648473521678251e-07,
+      "loss": 0.1796,
+      "step": 18561
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.298872067227049,
+      "learning_rate": 1.645496722597084e-07,
+      "loss": 0.1715,
+      "step": 18562
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8674647051865187,
+      "learning_rate": 1.642522591387352e-07,
+      "loss": 0.1452,
+      "step": 18563
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0411258487676804,
+      "learning_rate": 1.6395511281297682e-07,
+      "loss": 0.1454,
+      "step": 18564
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2294606917015065,
+      "learning_rate": 1.6365823329049124e-07,
+      "loss": 0.1551,
+      "step": 18565
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2626410672708395,
+      "learning_rate": 1.633616205793309e-07,
+      "loss": 0.1647,
+      "step": 18566
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8820935476550902,
+      "learning_rate": 1.6306527468754384e-07,
+      "loss": 0.1538,
+      "step": 18567
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.076423859688967,
+      "learning_rate": 1.6276919562316475e-07,
+      "loss": 0.1607,
+      "step": 18568
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.058599832374519,
+      "learning_rate": 1.6247338339422823e-07,
+      "loss": 0.1652,
+      "step": 18569
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9862983112602691,
+      "learning_rate": 1.6217783800875576e-07,
+      "loss": 0.1525,
+      "step": 18570
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1039987539964633,
+      "learning_rate": 1.618825594747664e-07,
+      "loss": 0.1691,
+      "step": 18571
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1355796680292811,
+      "learning_rate": 1.615875478002671e-07,
+      "loss": 0.1444,
+      "step": 18572
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1295020685609354,
+      "learning_rate": 1.6129280299326144e-07,
+      "loss": 0.1779,
+      "step": 18573
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1228313989287582,
+      "learning_rate": 1.6099832506174419e-07,
+      "loss": 0.1544,
+      "step": 18574
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3138722407016137,
+      "learning_rate": 1.6070411401370335e-07,
+      "loss": 0.1783,
+      "step": 18575
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1715432450412604,
+      "learning_rate": 1.6041016985711923e-07,
+      "loss": 0.161,
+      "step": 18576
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3627767559615154,
+      "learning_rate": 1.6011649259996541e-07,
+      "loss": 0.172,
+      "step": 18577
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1528273504272677,
+      "learning_rate": 1.598230822502067e-07,
+      "loss": 0.1569,
+      "step": 18578
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5197766356269795,
+      "learning_rate": 1.5952993881580336e-07,
+      "loss": 0.1562,
+      "step": 18579
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6041305313914618,
+      "learning_rate": 1.592370623047046e-07,
+      "loss": 0.1507,
+      "step": 18580
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8985954018484826,
+      "learning_rate": 1.5894445272485736e-07,
+      "loss": 0.1629,
+      "step": 18581
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0707360930908472,
+      "learning_rate": 1.586521100841987e-07,
+      "loss": 0.1543,
+      "step": 18582
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3859762485376421,
+      "learning_rate": 1.583600343906566e-07,
+      "loss": 0.1772,
+      "step": 18583
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9957403277629844,
+      "learning_rate": 1.5806822565215373e-07,
+      "loss": 0.1627,
+      "step": 18584
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4046400140650128,
+      "learning_rate": 1.5777668387660706e-07,
+      "loss": 0.1851,
+      "step": 18585
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.891535151011113,
+      "learning_rate": 1.5748540907192356e-07,
+      "loss": 0.2006,
+      "step": 18586
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5090297703983357,
+      "learning_rate": 1.571944012460036e-07,
+      "loss": 0.1712,
+      "step": 18587
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9003120205448941,
+      "learning_rate": 1.569036604067431e-07,
+      "loss": 0.1649,
+      "step": 18588
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1480753994260278,
+      "learning_rate": 1.566131865620246e-07,
+      "loss": 0.158,
+      "step": 18589
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9565359080976353,
+      "learning_rate": 1.5632297971972966e-07,
+      "loss": 0.1721,
+      "step": 18590
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0272957499468065,
+      "learning_rate": 1.5603303988773078e-07,
+      "loss": 0.1378,
+      "step": 18591
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9377209184491478,
+      "learning_rate": 1.5574336707389171e-07,
+      "loss": 0.1706,
+      "step": 18592
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7620016285666292,
+      "learning_rate": 1.554539612860695e-07,
+      "loss": 0.14,
+      "step": 18593
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9909527907118318,
+      "learning_rate": 1.551648225321145e-07,
+      "loss": 0.1582,
+      "step": 18594
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9335941141651837,
+      "learning_rate": 1.548759508198694e-07,
+      "loss": 0.1469,
+      "step": 18595
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7752699845301906,
+      "learning_rate": 1.545873461571712e-07,
+      "loss": 0.1461,
+      "step": 18596
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9913274331233224,
+      "learning_rate": 1.54299008551847e-07,
+      "loss": 0.1474,
+      "step": 18597
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2023120555067652,
+      "learning_rate": 1.5401093801171828e-07,
+      "loss": 0.1873,
+      "step": 18598
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.071690745494057,
+      "learning_rate": 1.5372313454459887e-07,
+      "loss": 0.1433,
+      "step": 18599
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.053479587695811,
+      "learning_rate": 1.5343559815829468e-07,
+      "loss": 0.1503,
+      "step": 18600
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9531335525572882,
+      "learning_rate": 1.5314832886060727e-07,
+      "loss": 0.1471,
+      "step": 18601
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.036438135675386,
+      "learning_rate": 1.5286132665932706e-07,
+      "loss": 0.1672,
+      "step": 18602
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9933283290691903,
+      "learning_rate": 1.525745915622401e-07,
+      "loss": 0.1593,
+      "step": 18603
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1621345115927901,
+      "learning_rate": 1.5228812357712231e-07,
+      "loss": 0.1512,
+      "step": 18604
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.860499166036676,
+      "learning_rate": 1.520019227117464e-07,
+      "loss": 0.167,
+      "step": 18605
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1558556579174626,
+      "learning_rate": 1.5171598897387395e-07,
+      "loss": 0.154,
+      "step": 18606
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9880628780813716,
+      "learning_rate": 1.514303223712621e-07,
+      "loss": 0.1725,
+      "step": 18607
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1171775767513268,
+      "learning_rate": 1.5114492291165794e-07,
+      "loss": 0.1747,
+      "step": 18608
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1338849876747592,
+      "learning_rate": 1.508597906028053e-07,
+      "loss": 0.1733,
+      "step": 18609
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1072484917924614,
+      "learning_rate": 1.505749254524358e-07,
+      "loss": 0.1475,
+      "step": 18610
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9919583007519222,
+      "learning_rate": 1.5029032746827875e-07,
+      "loss": 0.1575,
+      "step": 18611
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8184994189922,
+      "learning_rate": 1.500059966580525e-07,
+      "loss": 0.1411,
+      "step": 18612
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.398520475493659,
+      "learning_rate": 1.497219330294708e-07,
+      "loss": 0.1708,
+      "step": 18613
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0964778320968296,
+      "learning_rate": 1.4943813659023753e-07,
+      "loss": 0.1693,
+      "step": 18614
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6528757245841295,
+      "learning_rate": 1.49154607348051e-07,
+      "loss": 0.1722,
+      "step": 18615
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9692735415387215,
+      "learning_rate": 1.4887134531060165e-07,
+      "loss": 0.1523,
+      "step": 18616
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0528475743073271,
+      "learning_rate": 1.485883504855734e-07,
+      "loss": 0.1507,
+      "step": 18617
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9032628322727292,
+      "learning_rate": 1.4830562288064344e-07,
+      "loss": 0.1719,
+      "step": 18618
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0427542611832965,
+      "learning_rate": 1.4802316250347893e-07,
+      "loss": 0.1806,
+      "step": 18619
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9516894573352137,
+      "learning_rate": 1.4774096936174376e-07,
+      "loss": 0.1616,
+      "step": 18620
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.189078582059111,
+      "learning_rate": 1.474590434630907e-07,
+      "loss": 0.1558,
+      "step": 18621
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0663919954221222,
+      "learning_rate": 1.4717738481516808e-07,
+      "loss": 0.1742,
+      "step": 18622
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0112513802414314,
+      "learning_rate": 1.4689599342561423e-07,
+      "loss": 0.1508,
+      "step": 18623
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9221034568216941,
+      "learning_rate": 1.4661486930206415e-07,
+      "loss": 0.1558,
+      "step": 18624
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.141316296046559,
+      "learning_rate": 1.4633401245214064e-07,
+      "loss": 0.156,
+      "step": 18625
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0125885082259565,
+      "learning_rate": 1.4605342288346536e-07,
+      "loss": 0.1609,
+      "step": 18626
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2259660659391147,
+      "learning_rate": 1.4577310060364558e-07,
+      "loss": 0.1641,
+      "step": 18627
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0070231515282992,
+      "learning_rate": 1.4549304562028966e-07,
+      "loss": 0.1706,
+      "step": 18628
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8620628643100052,
+      "learning_rate": 1.4521325794098928e-07,
+      "loss": 0.1614,
+      "step": 18629
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9305244147767624,
+      "learning_rate": 1.449337375733373e-07,
+      "loss": 0.1551,
+      "step": 18630
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5742305713151112,
+      "learning_rate": 1.4465448452491315e-07,
+      "loss": 0.1686,
+      "step": 18631
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2104559628853415,
+      "learning_rate": 1.4437549880329415e-07,
+      "loss": 0.16,
+      "step": 18632
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.239220991219772,
+      "learning_rate": 1.4409678041604426e-07,
+      "loss": 0.1666,
+      "step": 18633
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.786747929668593,
+      "learning_rate": 1.4381832937072737e-07,
+      "loss": 0.1394,
+      "step": 18634
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2145595739356847,
+      "learning_rate": 1.4354014567489528e-07,
+      "loss": 0.1649,
+      "step": 18635
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9754466948839298,
+      "learning_rate": 1.432622293360919e-07,
+      "loss": 0.1846,
+      "step": 18636
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0548122289678061,
+      "learning_rate": 1.4298458036185903e-07,
+      "loss": 0.1495,
+      "step": 18637
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.403480202485536,
+      "learning_rate": 1.4270719875972506e-07,
+      "loss": 0.1592,
+      "step": 18638
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.063392804506902,
+      "learning_rate": 1.424300845372162e-07,
+      "loss": 0.1699,
+      "step": 18639
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2431406279033599,
+      "learning_rate": 1.4215323770184642e-07,
+      "loss": 0.1733,
+      "step": 18640
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9805121919052685,
+      "learning_rate": 1.418766582611286e-07,
+      "loss": 0.1477,
+      "step": 18641
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1772098397613278,
+      "learning_rate": 1.4160034622256125e-07,
+      "loss": 0.1678,
+      "step": 18642
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2741303245522206,
+      "learning_rate": 1.4132430159364273e-07,
+      "loss": 0.154,
+      "step": 18643
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0401178606701769,
+      "learning_rate": 1.4104852438185823e-07,
+      "loss": 0.1571,
+      "step": 18644
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3489124060031128,
+      "learning_rate": 1.4077301459469062e-07,
+      "loss": 0.1594,
+      "step": 18645
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1101616198916429,
+      "learning_rate": 1.404977722396106e-07,
+      "loss": 0.1495,
+      "step": 18646
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.266653077152881,
+      "learning_rate": 1.4022279732408661e-07,
+      "loss": 0.1544,
+      "step": 18647
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3044435715874434,
+      "learning_rate": 1.3994808985557497e-07,
+      "loss": 0.1595,
+      "step": 18648
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.865404252667061,
+      "learning_rate": 1.3967364984152965e-07,
+      "loss": 0.1569,
+      "step": 18649
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1375610633591677,
+      "learning_rate": 1.393994772893925e-07,
+      "loss": 0.1546,
+      "step": 18650
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3318447589511289,
+      "learning_rate": 1.3912557220660206e-07,
+      "loss": 0.1477,
+      "step": 18651
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.155841518237319,
+      "learning_rate": 1.3885193460058676e-07,
+      "loss": 0.1355,
+      "step": 18652
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.148309023639851,
+      "learning_rate": 1.3857856447876962e-07,
+      "loss": 0.1764,
+      "step": 18653
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.087533529247148,
+      "learning_rate": 1.3830546184856687e-07,
+      "loss": 0.1581,
+      "step": 18654
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.167221914578398,
+      "learning_rate": 1.380326267173848e-07,
+      "loss": 0.1784,
+      "step": 18655
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8385862560019581,
+      "learning_rate": 1.3776005909262423e-07,
+      "loss": 0.1415,
+      "step": 18656
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1988900522919759,
+      "learning_rate": 1.374877589816792e-07,
+      "loss": 0.1595,
+      "step": 18657
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5518572845652296,
+      "learning_rate": 1.3721572639193714e-07,
+      "loss": 0.1569,
+      "step": 18658
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3188198072119175,
+      "learning_rate": 1.3694396133077436e-07,
+      "loss": 0.1524,
+      "step": 18659
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3273904922071935,
+      "learning_rate": 1.3667246380556386e-07,
+      "loss": 0.1603,
+      "step": 18660
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9632445855587911,
+      "learning_rate": 1.3640123382366977e-07,
+      "loss": 0.1464,
+      "step": 18661
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1982344972555086,
+      "learning_rate": 1.361302713924495e-07,
+      "loss": 0.1712,
+      "step": 18662
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.628306264095215,
+      "learning_rate": 1.3585957651925274e-07,
+      "loss": 0.1512,
+      "step": 18663
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7487833080827206,
+      "learning_rate": 1.355891492114214e-07,
+      "loss": 0.1611,
+      "step": 18664
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9688560463974245,
+      "learning_rate": 1.3531898947629296e-07,
+      "loss": 0.161,
+      "step": 18665
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.263410858922274,
+      "learning_rate": 1.3504909732119266e-07,
+      "loss": 0.1701,
+      "step": 18666
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.545720766035433,
+      "learning_rate": 1.347794727534435e-07,
+      "loss": 0.1691,
+      "step": 18667
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.280358835845226,
+      "learning_rate": 1.3451011578035856e-07,
+      "loss": 0.1576,
+      "step": 18668
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.018370216604535,
+      "learning_rate": 1.3424102640924307e-07,
+      "loss": 0.1566,
+      "step": 18669
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.649240179706749,
+      "learning_rate": 1.339722046473979e-07,
+      "loss": 0.1624,
+      "step": 18670
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.585373607840606,
+      "learning_rate": 1.3370365050211387e-07,
+      "loss": 0.1445,
+      "step": 18671
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9880163861646013,
+      "learning_rate": 1.3343536398067513e-07,
+      "loss": 0.1682,
+      "step": 18672
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1272581002377124,
+      "learning_rate": 1.3316734509035922e-07,
+      "loss": 0.1606,
+      "step": 18673
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9879019391840964,
+      "learning_rate": 1.3289959383843698e-07,
+      "loss": 0.1414,
+      "step": 18674
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9941590979821857,
+      "learning_rate": 1.3263211023217038e-07,
+      "loss": 0.1523,
+      "step": 18675
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0029671973975969,
+      "learning_rate": 1.323648942788147e-07,
+      "loss": 0.1646,
+      "step": 18676
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9179981656127324,
+      "learning_rate": 1.3209794598561864e-07,
+      "loss": 0.1772,
+      "step": 18677
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9915365118700741,
+      "learning_rate": 1.3183126535982306e-07,
+      "loss": 0.1552,
+      "step": 18678
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6295534645843557,
+      "learning_rate": 1.3156485240866213e-07,
+      "loss": 0.1526,
+      "step": 18679
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1206894933896525,
+      "learning_rate": 1.312987071393612e-07,
+      "loss": 0.165,
+      "step": 18680
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1012662595538143,
+      "learning_rate": 1.310328295591412e-07,
+      "loss": 0.171,
+      "step": 18681
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.112825275192214,
+      "learning_rate": 1.30767219675213e-07,
+      "loss": 0.1534,
+      "step": 18682
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.042100924816748,
+      "learning_rate": 1.3050187749478192e-07,
+      "loss": 0.1683,
+      "step": 18683
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.4419620222320186,
+      "learning_rate": 1.3023680302504338e-07,
+      "loss": 0.1751,
+      "step": 18684
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9906677459433227,
+      "learning_rate": 1.2997199627319047e-07,
+      "loss": 0.1514,
+      "step": 18685
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9638069201467271,
+      "learning_rate": 1.29707457246403e-07,
+      "loss": 0.1733,
+      "step": 18686
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1689158639836776,
+      "learning_rate": 1.2944318595185855e-07,
+      "loss": 0.1571,
+      "step": 18687
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9465121862654976,
+      "learning_rate": 1.29179182396727e-07,
+      "loss": 0.1454,
+      "step": 18688
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.490398094105073,
+      "learning_rate": 1.289154465881659e-07,
+      "loss": 0.1399,
+      "step": 18689
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0151831663290165,
+      "learning_rate": 1.2865197853333179e-07,
+      "loss": 0.1606,
+      "step": 18690
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0566143989903571,
+      "learning_rate": 1.283887782393689e-07,
+      "loss": 0.1635,
+      "step": 18691
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.450115053552093,
+      "learning_rate": 1.2812584571341936e-07,
+      "loss": 0.1679,
+      "step": 18692
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.32553033080943,
+      "learning_rate": 1.2786318096261298e-07,
+      "loss": 0.1729,
+      "step": 18693
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0099512930173309,
+      "learning_rate": 1.2760078399407626e-07,
+      "loss": 0.1483,
+      "step": 18694
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0971077485331466,
+      "learning_rate": 1.273386548149247e-07,
+      "loss": 0.1406,
+      "step": 18695
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9603529997571104,
+      "learning_rate": 1.270767934322703e-07,
+      "loss": 0.1662,
+      "step": 18696
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7585292023883128,
+      "learning_rate": 1.2681519985321522e-07,
+      "loss": 0.1954,
+      "step": 18697
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3029394904526037,
+      "learning_rate": 1.2655387408485597e-07,
+      "loss": 0.1469,
+      "step": 18698
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9517084868877973,
+      "learning_rate": 1.2629281613428024e-07,
+      "loss": 0.1617,
+      "step": 18699
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8858492726075213,
+      "learning_rate": 1.260320260085701e-07,
+      "loss": 0.1501,
+      "step": 18700
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.061871176835859,
+      "learning_rate": 1.2577150371479884e-07,
+      "loss": 0.1422,
+      "step": 18701
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2908260521128192,
+      "learning_rate": 1.25511249260033e-07,
+      "loss": 0.165,
+      "step": 18702
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1293124330242905,
+      "learning_rate": 1.2525126265133137e-07,
+      "loss": 0.1873,
+      "step": 18703
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.085955564601038,
+      "learning_rate": 1.249915438957483e-07,
+      "loss": 0.1599,
+      "step": 18704
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.021664792923056,
+      "learning_rate": 1.2473209300032706e-07,
+      "loss": 0.1607,
+      "step": 18705
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3178651427571242,
+      "learning_rate": 1.2447290997210426e-07,
+      "loss": 0.1552,
+      "step": 18706
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.5516786445283004,
+      "learning_rate": 1.2421399481811313e-07,
+      "loss": 0.1702,
+      "step": 18707
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.329356585607747,
+      "learning_rate": 1.2395534754537475e-07,
+      "loss": 0.1616,
+      "step": 18708
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1807373706580682,
+      "learning_rate": 1.236969681609057e-07,
+      "loss": 0.1495,
+      "step": 18709
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5458064139687657,
+      "learning_rate": 1.2343885667171373e-07,
+      "loss": 0.1954,
+      "step": 18710
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.94979791393677,
+      "learning_rate": 1.23181013084801e-07,
+      "loss": 0.1673,
+      "step": 18711
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3667713860642676,
+      "learning_rate": 1.2292343740715973e-07,
+      "loss": 0.1695,
+      "step": 18712
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8269227975604101,
+      "learning_rate": 1.2266612964577984e-07,
+      "loss": 0.1403,
+      "step": 18713
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1395132914278172,
+      "learning_rate": 1.22409089807638e-07,
+      "loss": 0.1547,
+      "step": 18714
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3599861623733471,
+      "learning_rate": 1.221523178997075e-07,
+      "loss": 0.1406,
+      "step": 18715
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1998746875068946,
+      "learning_rate": 1.2189581392895388e-07,
+      "loss": 0.1695,
+      "step": 18716
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0641164731288262,
+      "learning_rate": 1.2163957790233382e-07,
+      "loss": 0.1664,
+      "step": 18717
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.188292593607651,
+      "learning_rate": 1.2138360982679842e-07,
+      "loss": 0.1845,
+      "step": 18718
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.253121358483849,
+      "learning_rate": 1.21127909709291e-07,
+      "loss": 0.1712,
+      "step": 18719
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9717715577170544,
+      "learning_rate": 1.2087247755674603e-07,
+      "loss": 0.1675,
+      "step": 18720
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.258572516156622,
+      "learning_rate": 1.206173133760935e-07,
+      "loss": 0.1513,
+      "step": 18721
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0221084000142802,
+      "learning_rate": 1.2036241717425456e-07,
+      "loss": 0.1663,
+      "step": 18722
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2012471054237224,
+      "learning_rate": 1.2010778895814258e-07,
+      "loss": 0.1562,
+      "step": 18723
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.594584266408908,
+      "learning_rate": 1.1985342873466532e-07,
+      "loss": 0.178,
+      "step": 18724
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.201800662850403,
+      "learning_rate": 1.1959933651072065e-07,
+      "loss": 0.1562,
+      "step": 18725
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.910808606905359,
+      "learning_rate": 1.1934551229320413e-07,
+      "loss": 0.1586,
+      "step": 18726
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.004299274296927,
+      "learning_rate": 1.1909195608899694e-07,
+      "loss": 0.1672,
+      "step": 18727
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0095444901687516,
+      "learning_rate": 1.1883866790497911e-07,
+      "loss": 0.1509,
+      "step": 18728
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8761406814468267,
+      "learning_rate": 1.1858564774802073e-07,
+      "loss": 0.1589,
+      "step": 18729
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.034883019557606,
+      "learning_rate": 1.1833289562498406e-07,
+      "loss": 0.1614,
+      "step": 18730
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1760568222090761,
+      "learning_rate": 1.1808041154272587e-07,
+      "loss": 0.1413,
+      "step": 18731
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8505003865615471,
+      "learning_rate": 1.178281955080951e-07,
+      "loss": 0.1571,
+      "step": 18732
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9717764432364998,
+      "learning_rate": 1.1757624752793184e-07,
+      "loss": 0.1698,
+      "step": 18733
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1320680343929483,
+      "learning_rate": 1.1732456760907174e-07,
+      "loss": 0.1648,
+      "step": 18734
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9709646357714493,
+      "learning_rate": 1.1707315575834044e-07,
+      "loss": 0.1525,
+      "step": 18735
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8168264220250449,
+      "learning_rate": 1.1682201198255916e-07,
+      "loss": 0.1456,
+      "step": 18736
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.953615837573349,
+      "learning_rate": 1.16571136288538e-07,
+      "loss": 0.1477,
+      "step": 18737
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1923327503063306,
+      "learning_rate": 1.1632052868308375e-07,
+      "loss": 0.1763,
+      "step": 18738
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0083921240352136,
+      "learning_rate": 1.1607018917299207e-07,
+      "loss": 0.1484,
+      "step": 18739
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3179243790087516,
+      "learning_rate": 1.158201177650553e-07,
+      "loss": 0.1573,
+      "step": 18740
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2500627713525843,
+      "learning_rate": 1.1557031446605693e-07,
+      "loss": 0.1505,
+      "step": 18741
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.239246970452079,
+      "learning_rate": 1.1532077928277152e-07,
+      "loss": 0.1716,
+      "step": 18742
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8462484777876413,
+      "learning_rate": 1.1507151222196811e-07,
+      "loss": 0.1404,
+      "step": 18743
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.059018702238596,
+      "learning_rate": 1.1482251329040795e-07,
+      "loss": 0.1819,
+      "step": 18744
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4355337392067744,
+      "learning_rate": 1.1457378249484674e-07,
+      "loss": 0.1705,
+      "step": 18745
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9243722354509532,
+      "learning_rate": 1.143253198420291e-07,
+      "loss": 0.1648,
+      "step": 18746
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7227231193621009,
+      "learning_rate": 1.1407712533869519e-07,
+      "loss": 0.152,
+      "step": 18747
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1571786371702186,
+      "learning_rate": 1.1382919899157852e-07,
+      "loss": 0.1618,
+      "step": 18748
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0542535821163654,
+      "learning_rate": 1.1358154080740147e-07,
+      "loss": 0.1438,
+      "step": 18749
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.031459382470392,
+      "learning_rate": 1.1333415079288424e-07,
+      "loss": 0.1706,
+      "step": 18750
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1376771610307743,
+      "learning_rate": 1.130870289547381e-07,
+      "loss": 0.1569,
+      "step": 18751
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1361995973113577,
+      "learning_rate": 1.1284017529966329e-07,
+      "loss": 0.1689,
+      "step": 18752
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9430514064426113,
+      "learning_rate": 1.1259358983435775e-07,
+      "loss": 0.1825,
+      "step": 18753
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0092812343386752,
+      "learning_rate": 1.1234727256550837e-07,
+      "loss": 0.1763,
+      "step": 18754
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8164572156621775,
+      "learning_rate": 1.1210122349979979e-07,
+      "loss": 0.154,
+      "step": 18755
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.327723266994876,
+      "learning_rate": 1.1185544264390225e-07,
+      "loss": 0.1733,
+      "step": 18756
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6508728885834751,
+      "learning_rate": 1.1160993000448372e-07,
+      "loss": 0.1735,
+      "step": 18757
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9180275397058472,
+      "learning_rate": 1.1136468558820668e-07,
+      "loss": 0.1582,
+      "step": 18758
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.101850420181442,
+      "learning_rate": 1.1111970940171912e-07,
+      "loss": 0.1516,
+      "step": 18759
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1555854613546752,
+      "learning_rate": 1.1087500145166908e-07,
+      "loss": 0.1411,
+      "step": 18760
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2756787789402384,
+      "learning_rate": 1.1063056174469234e-07,
+      "loss": 0.1396,
+      "step": 18761
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4078874087782576,
+      "learning_rate": 1.1038639028742138e-07,
+      "loss": 0.194,
+      "step": 18762
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8496709042888108,
+      "learning_rate": 1.1014248708647645e-07,
+      "loss": 0.1247,
+      "step": 18763
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8844652923011713,
+      "learning_rate": 1.0989885214847673e-07,
+      "loss": 0.1586,
+      "step": 18764
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.001028423834968,
+      "learning_rate": 1.0965548548002802e-07,
+      "loss": 0.1641,
+      "step": 18765
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.865470853867464,
+      "learning_rate": 1.0941238708773283e-07,
+      "loss": 0.1806,
+      "step": 18766
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.408923759241474,
+      "learning_rate": 1.0916955697818587e-07,
+      "loss": 0.1504,
+      "step": 18767
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2024386779570817,
+      "learning_rate": 1.0892699515797411e-07,
+      "loss": 0.1791,
+      "step": 18768
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0923015100986826,
+      "learning_rate": 1.086847016336745e-07,
+      "loss": 0.1392,
+      "step": 18769
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8647220030074458,
+      "learning_rate": 1.084426764118629e-07,
+      "loss": 0.1486,
+      "step": 18770
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7755338691389857,
+      "learning_rate": 1.0820091949910072e-07,
+      "loss": 0.1672,
+      "step": 18771
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8298954613901326,
+      "learning_rate": 1.0795943090194827e-07,
+      "loss": 0.1649,
+      "step": 18772
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9795338903069379,
+      "learning_rate": 1.0771821062695476e-07,
+      "loss": 0.1489,
+      "step": 18773
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0089380968964687,
+      "learning_rate": 1.0747725868066383e-07,
+      "loss": 0.151,
+      "step": 18774
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2822032644898074,
+      "learning_rate": 1.0723657506961027e-07,
+      "loss": 0.172,
+      "step": 18775
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7322298865698078,
+      "learning_rate": 1.069961598003233e-07,
+      "loss": 0.1599,
+      "step": 18776
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.046246918063842,
+      "learning_rate": 1.0675601287932547e-07,
+      "loss": 0.1966,
+      "step": 18777
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.328086242939519,
+      "learning_rate": 1.0651613431312824e-07,
+      "loss": 0.1464,
+      "step": 18778
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.294347819359744,
+      "learning_rate": 1.0627652410823975e-07,
+      "loss": 0.1369,
+      "step": 18779
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7989102185537711,
+      "learning_rate": 1.0603718227116034e-07,
+      "loss": 0.1613,
+      "step": 18780
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.036952896112565,
+      "learning_rate": 1.0579810880838037e-07,
+      "loss": 0.1639,
+      "step": 18781
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2476287988888424,
+      "learning_rate": 1.0555930372638578e-07,
+      "loss": 0.1744,
+      "step": 18782
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.37060572453245,
+      "learning_rate": 1.0532076703165362e-07,
+      "loss": 0.1899,
+      "step": 18783
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9596618628758464,
+      "learning_rate": 1.0508249873065424e-07,
+      "loss": 0.15,
+      "step": 18784
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.185689083060517,
+      "learning_rate": 1.0484449882985138e-07,
+      "loss": 0.1541,
+      "step": 18785
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8349670685171569,
+      "learning_rate": 1.0460676733570096e-07,
+      "loss": 0.166,
+      "step": 18786
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0295126853380023,
+      "learning_rate": 1.0436930425465008e-07,
+      "loss": 0.1874,
+      "step": 18787
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0593615747004992,
+      "learning_rate": 1.0413210959314135e-07,
+      "loss": 0.1595,
+      "step": 18788
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.168075177134784,
+      "learning_rate": 1.038951833576074e-07,
+      "loss": 0.1723,
+      "step": 18789
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1218155857630336,
+      "learning_rate": 1.0365852555447642e-07,
+      "loss": 0.1753,
+      "step": 18790
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8050559077961082,
+      "learning_rate": 1.0342213619016661e-07,
+      "loss": 0.1587,
+      "step": 18791
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9181660602621501,
+      "learning_rate": 1.0318601527108952e-07,
+      "loss": 0.1496,
+      "step": 18792
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1033877829228513,
+      "learning_rate": 1.0295016280365111e-07,
+      "loss": 0.1379,
+      "step": 18793
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9590634235469165,
+      "learning_rate": 1.0271457879424851e-07,
+      "loss": 0.1546,
+      "step": 18794
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.952709206529511,
+      "learning_rate": 1.0247926324927215e-07,
+      "loss": 0.1687,
+      "step": 18795
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1759733030062076,
+      "learning_rate": 1.0224421617510471e-07,
+      "loss": 0.1583,
+      "step": 18796
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2763520910574768,
+      "learning_rate": 1.020094375781222e-07,
+      "loss": 0.1571,
+      "step": 18797
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7511259754137039,
+      "learning_rate": 1.0177492746469286e-07,
+      "loss": 0.1676,
+      "step": 18798
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.148516859847391,
+      "learning_rate": 1.0154068584117716e-07,
+      "loss": 0.1773,
+      "step": 18799
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9627330663214961,
+      "learning_rate": 1.0130671271392889e-07,
+      "loss": 0.141,
+      "step": 18800
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9745604777780793,
+      "learning_rate": 1.0107300808929522e-07,
+      "loss": 0.1718,
+      "step": 18801
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8737543564782753,
+      "learning_rate": 1.008395719736166e-07,
+      "loss": 0.1648,
+      "step": 18802
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0772967372661506,
+      "learning_rate": 1.0060640437322244e-07,
+      "loss": 0.1645,
+      "step": 18803
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4726543395679172,
+      "learning_rate": 1.003735052944399e-07,
+      "loss": 0.1848,
+      "step": 18804
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.169918964022628,
+      "learning_rate": 1.0014087474358392e-07,
+      "loss": 0.1543,
+      "step": 18805
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.216989418510401,
+      "learning_rate": 9.990851272696722e-08,
+      "loss": 0.1576,
+      "step": 18806
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1686919692769921,
+      "learning_rate": 9.967641925089033e-08,
+      "loss": 0.173,
+      "step": 18807
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4996681879717797,
+      "learning_rate": 9.944459432165044e-08,
+      "loss": 0.1664,
+      "step": 18808
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1045459523269272,
+      "learning_rate": 9.92130379455336e-08,
+      "loss": 0.1522,
+      "step": 18809
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9910006468676175,
+      "learning_rate": 9.89817501288226e-08,
+      "loss": 0.1373,
+      "step": 18810
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1419457094601917,
+      "learning_rate": 9.875073087779241e-08,
+      "loss": 0.1856,
+      "step": 18811
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0739981086468395,
+      "learning_rate": 9.85199801987069e-08,
+      "loss": 0.1716,
+      "step": 18812
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9680114664815531,
+      "learning_rate": 9.828949809782662e-08,
+      "loss": 0.1678,
+      "step": 18813
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0786498296590985,
+      "learning_rate": 9.805928458140212e-08,
+      "loss": 0.1591,
+      "step": 18814
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.64021686468835,
+      "learning_rate": 9.782933965567953e-08,
+      "loss": 0.1626,
+      "step": 18815
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1278692654493154,
+      "learning_rate": 9.759966332689497e-08,
+      "loss": 0.1598,
+      "step": 18816
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1266880264484171,
+      "learning_rate": 9.737025560127899e-08,
+      "loss": 0.1598,
+      "step": 18817
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1920107831865574,
+      "learning_rate": 9.714111648505442e-08,
+      "loss": 0.1857,
+      "step": 18818
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1175544268483446,
+      "learning_rate": 9.691224598443515e-08,
+      "loss": 0.1506,
+      "step": 18819
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0988902168208523,
+      "learning_rate": 9.66836441056318e-08,
+      "loss": 0.1727,
+      "step": 18820
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.826008174194182,
+      "learning_rate": 9.645531085484383e-08,
+      "loss": 0.1501,
+      "step": 18821
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.292188236065506,
+      "learning_rate": 9.622724623826407e-08,
+      "loss": 0.1771,
+      "step": 18822
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4458397407326542,
+      "learning_rate": 9.59994502620809e-08,
+      "loss": 0.1767,
+      "step": 18823
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2335768493154349,
+      "learning_rate": 9.57719229324705e-08,
+      "loss": 0.17,
+      "step": 18824
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9917330277547299,
+      "learning_rate": 9.554466425560793e-08,
+      "loss": 0.1501,
+      "step": 18825
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9023946695552103,
+      "learning_rate": 9.531767423765381e-08,
+      "loss": 0.1746,
+      "step": 18826
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1679938172110933,
+      "learning_rate": 9.509095288476767e-08,
+      "loss": 0.1623,
+      "step": 18827
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3085712269367609,
+      "learning_rate": 9.486450020310011e-08,
+      "loss": 0.1536,
+      "step": 18828
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1529943097730873,
+      "learning_rate": 9.463831619879183e-08,
+      "loss": 0.1537,
+      "step": 18829
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0872766259105977,
+      "learning_rate": 9.441240087797787e-08,
+      "loss": 0.172,
+      "step": 18830
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0909881135030688,
+      "learning_rate": 9.41867542467878e-08,
+      "loss": 0.1737,
+      "step": 18831
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8379483903814336,
+      "learning_rate": 9.396137631134116e-08,
+      "loss": 0.1478,
+      "step": 18832
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3761318998390024,
+      "learning_rate": 9.373626707775196e-08,
+      "loss": 0.1652,
+      "step": 18833
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2279841676392231,
+      "learning_rate": 9.351142655212642e-08,
+      "loss": 0.1867,
+      "step": 18834
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0752650044873828,
+      "learning_rate": 9.328685474056187e-08,
+      "loss": 0.1638,
+      "step": 18835
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2379873953884408,
+      "learning_rate": 9.306255164915123e-08,
+      "loss": 0.1366,
+      "step": 18836
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2130097579213732,
+      "learning_rate": 9.283851728397853e-08,
+      "loss": 0.1628,
+      "step": 18837
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8947669982322405,
+      "learning_rate": 9.26147516511211e-08,
+      "loss": 0.1507,
+      "step": 18838
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0036150550246519,
+      "learning_rate": 9.239125475664746e-08,
+      "loss": 0.159,
+      "step": 18839
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0657812870787666,
+      "learning_rate": 9.216802660662161e-08,
+      "loss": 0.1743,
+      "step": 18840
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9154068268279001,
+      "learning_rate": 9.194506720709651e-08,
+      "loss": 0.1747,
+      "step": 18841
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9264247298220879,
+      "learning_rate": 9.172237656412175e-08,
+      "loss": 0.1458,
+      "step": 18842
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9801921323891072,
+      "learning_rate": 9.149995468373696e-08,
+      "loss": 0.1621,
+      "step": 18843
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1632621863958537,
+      "learning_rate": 9.127780157197619e-08,
+      "loss": 0.1676,
+      "step": 18844
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1317424244696488,
+      "learning_rate": 9.105591723486352e-08,
+      "loss": 0.1542,
+      "step": 18845
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0736880815431518,
+      "learning_rate": 9.083430167841856e-08,
+      "loss": 0.1511,
+      "step": 18846
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0161957878733108,
+      "learning_rate": 9.061295490865429e-08,
+      "loss": 0.1674,
+      "step": 18847
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7928674453212935,
+      "learning_rate": 9.039187693157147e-08,
+      "loss": 0.1398,
+      "step": 18848
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9045398606870544,
+      "learning_rate": 9.017106775317086e-08,
+      "loss": 0.1855,
+      "step": 18849
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0535097383038936,
+      "learning_rate": 8.995052737943766e-08,
+      "loss": 0.163,
+      "step": 18850
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4496573427182988,
+      "learning_rate": 8.973025581635819e-08,
+      "loss": 0.1817,
+      "step": 18851
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0777957220050927,
+      "learning_rate": 8.951025306990324e-08,
+      "loss": 0.1563,
+      "step": 18852
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0506377567143983,
+      "learning_rate": 8.929051914604359e-08,
+      "loss": 0.1916,
+      "step": 18853
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.002996556765619,
+      "learning_rate": 8.907105405073779e-08,
+      "loss": 0.1759,
+      "step": 18854
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.998730610245173,
+      "learning_rate": 8.885185778993999e-08,
+      "loss": 0.1495,
+      "step": 18855
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2495765988017125,
+      "learning_rate": 8.863293036959431e-08,
+      "loss": 0.165,
+      "step": 18856
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0130819408578617,
+      "learning_rate": 8.841427179564154e-08,
+      "loss": 0.146,
+      "step": 18857
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8272516627383144,
+      "learning_rate": 8.819588207401142e-08,
+      "loss": 0.1715,
+      "step": 18858
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9948517329878664,
+      "learning_rate": 8.797776121062696e-08,
+      "loss": 0.1534,
+      "step": 18859
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3309641860820505,
+      "learning_rate": 8.775990921140565e-08,
+      "loss": 0.1656,
+      "step": 18860
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.444462781797579,
+      "learning_rate": 8.754232608225722e-08,
+      "loss": 0.1625,
+      "step": 18861
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8860214555939581,
+      "learning_rate": 8.732501182908249e-08,
+      "loss": 0.1835,
+      "step": 18862
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.300447118415056,
+      "learning_rate": 8.710796645777674e-08,
+      "loss": 0.1569,
+      "step": 18863
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.011564783440716,
+      "learning_rate": 8.689118997422752e-08,
+      "loss": 0.1413,
+      "step": 18864
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.543969469262822,
+      "learning_rate": 8.667468238431453e-08,
+      "loss": 0.151,
+      "step": 18865
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6519119012480639,
+      "learning_rate": 8.645844369391088e-08,
+      "loss": 0.155,
+      "step": 18866
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.790938070722103,
+      "learning_rate": 8.624247390888186e-08,
+      "loss": 0.1392,
+      "step": 18867
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.509865696614438,
+      "learning_rate": 8.602677303508611e-08,
+      "loss": 0.1826,
+      "step": 18868
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4021515201833346,
+      "learning_rate": 8.581134107837341e-08,
+      "loss": 0.1487,
+      "step": 18869
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6080350955718345,
+      "learning_rate": 8.559617804458798e-08,
+      "loss": 0.1771,
+      "step": 18870
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9752358089227612,
+      "learning_rate": 8.538128393956624e-08,
+      "loss": 0.1662,
+      "step": 18871
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3198701123217553,
+      "learning_rate": 8.516665876913687e-08,
+      "loss": 0.1725,
+      "step": 18872
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9611312290629432,
+      "learning_rate": 8.49523025391219e-08,
+      "loss": 0.1626,
+      "step": 18873
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2329227273918022,
+      "learning_rate": 8.473821525533665e-08,
+      "loss": 0.1747,
+      "step": 18874
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9490149018446066,
+      "learning_rate": 8.452439692358649e-08,
+      "loss": 0.172,
+      "step": 18875
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0099470775860748,
+      "learning_rate": 8.431084754967345e-08,
+      "loss": 0.1378,
+      "step": 18876
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7531190523770079,
+      "learning_rate": 8.409756713938733e-08,
+      "loss": 0.1473,
+      "step": 18877
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0347859261232761,
+      "learning_rate": 8.388455569851461e-08,
+      "loss": 0.1719,
+      "step": 18878
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8585341785400499,
+      "learning_rate": 8.367181323283402e-08,
+      "loss": 0.1678,
+      "step": 18879
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9211259975521868,
+      "learning_rate": 8.345933974811537e-08,
+      "loss": 0.1711,
+      "step": 18880
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.208502773218923,
+      "learning_rate": 8.324713525012185e-08,
+      "loss": 0.1451,
+      "step": 18881
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.854440510579316,
+      "learning_rate": 8.303519974461106e-08,
+      "loss": 0.1565,
+      "step": 18882
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9862239698761738,
+      "learning_rate": 8.28235332373295e-08,
+      "loss": 0.1941,
+      "step": 18883
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5255599255822914,
+      "learning_rate": 8.261213573402038e-08,
+      "loss": 0.159,
+      "step": 18884
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6668881972462515,
+      "learning_rate": 8.240100724041689e-08,
+      "loss": 0.1566,
+      "step": 18885
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9173579531264667,
+      "learning_rate": 8.219014776224555e-08,
+      "loss": 0.1594,
+      "step": 18886
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4391060778764997,
+      "learning_rate": 8.197955730522733e-08,
+      "loss": 0.1432,
+      "step": 18887
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7719058427891368,
+      "learning_rate": 8.176923587507434e-08,
+      "loss": 0.1642,
+      "step": 18888
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9890541847862871,
+      "learning_rate": 8.155918347748981e-08,
+      "loss": 0.15,
+      "step": 18889
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2971062282347046,
+      "learning_rate": 8.13494001181725e-08,
+      "loss": 0.1396,
+      "step": 18890
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.89297258092725,
+      "learning_rate": 8.113988580281451e-08,
+      "loss": 0.1593,
+      "step": 18891
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1064976604626795,
+      "learning_rate": 8.093064053709576e-08,
+      "loss": 0.1901,
+      "step": 18892
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9866375316992192,
+      "learning_rate": 8.072166432669503e-08,
+      "loss": 0.1433,
+      "step": 18893
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1940034352979834,
+      "learning_rate": 8.051295717727892e-08,
+      "loss": 0.1657,
+      "step": 18894
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1200969835311203,
+      "learning_rate": 8.030451909450842e-08,
+      "loss": 0.1698,
+      "step": 18895
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4641811009107635,
+      "learning_rate": 8.009635008403904e-08,
+      "loss": 0.1399,
+      "step": 18896
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9307287020748609,
+      "learning_rate": 7.988845015151513e-08,
+      "loss": 0.1617,
+      "step": 18897
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2935854651952166,
+      "learning_rate": 7.968081930257887e-08,
+      "loss": 0.1524,
+      "step": 18898
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1529460938502556,
+      "learning_rate": 7.947345754285906e-08,
+      "loss": 0.1661,
+      "step": 18899
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6332754029727536,
+      "learning_rate": 7.926636487798345e-08,
+      "loss": 0.1597,
+      "step": 18900
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9341773562612538,
+      "learning_rate": 7.905954131356752e-08,
+      "loss": 0.1714,
+      "step": 18901
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7598173351821158,
+      "learning_rate": 7.885298685522235e-08,
+      "loss": 0.1494,
+      "step": 18902
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9162616507880349,
+      "learning_rate": 7.864670150854903e-08,
+      "loss": 0.1413,
+      "step": 18903
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9287105319838184,
+      "learning_rate": 7.844068527914528e-08,
+      "loss": 0.1839,
+      "step": 18904
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.144598902170974,
+      "learning_rate": 7.823493817259776e-08,
+      "loss": 0.1772,
+      "step": 18905
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.147242897737404,
+      "learning_rate": 7.802946019448864e-08,
+      "loss": 0.1854,
+      "step": 18906
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2450550086501007,
+      "learning_rate": 7.782425135039018e-08,
+      "loss": 0.1666,
+      "step": 18907
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0442645415313312,
+      "learning_rate": 7.7619311645869e-08,
+      "loss": 0.1513,
+      "step": 18908
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9659641393581991,
+      "learning_rate": 7.741464108648511e-08,
+      "loss": 0.1479,
+      "step": 18909
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0802928177377895,
+      "learning_rate": 7.721023967778962e-08,
+      "loss": 0.1478,
+      "step": 18910
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2944293111076215,
+      "learning_rate": 7.700610742532588e-08,
+      "loss": 0.165,
+      "step": 18911
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1730839439653735,
+      "learning_rate": 7.68022443346328e-08,
+      "loss": 0.1776,
+      "step": 18912
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9910193731834973,
+      "learning_rate": 7.659865041123926e-08,
+      "loss": 0.1662,
+      "step": 18913
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.056809813408625,
+      "learning_rate": 7.639532566066755e-08,
+      "loss": 0.1683,
+      "step": 18914
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0770889532907564,
+      "learning_rate": 7.619227008843322e-08,
+      "loss": 0.1468,
+      "step": 18915
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.033875052227631,
+      "learning_rate": 7.598948370004412e-08,
+      "loss": 0.1914,
+      "step": 18916
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0503186323084912,
+      "learning_rate": 7.578696650100025e-08,
+      "loss": 0.1652,
+      "step": 18917
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.061638624902994,
+      "learning_rate": 7.558471849679616e-08,
+      "loss": 0.1609,
+      "step": 18918
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.874059333965608,
+      "learning_rate": 7.538273969291743e-08,
+      "loss": 0.1628,
+      "step": 18919
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9470801203586013,
+      "learning_rate": 7.518103009484079e-08,
+      "loss": 0.1403,
+      "step": 18920
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.200812196227737,
+      "learning_rate": 7.497958970804076e-08,
+      "loss": 0.1667,
+      "step": 18921
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0078825382674224,
+      "learning_rate": 7.477841853797963e-08,
+      "loss": 0.1674,
+      "step": 18922
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1759053898946894,
+      "learning_rate": 7.457751659011414e-08,
+      "loss": 0.1676,
+      "step": 18923
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8222529337220591,
+      "learning_rate": 7.437688386989438e-08,
+      "loss": 0.1599,
+      "step": 18924
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8533396031334325,
+      "learning_rate": 7.417652038276157e-08,
+      "loss": 0.15,
+      "step": 18925
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1056761413519485,
+      "learning_rate": 7.397642613415245e-08,
+      "loss": 0.1585,
+      "step": 18926
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3920512039464155,
+      "learning_rate": 7.37766011294927e-08,
+      "loss": 0.1554,
+      "step": 18927
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5934123887083915,
+      "learning_rate": 7.357704537420351e-08,
+      "loss": 0.1708,
+      "step": 18928
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2091051382121403,
+      "learning_rate": 7.337775887369725e-08,
+      "loss": 0.1646,
+      "step": 18929
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.148339743268307,
+      "learning_rate": 7.317874163338068e-08,
+      "loss": 0.1679,
+      "step": 18930
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0786303167088929,
+      "learning_rate": 7.29799936586506e-08,
+      "loss": 0.1727,
+      "step": 18931
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.132990580590024,
+      "learning_rate": 7.278151495489938e-08,
+      "loss": 0.1621,
+      "step": 18932
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.839523001421132,
+      "learning_rate": 7.258330552750936e-08,
+      "loss": 0.1708,
+      "step": 18933
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1078873139031478,
+      "learning_rate": 7.238536538185848e-08,
+      "loss": 0.1629,
+      "step": 18934
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.149341920558963,
+      "learning_rate": 7.218769452331576e-08,
+      "loss": 0.1722,
+      "step": 18935
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.08472091194159,
+      "learning_rate": 7.199029295724247e-08,
+      "loss": 0.1674,
+      "step": 18936
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.229463761384355,
+      "learning_rate": 7.179316068899211e-08,
+      "loss": 0.1659,
+      "step": 18937
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4155063754897228,
+      "learning_rate": 7.159629772391485e-08,
+      "loss": 0.1902,
+      "step": 18938
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9655177285562941,
+      "learning_rate": 7.139970406734753e-08,
+      "loss": 0.1753,
+      "step": 18939
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.118473850390554,
+      "learning_rate": 7.120337972462365e-08,
+      "loss": 0.1529,
+      "step": 18940
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.241518882169247,
+      "learning_rate": 7.100732470107008e-08,
+      "loss": 0.156,
+      "step": 18941
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.913325622858883,
+      "learning_rate": 7.081153900200255e-08,
+      "loss": 0.1557,
+      "step": 18942
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2350006356643224,
+      "learning_rate": 7.061602263273237e-08,
+      "loss": 0.172,
+      "step": 18943
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1409555515039296,
+      "learning_rate": 7.04207755985653e-08,
+      "loss": 0.159,
+      "step": 18944
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8621408823943831,
+      "learning_rate": 7.022579790479378e-08,
+      "loss": 0.1437,
+      "step": 18945
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8696956434949746,
+      "learning_rate": 7.003108955670911e-08,
+      "loss": 0.1808,
+      "step": 18946
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0370223435433659,
+      "learning_rate": 6.983665055959154e-08,
+      "loss": 0.1462,
+      "step": 18947
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0713365572057763,
+      "learning_rate": 6.964248091871683e-08,
+      "loss": 0.1703,
+      "step": 18948
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9865314911287251,
+      "learning_rate": 6.944858063934967e-08,
+      "loss": 0.1432,
+      "step": 18949
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0416169563655078,
+      "learning_rate": 6.925494972675029e-08,
+      "loss": 0.1425,
+      "step": 18950
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2588921376373747,
+      "learning_rate": 6.906158818617226e-08,
+      "loss": 0.1717,
+      "step": 18951
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0330787653455085,
+      "learning_rate": 6.886849602285916e-08,
+      "loss": 0.1566,
+      "step": 18952
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0131289400465249,
+      "learning_rate": 6.867567324204905e-08,
+      "loss": 0.1638,
+      "step": 18953
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8849718675951277,
+      "learning_rate": 6.848311984897216e-08,
+      "loss": 0.1354,
+      "step": 18954
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2590838070351569,
+      "learning_rate": 6.82908358488521e-08,
+      "loss": 0.1656,
+      "step": 18955
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.542926066213727,
+      "learning_rate": 6.809882124690358e-08,
+      "loss": 0.1639,
+      "step": 18956
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3956886472564511,
+      "learning_rate": 6.79070760483358e-08,
+      "loss": 0.1853,
+      "step": 18957
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3101703160806484,
+      "learning_rate": 6.771560025834901e-08,
+      "loss": 0.1963,
+      "step": 18958
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5623705229321303,
+      "learning_rate": 6.752439388213682e-08,
+      "loss": 0.1965,
+      "step": 18959
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.12056540940211,
+      "learning_rate": 6.733345692488736e-08,
+      "loss": 0.1622,
+      "step": 18960
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0543695819488192,
+      "learning_rate": 6.714278939177976e-08,
+      "loss": 0.1504,
+      "step": 18961
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2197686288091252,
+      "learning_rate": 6.695239128798325e-08,
+      "loss": 0.1545,
+      "step": 18962
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.819372886682206,
+      "learning_rate": 6.676226261866591e-08,
+      "loss": 0.1587,
+      "step": 18963
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1616646556897137,
+      "learning_rate": 6.657240338898141e-08,
+      "loss": 0.1658,
+      "step": 18964
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4269534756249098,
+      "learning_rate": 6.638281360408339e-08,
+      "loss": 0.1673,
+      "step": 18965
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.578288524551596,
+      "learning_rate": 6.619349326911218e-08,
+      "loss": 0.143,
+      "step": 18966
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2967738604121724,
+      "learning_rate": 6.600444238920256e-08,
+      "loss": 0.1589,
+      "step": 18967
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9378892547102705,
+      "learning_rate": 6.581566096948377e-08,
+      "loss": 0.1514,
+      "step": 18968
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0222674207787896,
+      "learning_rate": 6.562714901507616e-08,
+      "loss": 0.1551,
+      "step": 18969
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4522802005044373,
+      "learning_rate": 6.543890653109341e-08,
+      "loss": 0.1993,
+      "step": 18970
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9929201580900462,
+      "learning_rate": 6.525093352264145e-08,
+      "loss": 0.1565,
+      "step": 18971
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2075074305444102,
+      "learning_rate": 6.506322999481951e-08,
+      "loss": 0.1568,
+      "step": 18972
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9119396547683619,
+      "learning_rate": 6.487579595271798e-08,
+      "loss": 0.1385,
+      "step": 18973
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9468835404050955,
+      "learning_rate": 6.468863140142279e-08,
+      "loss": 0.1522,
+      "step": 18974
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0360263939289716,
+      "learning_rate": 6.450173634600876e-08,
+      "loss": 0.1544,
+      "step": 18975
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9384445234929053,
+      "learning_rate": 6.431511079154517e-08,
+      "loss": 0.1421,
+      "step": 18976
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9717830054475515,
+      "learning_rate": 6.412875474309688e-08,
+      "loss": 0.143,
+      "step": 18977
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.242839667500306,
+      "learning_rate": 6.394266820571649e-08,
+      "loss": 0.1657,
+      "step": 18978
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8385638418607924,
+      "learning_rate": 6.375685118445329e-08,
+      "loss": 0.1526,
+      "step": 18979
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0274904096006774,
+      "learning_rate": 6.35713036843455e-08,
+      "loss": 0.1462,
+      "step": 18980
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.955038906916783,
+      "learning_rate": 6.338602571042795e-08,
+      "loss": 0.1813,
+      "step": 18981
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4956850451648465,
+      "learning_rate": 6.320101726772665e-08,
+      "loss": 0.1594,
+      "step": 18982
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2377378344046592,
+      "learning_rate": 6.301627836125757e-08,
+      "loss": 0.1779,
+      "step": 18983
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9260561719978769,
+      "learning_rate": 6.283180899603447e-08,
+      "loss": 0.171,
+      "step": 18984
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9400856349983138,
+      "learning_rate": 6.264760917705782e-08,
+      "loss": 0.1605,
+      "step": 18985
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4391237828833978,
+      "learning_rate": 6.246367890932803e-08,
+      "loss": 0.1747,
+      "step": 18986
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0650353212909858,
+      "learning_rate": 6.228001819783113e-08,
+      "loss": 0.1503,
+      "step": 18987
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2264828665242962,
+      "learning_rate": 6.209662704754982e-08,
+      "loss": 0.1679,
+      "step": 18988
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2608897282732527,
+      "learning_rate": 6.191350546346008e-08,
+      "loss": 0.1607,
+      "step": 18989
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0195823417707264,
+      "learning_rate": 6.173065345052687e-08,
+      "loss": 0.1584,
+      "step": 18990
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.3217468456080654,
+      "learning_rate": 6.154807101371063e-08,
+      "loss": 0.1415,
+      "step": 18991
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3982193219953636,
+      "learning_rate": 6.136575815796519e-08,
+      "loss": 0.1604,
+      "step": 18992
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.039154864355294,
+      "learning_rate": 6.118371488823439e-08,
+      "loss": 0.171,
+      "step": 18993
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.449132551538369,
+      "learning_rate": 6.100194120945645e-08,
+      "loss": 0.185,
+      "step": 18994
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7339355163445271,
+      "learning_rate": 6.082043712656305e-08,
+      "loss": 0.1715,
+      "step": 18995
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6051479995279934,
+      "learning_rate": 6.063920264447576e-08,
+      "loss": 0.1468,
+      "step": 18996
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0033063919052598,
+      "learning_rate": 6.045823776811177e-08,
+      "loss": 0.1525,
+      "step": 18997
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9764565540340898,
+      "learning_rate": 6.027754250237938e-08,
+      "loss": 0.1755,
+      "step": 18998
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3922419999429214,
+      "learning_rate": 6.009711685218134e-08,
+      "loss": 0.1729,
+      "step": 18999
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1355283592252898,
+      "learning_rate": 5.991696082240817e-08,
+      "loss": 0.1741,
+      "step": 19000
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0226456454781145,
+      "learning_rate": 5.97370744179504e-08,
+      "loss": 0.1596,
+      "step": 19001
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.022674255084776,
+      "learning_rate": 5.9557457643685257e-08,
+      "loss": 0.1534,
+      "step": 19002
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3787117157844775,
+      "learning_rate": 5.93781105044855e-08,
+      "loss": 0.1637,
+      "step": 19003
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5550964475153704,
+      "learning_rate": 5.9199033005217233e-08,
+      "loss": 0.157,
+      "step": 19004
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2498322296452542,
+      "learning_rate": 5.9020225150735466e-08,
+      "loss": 0.1518,
+      "step": 19005
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5263437302645886,
+      "learning_rate": 5.8841686945891874e-08,
+      "loss": 0.1414,
+      "step": 19006
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0709694859747907,
+      "learning_rate": 5.866341839552814e-08,
+      "loss": 0.1782,
+      "step": 19007
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4903520322067012,
+      "learning_rate": 5.848541950448261e-08,
+      "loss": 0.1648,
+      "step": 19008
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9290190400309384,
+      "learning_rate": 5.830769027757921e-08,
+      "loss": 0.1842,
+      "step": 19009
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1113056250270656,
+      "learning_rate": 5.813023071964296e-08,
+      "loss": 0.1524,
+      "step": 19010
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8503388097141104,
+      "learning_rate": 5.795304083548559e-08,
+      "loss": 0.1451,
+      "step": 19011
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0071646211516418,
+      "learning_rate": 5.777612062991211e-08,
+      "loss": 0.1571,
+      "step": 19012
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9059092713987779,
+      "learning_rate": 5.759947010772426e-08,
+      "loss": 0.1576,
+      "step": 19013
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9766952791494546,
+      "learning_rate": 5.742308927371154e-08,
+      "loss": 0.1701,
+      "step": 19014
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9185439725383592,
+      "learning_rate": 5.7246978132659e-08,
+      "loss": 0.148,
+      "step": 19015
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.419729476220642,
+      "learning_rate": 5.707113668934505e-08,
+      "loss": 0.1695,
+      "step": 19016
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1401721563452156,
+      "learning_rate": 5.6895564948536984e-08,
+      "loss": 0.1611,
+      "step": 19017
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.053616248533265,
+      "learning_rate": 5.672026291499877e-08,
+      "loss": 0.1598,
+      "step": 19018
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.463192697506729,
+      "learning_rate": 5.6545230593484376e-08,
+      "loss": 0.1678,
+      "step": 19019
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9548696767406545,
+      "learning_rate": 5.637046798874335e-08,
+      "loss": 0.1494,
+      "step": 19020
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0769999525229907,
+      "learning_rate": 5.619597510551411e-08,
+      "loss": 0.1516,
+      "step": 19021
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1269093531874508,
+      "learning_rate": 5.6021751948530656e-08,
+      "loss": 0.1603,
+      "step": 19022
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9860154062923643,
+      "learning_rate": 5.58477985225192e-08,
+      "loss": 0.1511,
+      "step": 19023
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3408468901522745,
+      "learning_rate": 5.567411483219709e-08,
+      "loss": 0.1514,
+      "step": 19024
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0965825124126323,
+      "learning_rate": 5.550070088227721e-08,
+      "loss": 0.1573,
+      "step": 19025
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.023824284466048,
+      "learning_rate": 5.532755667746248e-08,
+      "loss": 0.1946,
+      "step": 19026
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0445023365005794,
+      "learning_rate": 5.515468222244913e-08,
+      "loss": 0.1519,
+      "step": 19027
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7889943973151463,
+      "learning_rate": 5.4982077521926744e-08,
+      "loss": 0.1603,
+      "step": 19028
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.113921879902848,
+      "learning_rate": 5.4809742580577144e-08,
+      "loss": 0.1807,
+      "step": 19029
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2384246609536174,
+      "learning_rate": 5.4637677403074355e-08,
+      "loss": 0.1507,
+      "step": 19030
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1053062410513481,
+      "learning_rate": 5.4465881994087976e-08,
+      "loss": 0.159,
+      "step": 19031
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.317934675993351,
+      "learning_rate": 5.4294356358274294e-08,
+      "loss": 0.1587,
+      "step": 19032
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0281026923641639,
+      "learning_rate": 5.4123100500289574e-08,
+      "loss": 0.1613,
+      "step": 19033
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.385724210989773,
+      "learning_rate": 5.395211442477677e-08,
+      "loss": 0.1577,
+      "step": 19034
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.002683221703717,
+      "learning_rate": 5.3781398136374394e-08,
+      "loss": 0.1447,
+      "step": 19035
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3531188851630518,
+      "learning_rate": 5.3610951639714305e-08,
+      "loss": 0.1619,
+      "step": 19036
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7890219249827146,
+      "learning_rate": 5.344077493941835e-08,
+      "loss": 0.1771,
+      "step": 19037
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1012008101422308,
+      "learning_rate": 5.327086804010284e-08,
+      "loss": 0.1614,
+      "step": 19038
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0428208001568597,
+      "learning_rate": 5.3101230946376314e-08,
+      "loss": 0.1754,
+      "step": 19039
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.026144332646727,
+      "learning_rate": 5.2931863662841755e-08,
+      "loss": 0.1688,
+      "step": 19040
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.603608003297128,
+      "learning_rate": 5.276276619409215e-08,
+      "loss": 0.1589,
+      "step": 19041
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1969975578810148,
+      "learning_rate": 5.259393854471384e-08,
+      "loss": 0.1777,
+      "step": 19042
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0041848312681114,
+      "learning_rate": 5.242538071928649e-08,
+      "loss": 0.174,
+      "step": 19043
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2724602233865854,
+      "learning_rate": 5.225709272238311e-08,
+      "loss": 0.1433,
+      "step": 19044
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0626391350708608,
+      "learning_rate": 5.208907455856782e-08,
+      "loss": 0.1853,
+      "step": 19045
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9280447894665182,
+      "learning_rate": 5.19213262323981e-08,
+      "loss": 0.1554,
+      "step": 19046
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4995114449645333,
+      "learning_rate": 5.1753847748424735e-08,
+      "loss": 0.1609,
+      "step": 19047
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8714223222188999,
+      "learning_rate": 5.158663911118966e-08,
+      "loss": 0.1459,
+      "step": 19048
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1422759045113517,
+      "learning_rate": 5.141970032522925e-08,
+      "loss": 0.1769,
+      "step": 19049
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9078013270129156,
+      "learning_rate": 5.1253031395070985e-08,
+      "loss": 0.1561,
+      "step": 19050
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4725725563008483,
+      "learning_rate": 5.1086632325236804e-08,
+      "loss": 0.177,
+      "step": 19051
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0447898313350958,
+      "learning_rate": 5.0920503120239775e-08,
+      "loss": 0.1582,
+      "step": 19052
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.403622306022244,
+      "learning_rate": 5.075464378458517e-08,
+      "loss": 0.1592,
+      "step": 19053
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2749760646824433,
+      "learning_rate": 5.0589054322772725e-08,
+      "loss": 0.1795,
+      "step": 19054
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.151845211047926,
+      "learning_rate": 5.04237347392944e-08,
+      "loss": 0.1592,
+      "step": 19055
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0014197338720585,
+      "learning_rate": 5.0258685038634406e-08,
+      "loss": 0.151,
+      "step": 19056
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.381306982875491,
+      "learning_rate": 5.009390522526914e-08,
+      "loss": 0.1544,
+      "step": 19057
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6958918491548332,
+      "learning_rate": 4.992939530366947e-08,
+      "loss": 0.1547,
+      "step": 19058
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4727261828084512,
+      "learning_rate": 4.9765155278296284e-08,
+      "loss": 0.1544,
+      "step": 19059
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.328137183771419,
+      "learning_rate": 4.96011851536049e-08,
+      "loss": 0.169,
+      "step": 19060
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.162867331490707,
+      "learning_rate": 4.9437484934043987e-08,
+      "loss": 0.1507,
+      "step": 19061
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9319785655569531,
+      "learning_rate": 4.927405462405332e-08,
+      "loss": 0.1596,
+      "step": 19062
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1297107255840486,
+      "learning_rate": 4.911089422806603e-08,
+      "loss": 0.1448,
+      "step": 19063
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0959641369353819,
+      "learning_rate": 4.8948003750507455e-08,
+      "loss": 0.1556,
+      "step": 19064
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4195341111796325,
+      "learning_rate": 4.878538319579629e-08,
+      "loss": 0.1801,
+      "step": 19065
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3786838067360543,
+      "learning_rate": 4.862303256834344e-08,
+      "loss": 0.173,
+      "step": 19066
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.551126828679602,
+      "learning_rate": 4.846095187255318e-08,
+      "loss": 0.1725,
+      "step": 19067
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0936716110028684,
+      "learning_rate": 4.8299141112821966e-08,
+      "loss": 0.1677,
+      "step": 19068
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1227017299371511,
+      "learning_rate": 4.8137600293538536e-08,
+      "loss": 0.1773,
+      "step": 19069
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.021515630146002,
+      "learning_rate": 4.7976329419084924e-08,
+      "loss": 0.1722,
+      "step": 19070
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1159643409886524,
+      "learning_rate": 4.781532849383541e-08,
+      "loss": 0.1559,
+      "step": 19071
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.508985334043882,
+      "learning_rate": 4.765459752215651e-08,
+      "loss": 0.1624,
+      "step": 19072
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4314591906020395,
+      "learning_rate": 4.749413650840917e-08,
+      "loss": 0.1639,
+      "step": 19073
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1081591206295462,
+      "learning_rate": 4.7333945456945474e-08,
+      "loss": 0.1629,
+      "step": 19074
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9484977524062326,
+      "learning_rate": 4.717402437211083e-08,
+      "loss": 0.1546,
+      "step": 19075
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.032321978750882,
+      "learning_rate": 4.701437325824287e-08,
+      "loss": 0.1548,
+      "step": 19076
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0338695817692656,
+      "learning_rate": 4.6854992119671484e-08,
+      "loss": 0.1546,
+      "step": 19077
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8415263499415587,
+      "learning_rate": 4.6695880960722085e-08,
+      "loss": 0.1634,
+      "step": 19078
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1299997587136064,
+      "learning_rate": 4.653703978570789e-08,
+      "loss": 0.1507,
+      "step": 19079
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.050521335679299,
+      "learning_rate": 4.6378468598938794e-08,
+      "loss": 0.1812,
+      "step": 19080
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.361031673379913,
+      "learning_rate": 4.62201674047158e-08,
+      "loss": 0.1644,
+      "step": 19081
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9886945750398041,
+      "learning_rate": 4.606213620733324e-08,
+      "loss": 0.1619,
+      "step": 19082
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1377023965397417,
+      "learning_rate": 4.59043750110777e-08,
+      "loss": 0.1677,
+      "step": 19083
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3184405498120437,
+      "learning_rate": 4.574688382022796e-08,
+      "loss": 0.1714,
+      "step": 19084
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8280553700474405,
+      "learning_rate": 4.5589662639056175e-08,
+      "loss": 0.1735,
+      "step": 19085
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.005031500161824,
+      "learning_rate": 4.5432711471826704e-08,
+      "loss": 0.159,
+      "step": 19086
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1721295729425816,
+      "learning_rate": 4.527603032279726e-08,
+      "loss": 0.1513,
+      "step": 19087
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7713609394550736,
+      "learning_rate": 4.511961919621888e-08,
+      "loss": 0.1603,
+      "step": 19088
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.941106674306136,
+      "learning_rate": 4.496347809633261e-08,
+      "loss": 0.1454,
+      "step": 19089
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.824939947863532,
+      "learning_rate": 4.480760702737286e-08,
+      "loss": 0.137,
+      "step": 19090
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.275105006746803,
+      "learning_rate": 4.465200599356956e-08,
+      "loss": 0.1616,
+      "step": 19091
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1215004426931332,
+      "learning_rate": 4.4496674999143786e-08,
+      "loss": 0.1473,
+      "step": 19092
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4440259031152056,
+      "learning_rate": 4.434161404830772e-08,
+      "loss": 0.1563,
+      "step": 19093
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7954277937536793,
+      "learning_rate": 4.41868231452669e-08,
+      "loss": 0.1615,
+      "step": 19094
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1079319217667891,
+      "learning_rate": 4.40323022942224e-08,
+      "loss": 0.1961,
+      "step": 19095
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3025911019288117,
+      "learning_rate": 4.387805149936197e-08,
+      "loss": 0.1534,
+      "step": 19096
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9653298618682854,
+      "learning_rate": 4.3724070764873396e-08,
+      "loss": 0.1516,
+      "step": 19097
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1497810602577705,
+      "learning_rate": 4.35703600949311e-08,
+      "loss": 0.158,
+      "step": 19098
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9031827811840533,
+      "learning_rate": 4.341691949370508e-08,
+      "loss": 0.1581,
+      "step": 19099
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0561575396976002,
+      "learning_rate": 4.326374896535757e-08,
+      "loss": 0.1816,
+      "step": 19100
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2191283042704604,
+      "learning_rate": 4.311084851404301e-08,
+      "loss": 0.1577,
+      "step": 19101
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7094001015154057,
+      "learning_rate": 4.2958218143909214e-08,
+      "loss": 0.1614,
+      "step": 19102
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0578646131166962,
+      "learning_rate": 4.280585785909619e-08,
+      "loss": 0.1417,
+      "step": 19103
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9456307144413473,
+      "learning_rate": 4.265376766373619e-08,
+      "loss": 0.1486,
+      "step": 19104
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9368889064225784,
+      "learning_rate": 4.2501947561955914e-08,
+      "loss": 0.1611,
+      "step": 19105
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9330791875403646,
+      "learning_rate": 4.235039755787318e-08,
+      "loss": 0.1539,
+      "step": 19106
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9165604635550397,
+      "learning_rate": 4.2199117655596924e-08,
+      "loss": 0.1688,
+      "step": 19107
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9110528720041452,
+      "learning_rate": 4.204810785923275e-08,
+      "loss": 0.1742,
+      "step": 19108
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9730364455869631,
+      "learning_rate": 4.1897368172875156e-08,
+      "loss": 0.146,
+      "step": 19109
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5814289522865332,
+      "learning_rate": 4.174689860061532e-08,
+      "loss": 0.1815,
+      "step": 19110
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.543582728055052,
+      "learning_rate": 4.159669914653219e-08,
+      "loss": 0.1706,
+      "step": 19111
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1285363465670573,
+      "learning_rate": 4.144676981470142e-08,
+      "loss": 0.1716,
+      "step": 19112
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3445171338848143,
+      "learning_rate": 4.1297110609189726e-08,
+      "loss": 0.1576,
+      "step": 19113
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8753483217197732,
+      "learning_rate": 4.1147721534056106e-08,
+      "loss": 0.1479,
+      "step": 19114
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9138793875980022,
+      "learning_rate": 4.099860259335287e-08,
+      "loss": 0.1736,
+      "step": 19115
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1239955889645314,
+      "learning_rate": 4.084975379112566e-08,
+      "loss": 0.1492,
+      "step": 19116
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.875353207317943,
+      "learning_rate": 4.070117513141014e-08,
+      "loss": 0.1633,
+      "step": 19117
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9480898716991211,
+      "learning_rate": 4.055286661823976e-08,
+      "loss": 0.1524,
+      "step": 19118
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0478210911230053,
+      "learning_rate": 4.040482825563352e-08,
+      "loss": 0.1739,
+      "step": 19119
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9717161847311233,
+      "learning_rate": 4.025706004760932e-08,
+      "loss": 0.1709,
+      "step": 19120
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9507746435779297,
+      "learning_rate": 4.010956199817506e-08,
+      "loss": 0.1628,
+      "step": 19121
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8743938983011655,
+      "learning_rate": 3.996233411133199e-08,
+      "loss": 0.1664,
+      "step": 19122
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0571851837849218,
+      "learning_rate": 3.981537639107247e-08,
+      "loss": 0.166,
+      "step": 19123
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0122948012698287,
+      "learning_rate": 3.966868884138442e-08,
+      "loss": 0.1588,
+      "step": 19124
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8442422327482219,
+      "learning_rate": 3.952227146624465e-08,
+      "loss": 0.161,
+      "step": 19125
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0852162833362355,
+      "learning_rate": 3.937612426962556e-08,
+      "loss": 0.1804,
+      "step": 19126
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9181077560303157,
+      "learning_rate": 3.923024725549285e-08,
+      "loss": 0.164,
+      "step": 19127
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2820486158378837,
+      "learning_rate": 3.908464042780114e-08,
+      "loss": 0.173,
+      "step": 19128
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.136925404121932,
+      "learning_rate": 3.8939303790501706e-08,
+      "loss": 0.1761,
+      "step": 19129
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1490818691885378,
+      "learning_rate": 3.879423734753585e-08,
+      "loss": 0.1519,
+      "step": 19130
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4095179194478367,
+      "learning_rate": 3.864944110284041e-08,
+      "loss": 0.1591,
+      "step": 19131
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0354443872710228,
+      "learning_rate": 3.850491506034004e-08,
+      "loss": 0.1597,
+      "step": 19132
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.927952075094008,
+      "learning_rate": 3.8360659223957155e-08,
+      "loss": 0.1622,
+      "step": 19133
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0288648343438391,
+      "learning_rate": 3.821667359760306e-08,
+      "loss": 0.1674,
+      "step": 19134
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9923904976945497,
+      "learning_rate": 3.8072958185184635e-08,
+      "loss": 0.1688,
+      "step": 19135
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9784718164833021,
+      "learning_rate": 3.7929512990600995e-08,
+      "loss": 0.1649,
+      "step": 19136
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 12.11685222340869,
+      "learning_rate": 3.778633801774123e-08,
+      "loss": 0.1644,
+      "step": 19137
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8243294680976933,
+      "learning_rate": 3.764343327048892e-08,
+      "loss": 0.1577,
+      "step": 19138
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9167141022086956,
+      "learning_rate": 3.750079875272206e-08,
+      "loss": 0.1739,
+      "step": 19139
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.072275320165148,
+      "learning_rate": 3.735843446830867e-08,
+      "loss": 0.1399,
+      "step": 19140
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9768691762964762,
+      "learning_rate": 3.721634042111011e-08,
+      "loss": 0.165,
+      "step": 19141
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.838464973730494,
+      "learning_rate": 3.707451661498107e-08,
+      "loss": 0.1578,
+      "step": 19142
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9608964549999129,
+      "learning_rate": 3.693296305376959e-08,
+      "loss": 0.1492,
+      "step": 19143
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.228769947210927,
+      "learning_rate": 3.679167974131259e-08,
+      "loss": 0.1453,
+      "step": 19144
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1040416367837809,
+      "learning_rate": 3.665066668144479e-08,
+      "loss": 0.1522,
+      "step": 19145
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8744148989593945,
+      "learning_rate": 3.65099238779909e-08,
+      "loss": 0.1632,
+      "step": 19146
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.922122532580692,
+      "learning_rate": 3.636945133476677e-08,
+      "loss": 0.1846,
+      "step": 19147
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0865718018611439,
+      "learning_rate": 3.622924905558489e-08,
+      "loss": 0.1593,
+      "step": 19148
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.973217513769555,
+      "learning_rate": 3.608931704424778e-08,
+      "loss": 0.1602,
+      "step": 19149
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2178422284953592,
+      "learning_rate": 3.5949655304550193e-08,
+      "loss": 0.1646,
+      "step": 19150
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8812935813974213,
+      "learning_rate": 3.581026384028019e-08,
+      "loss": 0.1523,
+      "step": 19151
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.8532607015421814,
+      "learning_rate": 3.567114265522031e-08,
+      "loss": 0.1392,
+      "step": 19152
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.116891255212077,
+      "learning_rate": 3.55322917531431e-08,
+      "loss": 0.161,
+      "step": 19153
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0735371441706794,
+      "learning_rate": 3.539371113781553e-08,
+      "loss": 0.1443,
+      "step": 19154
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.995049097027222,
+      "learning_rate": 3.525540081299683e-08,
+      "loss": 0.1607,
+      "step": 19155
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.372855437481362,
+      "learning_rate": 3.511736078243844e-08,
+      "loss": 0.1941,
+      "step": 19156
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9548197354253422,
+      "learning_rate": 3.497959104988291e-08,
+      "loss": 0.1552,
+      "step": 19157
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.171880449888493,
+      "learning_rate": 3.4842091619070594e-08,
+      "loss": 0.1554,
+      "step": 19158
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0613535166121013,
+      "learning_rate": 3.470486249372851e-08,
+      "loss": 0.1488,
+      "step": 19159
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0495709235018573,
+      "learning_rate": 3.456790367757923e-08,
+      "loss": 0.1404,
+      "step": 19160
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9657074899743266,
+      "learning_rate": 3.4431215174338675e-08,
+      "loss": 0.1589,
+      "step": 19161
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.615432278184323,
+      "learning_rate": 3.429479698771499e-08,
+      "loss": 0.1848,
+      "step": 19162
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8700346511067772,
+      "learning_rate": 3.415864912140743e-08,
+      "loss": 0.143,
+      "step": 19163
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.998348237566735,
+      "learning_rate": 3.402277157910971e-08,
+      "loss": 0.1488,
+      "step": 19164
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1098996119917584,
+      "learning_rate": 3.3887164364506676e-08,
+      "loss": 0.156,
+      "step": 19165
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9244051009838743,
+      "learning_rate": 3.375182748127759e-08,
+      "loss": 0.1975,
+      "step": 19166
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0719850242465663,
+      "learning_rate": 3.361676093309285e-08,
+      "loss": 0.1649,
+      "step": 19167
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4384884218914096,
+      "learning_rate": 3.34819647236162e-08,
+      "loss": 0.1911,
+      "step": 19168
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6804235632026177,
+      "learning_rate": 3.334743885650471e-08,
+      "loss": 0.1742,
+      "step": 19169
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5391305538864972,
+      "learning_rate": 3.321318333540546e-08,
+      "loss": 0.1745,
+      "step": 19170
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.821442405909109,
+      "learning_rate": 3.307919816396332e-08,
+      "loss": 0.1707,
+      "step": 19171
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9376385559888619,
+      "learning_rate": 3.294548334580982e-08,
+      "loss": 0.1571,
+      "step": 19172
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0596456111225416,
+      "learning_rate": 3.2812038884573185e-08,
+      "loss": 0.1722,
+      "step": 19173
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.512084876890349,
+      "learning_rate": 3.267886478387383e-08,
+      "loss": 0.1509,
+      "step": 19174
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9142596774458099,
+      "learning_rate": 3.254596104732222e-08,
+      "loss": 0.1706,
+      "step": 19175
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.227178005224331,
+      "learning_rate": 3.241332767852545e-08,
+      "loss": 0.1699,
+      "step": 19176
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.988269507234306,
+      "learning_rate": 3.228096468107844e-08,
+      "loss": 0.1521,
+      "step": 19177
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0376160910233094,
+      "learning_rate": 3.214887205857387e-08,
+      "loss": 0.1781,
+      "step": 19178
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0444876319228076,
+      "learning_rate": 3.201704981459441e-08,
+      "loss": 0.1569,
+      "step": 19179
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.151779282207466,
+      "learning_rate": 3.188549795271612e-08,
+      "loss": 0.1567,
+      "step": 19180
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1160958146018711,
+      "learning_rate": 3.175421647650612e-08,
+      "loss": 0.1439,
+      "step": 19181
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1713617665246183,
+      "learning_rate": 3.1623205389526015e-08,
+      "loss": 0.1822,
+      "step": 19182
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1129384740967305,
+      "learning_rate": 3.1492464695328517e-08,
+      "loss": 0.1473,
+      "step": 19183
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1710919498361196,
+      "learning_rate": 3.1361994397463015e-08,
+      "loss": 0.1641,
+      "step": 19184
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.47322006099083,
+      "learning_rate": 3.123179449946445e-08,
+      "loss": 0.1597,
+      "step": 19185
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0377026560653144,
+      "learning_rate": 3.110186500486778e-08,
+      "loss": 0.1523,
+      "step": 19186
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8996411914981404,
+      "learning_rate": 3.097220591719574e-08,
+      "loss": 0.1602,
+      "step": 19187
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1501427842312555,
+      "learning_rate": 3.084281723996552e-08,
+      "loss": 0.1571,
+      "step": 19188
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0734279089738592,
+      "learning_rate": 3.071369897668652e-08,
+      "loss": 0.1622,
+      "step": 19189
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9380478409377784,
+      "learning_rate": 3.0584851130861516e-08,
+      "loss": 0.1438,
+      "step": 19190
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1629182446898398,
+      "learning_rate": 3.0456273705986585e-08,
+      "loss": 0.1703,
+      "step": 19191
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1467098491253804,
+      "learning_rate": 3.032796670554783e-08,
+      "loss": 0.166,
+      "step": 19192
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8112423150515589,
+      "learning_rate": 3.01999301330258e-08,
+      "loss": 0.1689,
+      "step": 19193
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9835760332721184,
+      "learning_rate": 3.007216399189328e-08,
+      "loss": 0.1427,
+      "step": 19194
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0386973083640156,
+      "learning_rate": 2.9944668285617486e-08,
+      "loss": 0.16,
+      "step": 19195
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2121770744729188,
+      "learning_rate": 2.981744301765454e-08,
+      "loss": 0.1702,
+      "step": 19196
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1948688625434023,
+      "learning_rate": 2.9690488191457256e-08,
+      "loss": 0.1715,
+      "step": 19197
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0920289287961265,
+      "learning_rate": 2.9563803810468417e-08,
+      "loss": 0.1407,
+      "step": 19198
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6118283072831598,
+      "learning_rate": 2.943738987812528e-08,
+      "loss": 0.1629,
+      "step": 19199
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0650245468422912,
+      "learning_rate": 2.9311246397855097e-08,
+      "loss": 0.1664,
+      "step": 19200
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9317474614007701,
+      "learning_rate": 2.9185373373080694e-08,
+      "loss": 0.1704,
+      "step": 19201
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0744841689521663,
+      "learning_rate": 2.9059770807217114e-08,
+      "loss": 0.1483,
+      "step": 19202
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0447025231148186,
+      "learning_rate": 2.893443870366941e-08,
+      "loss": 0.1695,
+      "step": 19203
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0831017244705572,
+      "learning_rate": 2.8809377065838195e-08,
+      "loss": 0.1731,
+      "step": 19204
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.957580011722551,
+      "learning_rate": 2.8684585897116313e-08,
+      "loss": 0.1442,
+      "step": 19205
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1950503252036402,
+      "learning_rate": 2.856006520088772e-08,
+      "loss": 0.1447,
+      "step": 19206
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0911236067523062,
+      "learning_rate": 2.843581498053083e-08,
+      "loss": 0.154,
+      "step": 19207
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3110033762325044,
+      "learning_rate": 2.8311835239415166e-08,
+      "loss": 0.166,
+      "step": 19208
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8672953809905334,
+      "learning_rate": 2.8188125980904702e-08,
+      "loss": 0.144,
+      "step": 19209
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1167029052994197,
+      "learning_rate": 2.8064687208354534e-08,
+      "loss": 0.1766,
+      "step": 19210
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0134270323241488,
+      "learning_rate": 2.7941518925113098e-08,
+      "loss": 0.1565,
+      "step": 19211
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3869272247853248,
+      "learning_rate": 2.7818621134521052e-08,
+      "loss": 0.1532,
+      "step": 19212
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1589186608695665,
+      "learning_rate": 2.7695993839912394e-08,
+      "loss": 0.1417,
+      "step": 19213
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8894785662670377,
+      "learning_rate": 2.7573637044612245e-08,
+      "loss": 0.1661,
+      "step": 19214
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.039413572033424,
+      "learning_rate": 2.7451550751941282e-08,
+      "loss": 0.1762,
+      "step": 19215
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1886792996172786,
+      "learning_rate": 2.7329734965210185e-08,
+      "loss": 0.1754,
+      "step": 19216
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9696007012494356,
+      "learning_rate": 2.7208189687722986e-08,
+      "loss": 0.1551,
+      "step": 19217
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.037129191027604,
+      "learning_rate": 2.708691492277704e-08,
+      "loss": 0.1538,
+      "step": 19218
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0038345821338919,
+      "learning_rate": 2.6965910673661945e-08,
+      "loss": 0.1361,
+      "step": 19219
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0297748191335272,
+      "learning_rate": 2.684517694365951e-08,
+      "loss": 0.1653,
+      "step": 19220
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9427532899833283,
+      "learning_rate": 2.6724713736044904e-08,
+      "loss": 0.1593,
+      "step": 19221
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2120297659638368,
+      "learning_rate": 2.6604521054085507e-08,
+      "loss": 0.1655,
+      "step": 19222
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0842678478940873,
+      "learning_rate": 2.6484598901042047e-08,
+      "loss": 0.1572,
+      "step": 19223
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0850936239860367,
+      "learning_rate": 2.6364947280167474e-08,
+      "loss": 0.1567,
+      "step": 19224
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1589672563300961,
+      "learning_rate": 2.6245566194706973e-08,
+      "loss": 0.1565,
+      "step": 19225
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9041978598526125,
+      "learning_rate": 2.6126455647899063e-08,
+      "loss": 0.1401,
+      "step": 19226
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1471922837033912,
+      "learning_rate": 2.6007615642973383e-08,
+      "loss": 0.18,
+      "step": 19227
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9977763246122786,
+      "learning_rate": 2.588904618315513e-08,
+      "loss": 0.1533,
+      "step": 19228
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9580627412826506,
+      "learning_rate": 2.5770747271659512e-08,
+      "loss": 0.1619,
+      "step": 19229
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.992200161492214,
+      "learning_rate": 2.5652718911696185e-08,
+      "loss": 0.1521,
+      "step": 19230
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0007268238753533,
+      "learning_rate": 2.5534961106465918e-08,
+      "loss": 0.1644,
+      "step": 19231
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8550931452612291,
+      "learning_rate": 2.5417473859162823e-08,
+      "loss": 0.147,
+      "step": 19232
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8227788485716492,
+      "learning_rate": 2.530025717297546e-08,
+      "loss": 0.1548,
+      "step": 19233
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1593496255775817,
+      "learning_rate": 2.5183311051080184e-08,
+      "loss": 0.1691,
+      "step": 19234
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1956712286111624,
+      "learning_rate": 2.5066635496652225e-08,
+      "loss": 0.1716,
+      "step": 19235
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9729699845844008,
+      "learning_rate": 2.4950230512854612e-08,
+      "loss": 0.1862,
+      "step": 19236
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0720151399962266,
+      "learning_rate": 2.4834096102845927e-08,
+      "loss": 0.167,
+      "step": 19237
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9591798117726331,
+      "learning_rate": 2.4718232269774767e-08,
+      "loss": 0.1603,
+      "step": 19238
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.343576188483149,
+      "learning_rate": 2.460263901678639e-08,
+      "loss": 0.1562,
+      "step": 19239
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.252390946762163,
+      "learning_rate": 2.4487316347013845e-08,
+      "loss": 0.1719,
+      "step": 19240
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.156646101992709,
+      "learning_rate": 2.4372264263586852e-08,
+      "loss": 0.1698,
+      "step": 19241
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.081421366726149,
+      "learning_rate": 2.425748276962514e-08,
+      "loss": 0.1574,
+      "step": 19242
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9253255105133054,
+      "learning_rate": 2.414297186824288e-08,
+      "loss": 0.1552,
+      "step": 19243
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8335899901045356,
+      "learning_rate": 2.402873156254537e-08,
+      "loss": 0.1423,
+      "step": 19244
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9997764295713396,
+      "learning_rate": 2.3914761855632353e-08,
+      "loss": 0.1778,
+      "step": 19245
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0051440205824753,
+      "learning_rate": 2.3801062750595794e-08,
+      "loss": 0.1605,
+      "step": 19246
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.016491155010764,
+      "learning_rate": 2.3687634250517676e-08,
+      "loss": 0.1457,
+      "step": 19247
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9058016869766914,
+      "learning_rate": 2.357447635847554e-08,
+      "loss": 0.1631,
+      "step": 19248
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9075130738463206,
+      "learning_rate": 2.3461589077540258e-08,
+      "loss": 0.1589,
+      "step": 19249
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9980310571120593,
+      "learning_rate": 2.3348972410772718e-08,
+      "loss": 0.1454,
+      "step": 19250
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0595001302371383,
+      "learning_rate": 2.3236626361227145e-08,
+      "loss": 0.1514,
+      "step": 19251
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5416509882505045,
+      "learning_rate": 2.3124550931952205e-08,
+      "loss": 0.1594,
+      "step": 19252
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8453900521274997,
+      "learning_rate": 2.3012746125987695e-08,
+      "loss": 0.1525,
+      "step": 19253
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7985186282417956,
+      "learning_rate": 2.290121194636452e-08,
+      "loss": 0.1423,
+      "step": 19254
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8561020250996914,
+      "learning_rate": 2.278994839611026e-08,
+      "loss": 0.151,
+      "step": 19255
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.27649659032301,
+      "learning_rate": 2.2678955478242504e-08,
+      "loss": 0.1467,
+      "step": 19256
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8932878466540868,
+      "learning_rate": 2.256823319577217e-08,
+      "loss": 0.1622,
+      "step": 19257
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9914841386610725,
+      "learning_rate": 2.2457781551700198e-08,
+      "loss": 0.1424,
+      "step": 19258
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0687596475817032,
+      "learning_rate": 2.2347600549025294e-08,
+      "loss": 0.1477,
+      "step": 19259
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8674035862177825,
+      "learning_rate": 2.2237690190736183e-08,
+      "loss": 0.1522,
+      "step": 19260
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1072519797129452,
+      "learning_rate": 2.2128050479812703e-08,
+      "loss": 0.1686,
+      "step": 19261
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8107126930749041,
+      "learning_rate": 2.2018681419229138e-08,
+      "loss": 0.1415,
+      "step": 19262
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0972507429958072,
+      "learning_rate": 2.1909583011952007e-08,
+      "loss": 0.1542,
+      "step": 19263
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9810825168290226,
+      "learning_rate": 2.1800755260942276e-08,
+      "loss": 0.1549,
+      "step": 19264
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7842847357181404,
+      "learning_rate": 2.169219816914869e-08,
+      "loss": 0.1415,
+      "step": 19265
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8654898219260009,
+      "learning_rate": 2.1583911739518904e-08,
+      "loss": 0.1685,
+      "step": 19266
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3440330628325314,
+      "learning_rate": 2.1475895974989446e-08,
+      "loss": 0.1689,
+      "step": 19267
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3828300319608515,
+      "learning_rate": 2.13681508784902e-08,
+      "loss": 0.1501,
+      "step": 19268
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.462503265865077,
+      "learning_rate": 2.1260676452942164e-08,
+      "loss": 0.1763,
+      "step": 19269
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9461929891605176,
+      "learning_rate": 2.1153472701263e-08,
+      "loss": 0.1723,
+      "step": 19270
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2117507915141943,
+      "learning_rate": 2.1046539626359274e-08,
+      "loss": 0.1671,
+      "step": 19271
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0426408318344735,
+      "learning_rate": 2.093987723113089e-08,
+      "loss": 0.1633,
+      "step": 19272
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0395904044309374,
+      "learning_rate": 2.0833485518473305e-08,
+      "loss": 0.1538,
+      "step": 19273
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0043395663080972,
+      "learning_rate": 2.0727364491269773e-08,
+      "loss": 0.1555,
+      "step": 19274
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3419287721746127,
+      "learning_rate": 2.0621514152401323e-08,
+      "loss": 0.1582,
+      "step": 19275
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1290578600458498,
+      "learning_rate": 2.0515934504736768e-08,
+      "loss": 0.1365,
+      "step": 19276
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6278224170282116,
+      "learning_rate": 2.0410625551141594e-08,
+      "loss": 0.1416,
+      "step": 19277
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5792626413568063,
+      "learning_rate": 2.0305587294472408e-08,
+      "loss": 0.1629,
+      "step": 19278
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1919448565898834,
+      "learning_rate": 2.0200819737576926e-08,
+      "loss": 0.1526,
+      "step": 19279
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.179775286640012,
+      "learning_rate": 2.0096322883298435e-08,
+      "loss": 0.1533,
+      "step": 19280
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0520319965190825,
+      "learning_rate": 1.9992096734471333e-08,
+      "loss": 0.1556,
+      "step": 19281
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2946183283037571,
+      "learning_rate": 1.988814129392114e-08,
+      "loss": 0.1473,
+      "step": 19282
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0521064047313005,
+      "learning_rate": 1.9784456564470035e-08,
+      "loss": 0.1545,
+      "step": 19283
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2723479920478658,
+      "learning_rate": 1.9681042548928e-08,
+      "loss": 0.1575,
+      "step": 19284
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9780784388943833,
+      "learning_rate": 1.9577899250101674e-08,
+      "loss": 0.1556,
+      "step": 19285
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9001383282508998,
+      "learning_rate": 1.9475026670788822e-08,
+      "loss": 0.1594,
+      "step": 19286
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.193731060326972,
+      "learning_rate": 1.9372424813779432e-08,
+      "loss": 0.1526,
+      "step": 19287
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1469736780532818,
+      "learning_rate": 1.9270093681856837e-08,
+      "loss": 0.1938,
+      "step": 19288
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.133842881745061,
+      "learning_rate": 1.9168033277796595e-08,
+      "loss": 0.1514,
+      "step": 19289
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4315479417567807,
+      "learning_rate": 1.9066243604367595e-08,
+      "loss": 0.1574,
+      "step": 19290
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1500103445417296,
+      "learning_rate": 1.896472466432986e-08,
+      "loss": 0.1771,
+      "step": 19291
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1258307209006846,
+      "learning_rate": 1.8863476460437847e-08,
+      "loss": 0.1519,
+      "step": 19292
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2529336735398418,
+      "learning_rate": 1.876249899543825e-08,
+      "loss": 0.1799,
+      "step": 19293
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8555299298372947,
+      "learning_rate": 1.866179227206888e-08,
+      "loss": 0.1459,
+      "step": 19294
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0055165033399343,
+      "learning_rate": 1.8561356293061995e-08,
+      "loss": 0.1629,
+      "step": 19295
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8929016490331868,
+      "learning_rate": 1.846119106114319e-08,
+      "loss": 0.1629,
+      "step": 19296
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4282189392830789,
+      "learning_rate": 1.8361296579026965e-08,
+      "loss": 0.173,
+      "step": 19297
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3290545789755563,
+      "learning_rate": 1.8261672849425593e-08,
+      "loss": 0.1651,
+      "step": 19298
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.506479397433706,
+      "learning_rate": 1.8162319875040247e-08,
+      "loss": 0.1668,
+      "step": 19299
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2558990219488457,
+      "learning_rate": 1.8063237658564325e-08,
+      "loss": 0.1554,
+      "step": 19300
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2926613820482529,
+      "learning_rate": 1.7964426202687902e-08,
+      "loss": 0.1858,
+      "step": 19301
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9604834521488833,
+      "learning_rate": 1.7865885510089943e-08,
+      "loss": 0.1639,
+      "step": 19302
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8900020514217135,
+      "learning_rate": 1.7767615583443865e-08,
+      "loss": 0.1464,
+      "step": 19303
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1007333087388065,
+      "learning_rate": 1.7669616425414203e-08,
+      "loss": 0.1428,
+      "step": 19304
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.255905003951268,
+      "learning_rate": 1.7571888038661056e-08,
+      "loss": 0.1575,
+      "step": 19305
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2103777182423623,
+      "learning_rate": 1.747443042583341e-08,
+      "loss": 0.1729,
+      "step": 19306
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8329630456146007,
+      "learning_rate": 1.737724358957582e-08,
+      "loss": 0.1635,
+      "step": 19307
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1297998214124445,
+      "learning_rate": 1.7280327532525066e-08,
+      "loss": 0.1525,
+      "step": 19308
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6314144516197124,
+      "learning_rate": 1.7183682257309043e-08,
+      "loss": 0.1684,
+      "step": 19309
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3026685814979018,
+      "learning_rate": 1.708730776654899e-08,
+      "loss": 0.1768,
+      "step": 19310
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8763899372471403,
+      "learning_rate": 1.6991204062859478e-08,
+      "loss": 0.145,
+      "step": 19311
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0921780067084008,
+      "learning_rate": 1.6895371148847316e-08,
+      "loss": 0.1505,
+      "step": 19312
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0594109223251242,
+      "learning_rate": 1.6799809027112645e-08,
+      "loss": 0.1549,
+      "step": 19313
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9525505933657265,
+      "learning_rate": 1.6704517700246725e-08,
+      "loss": 0.1458,
+      "step": 19314
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0551914198110557,
+      "learning_rate": 1.6609497170834154e-08,
+      "loss": 0.1677,
+      "step": 19315
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7643407849401922,
+      "learning_rate": 1.6514747441453983e-08,
+      "loss": 0.1662,
+      "step": 19316
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6991980825364204,
+      "learning_rate": 1.6420268514674153e-08,
+      "loss": 0.1714,
+      "step": 19317
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8688835282260733,
+      "learning_rate": 1.6326060393058174e-08,
+      "loss": 0.1686,
+      "step": 19318
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7232939795009312,
+      "learning_rate": 1.6232123079162887e-08,
+      "loss": 0.1594,
+      "step": 19319
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1199605143149651,
+      "learning_rate": 1.6138456575534034e-08,
+      "loss": 0.1828,
+      "step": 19320
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9937593945033635,
+      "learning_rate": 1.6045060884714027e-08,
+      "loss": 0.1543,
+      "step": 19321
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0113342428247933,
+      "learning_rate": 1.5951936009235282e-08,
+      "loss": 0.1813,
+      "step": 19322
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9151153716972111,
+      "learning_rate": 1.5859081951624668e-08,
+      "loss": 0.1466,
+      "step": 19323
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.10118412807842,
+      "learning_rate": 1.5766498714400168e-08,
+      "loss": 0.1761,
+      "step": 19324
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1592831567954685,
+      "learning_rate": 1.5674186300073113e-08,
+      "loss": 0.1723,
+      "step": 19325
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0636078897730215,
+      "learning_rate": 1.5582144711148163e-08,
+      "loss": 0.1616,
+      "step": 19326
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.157339395167924,
+      "learning_rate": 1.5490373950121097e-08,
+      "loss": 0.1554,
+      "step": 19327
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8743647019705796,
+      "learning_rate": 1.5398874019481037e-08,
+      "loss": 0.1766,
+      "step": 19328
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8649892968907565,
+      "learning_rate": 1.5307644921710442e-08,
+      "loss": 0.1613,
+      "step": 19329
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0898910343208772,
+      "learning_rate": 1.5216686659285106e-08,
+      "loss": 0.1548,
+      "step": 19330
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1210862777680415,
+      "learning_rate": 1.512599923467084e-08,
+      "loss": 0.1539,
+      "step": 19331
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8436561419359566,
+      "learning_rate": 1.5035582650326786e-08,
+      "loss": 0.1484,
+      "step": 19332
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0534986774048938,
+      "learning_rate": 1.4945436908707645e-08,
+      "loss": 0.1465,
+      "step": 19333
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1097802246883643,
+      "learning_rate": 1.4855562012257019e-08,
+      "loss": 0.1513,
+      "step": 19334
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9406791484763374,
+      "learning_rate": 1.4765957963412957e-08,
+      "loss": 0.1457,
+      "step": 19335
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9979331385771827,
+      "learning_rate": 1.467662476460574e-08,
+      "loss": 0.1778,
+      "step": 19336
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9978593170264436,
+      "learning_rate": 1.4587562418260092e-08,
+      "loss": 0.1857,
+      "step": 19337
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0801844337278366,
+      "learning_rate": 1.449877092679075e-08,
+      "loss": 0.1514,
+      "step": 19338
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9235210259060828,
+      "learning_rate": 1.4410250292605788e-08,
+      "loss": 0.1797,
+      "step": 19339
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5939303682114307,
+      "learning_rate": 1.4322000518106616e-08,
+      "loss": 0.168,
+      "step": 19340
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0240846558467789,
+      "learning_rate": 1.4234021605687987e-08,
+      "loss": 0.1645,
+      "step": 19341
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1303492834383382,
+      "learning_rate": 1.414631355773466e-08,
+      "loss": 0.1592,
+      "step": 19342
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0958579086559461,
+      "learning_rate": 1.405887637662695e-08,
+      "loss": 0.1577,
+      "step": 19343
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3423002977177827,
+      "learning_rate": 1.3971710064736299e-08,
+      "loss": 0.1589,
+      "step": 19344
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9730615483497294,
+      "learning_rate": 1.3884814624427478e-08,
+      "loss": 0.1797,
+      "step": 19345
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8705667810277697,
+      "learning_rate": 1.3798190058056383e-08,
+      "loss": 0.143,
+      "step": 19346
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0582401666488022,
+      "learning_rate": 1.3711836367973353e-08,
+      "loss": 0.1499,
+      "step": 19347
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2066413755415508,
+      "learning_rate": 1.362575355652096e-08,
+      "loss": 0.155,
+      "step": 19348
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6819423962631999,
+      "learning_rate": 1.3539941626034003e-08,
+      "loss": 0.1695,
+      "step": 19349
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8172723617428077,
+      "learning_rate": 1.345440057884062e-08,
+      "loss": 0.1632,
+      "step": 19350
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8142788410916596,
+      "learning_rate": 1.3369130417260067e-08,
+      "loss": 0.1624,
+      "step": 19351
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.973732870831668,
+      "learning_rate": 1.3284131143606049e-08,
+      "loss": 0.1711,
+      "step": 19352
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9400105566861664,
+      "learning_rate": 1.3199402760184499e-08,
+      "loss": 0.1644,
+      "step": 19353
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0797258749734406,
+      "learning_rate": 1.3114945269292468e-08,
+      "loss": 0.1795,
+      "step": 19354
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1797224027283473,
+      "learning_rate": 1.3030758673221456e-08,
+      "loss": 0.1561,
+      "step": 19355
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1102196689897572,
+      "learning_rate": 1.2946842974256301e-08,
+      "loss": 0.1649,
+      "step": 19356
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6320482620567274,
+      "learning_rate": 1.2863198174671853e-08,
+      "loss": 0.1408,
+      "step": 19357
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3116515018242307,
+      "learning_rate": 1.2779824276736298e-08,
+      "loss": 0.159,
+      "step": 19358
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.76738250679809,
+      "learning_rate": 1.2696721282712266e-08,
+      "loss": 0.1355,
+      "step": 19359
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1489894563325833,
+      "learning_rate": 1.2613889194854623e-08,
+      "loss": 0.1742,
+      "step": 19360
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.876093932556513,
+      "learning_rate": 1.253132801540935e-08,
+      "loss": 0.1541,
+      "step": 19361
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1655631323365128,
+      "learning_rate": 1.2449037746614657e-08,
+      "loss": 0.1569,
+      "step": 19362
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3719793760387626,
+      "learning_rate": 1.2367018390704311e-08,
+      "loss": 0.1394,
+      "step": 19363
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1327152068419841,
+      "learning_rate": 1.22852699499032e-08,
+      "loss": 0.1777,
+      "step": 19364
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3962861506837203,
+      "learning_rate": 1.220379242642844e-08,
+      "loss": 0.1614,
+      "step": 19365
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9929737564903329,
+      "learning_rate": 1.2122585822489374e-08,
+      "loss": 0.1599,
+      "step": 19366
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9344319731990219,
+      "learning_rate": 1.2041650140289796e-08,
+      "loss": 0.1611,
+      "step": 19367
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9210789177024352,
+      "learning_rate": 1.1960985382024615e-08,
+      "loss": 0.1423,
+      "step": 19368
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2308739590673927,
+      "learning_rate": 1.188059154988097e-08,
+      "loss": 0.1977,
+      "step": 19369
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9493833067382035,
+      "learning_rate": 1.1800468646041563e-08,
+      "loss": 0.1815,
+      "step": 19370
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8787135242048683,
+      "learning_rate": 1.1720616672676876e-08,
+      "loss": 0.1478,
+      "step": 19371
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0307096213171172,
+      "learning_rate": 1.1641035631956288e-08,
+      "loss": 0.1711,
+      "step": 19372
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2366301080695932,
+      "learning_rate": 1.156172552603585e-08,
+      "loss": 0.1562,
+      "step": 19373
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9931567520163074,
+      "learning_rate": 1.1482686357068284e-08,
+      "loss": 0.1371,
+      "step": 19374
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9024408195387488,
+      "learning_rate": 1.1403918127196323e-08,
+      "loss": 0.173,
+      "step": 19375
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0171629994494382,
+      "learning_rate": 1.1325420838558254e-08,
+      "loss": 0.1705,
+      "step": 19376
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7020732645492647,
+      "learning_rate": 1.1247194493281266e-08,
+      "loss": 0.1393,
+      "step": 19377
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9275481098725942,
+      "learning_rate": 1.1169239093489214e-08,
+      "loss": 0.1614,
+      "step": 19378
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0968512398274697,
+      "learning_rate": 1.1091554641294854e-08,
+      "loss": 0.146,
+      "step": 19379
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1071662113271206,
+      "learning_rate": 1.10141411388065e-08,
+      "loss": 0.1928,
+      "step": 19380
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0772512461637493,
+      "learning_rate": 1.0936998588124693e-08,
+      "loss": 0.145,
+      "step": 19381
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1481733489822117,
+      "learning_rate": 1.0860126991339982e-08,
+      "loss": 0.1415,
+      "step": 19382
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8551816592286479,
+      "learning_rate": 1.0783526350538476e-08,
+      "loss": 0.1608,
+      "step": 19383
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9707542368946773,
+      "learning_rate": 1.0707196667798513e-08,
+      "loss": 0.1539,
+      "step": 19384
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5645121903262291,
+      "learning_rate": 1.063113794518955e-08,
+      "loss": 0.1612,
+      "step": 19385
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3182108529350907,
+      "learning_rate": 1.0555350184775493e-08,
+      "loss": 0.1936,
+      "step": 19386
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9540155428082053,
+      "learning_rate": 1.047983338861136e-08,
+      "loss": 0.1589,
+      "step": 19387
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9832984324694535,
+      "learning_rate": 1.0404587558746626e-08,
+      "loss": 0.1438,
+      "step": 19388
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9006067158466624,
+      "learning_rate": 1.032961269722077e-08,
+      "loss": 0.1776,
+      "step": 19389
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.244743279426564,
+      "learning_rate": 1.0254908806068831e-08,
+      "loss": 0.1625,
+      "step": 19390
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9579273824568465,
+      "learning_rate": 1.0180475887316966e-08,
+      "loss": 0.1696,
+      "step": 19391
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0104067076033303,
+      "learning_rate": 1.010631394298467e-08,
+      "loss": 0.1642,
+      "step": 19392
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9938051532179932,
+      "learning_rate": 1.0032422975081446e-08,
+      "loss": 0.1532,
+      "step": 19393
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8113243547723034,
+      "learning_rate": 9.95880298561347e-09,
+      "loss": 0.1686,
+      "step": 19394
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6819415896919474,
+      "learning_rate": 9.88545397657692e-09,
+      "loss": 0.1578,
+      "step": 19395
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6053397774824008,
+      "learning_rate": 9.812375949962426e-09,
+      "loss": 0.1852,
+      "step": 19396
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.088283665461383,
+      "learning_rate": 9.739568907750629e-09,
+      "loss": 0.136,
+      "step": 19397
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9290486862927769,
+      "learning_rate": 9.667032851917723e-09,
+      "loss": 0.167,
+      "step": 19398
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1691011391086281,
+      "learning_rate": 9.594767784431025e-09,
+      "loss": 0.139,
+      "step": 19399
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8467822619073552,
+      "learning_rate": 9.522773707250077e-09,
+      "loss": 0.1709,
+      "step": 19400
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.471340651344358,
+      "learning_rate": 9.451050622328873e-09,
+      "loss": 0.16,
+      "step": 19401
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.273843813651816,
+      "learning_rate": 9.379598531611412e-09,
+      "loss": 0.157,
+      "step": 19402
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5288014134271248,
+      "learning_rate": 9.308417437037254e-09,
+      "loss": 0.1755,
+      "step": 19403
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9807843288626784,
+      "learning_rate": 9.237507340535968e-09,
+      "loss": 0.1631,
+      "step": 19404
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9258704245017536,
+      "learning_rate": 9.166868244031568e-09,
+      "loss": 0.1685,
+      "step": 19405
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1879016453660538,
+      "learning_rate": 9.096500149440302e-09,
+      "loss": 0.1534,
+      "step": 19406
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.866655679800848,
+      "learning_rate": 9.02640305867064e-09,
+      "loss": 0.1266,
+      "step": 19407
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0303505197830716,
+      "learning_rate": 8.956576973624398e-09,
+      "loss": 0.1772,
+      "step": 19408
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0919726382863375,
+      "learning_rate": 8.887021896195614e-09,
+      "loss": 0.1592,
+      "step": 19409
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.986364397224031,
+      "learning_rate": 8.817737828269446e-09,
+      "loss": 0.1547,
+      "step": 19410
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5733663737118595,
+      "learning_rate": 8.748724771727724e-09,
+      "loss": 0.1783,
+      "step": 19411
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1926579268620583,
+      "learning_rate": 8.679982728440061e-09,
+      "loss": 0.1575,
+      "step": 19412
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0728604790120746,
+      "learning_rate": 8.611511700272746e-09,
+      "loss": 0.1656,
+      "step": 19413
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9749045658625051,
+      "learning_rate": 8.543311689083177e-09,
+      "loss": 0.142,
+      "step": 19414
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9908315305488623,
+      "learning_rate": 8.47538269671988e-09,
+      "loss": 0.1469,
+      "step": 19415
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0193752846011748,
+      "learning_rate": 8.407724725025823e-09,
+      "loss": 0.1627,
+      "step": 19416
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.770904345379466,
+      "learning_rate": 8.340337775837316e-09,
+      "loss": 0.1468,
+      "step": 19417
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9309726993901151,
+      "learning_rate": 8.273221850980673e-09,
+      "loss": 0.1453,
+      "step": 19418
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1105475116378467,
+      "learning_rate": 8.206376952277772e-09,
+      "loss": 0.1756,
+      "step": 19419
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4908709404855125,
+      "learning_rate": 8.139803081540499e-09,
+      "loss": 0.1683,
+      "step": 19420
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.939681217803894,
+      "learning_rate": 8.073500240576292e-09,
+      "loss": 0.1752,
+      "step": 19421
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9335727034124158,
+      "learning_rate": 8.007468431182609e-09,
+      "loss": 0.1528,
+      "step": 19422
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8478138172436791,
+      "learning_rate": 7.941707655150233e-09,
+      "loss": 0.1453,
+      "step": 19423
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9989968451138482,
+      "learning_rate": 7.876217914264406e-09,
+      "loss": 0.1616,
+      "step": 19424
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.419711096650346,
+      "learning_rate": 7.810999210299263e-09,
+      "loss": 0.1573,
+      "step": 19425
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8033453385822357,
+      "learning_rate": 7.746051545025613e-09,
+      "loss": 0.1636,
+      "step": 19426
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2856780985716851,
+      "learning_rate": 7.681374920205375e-09,
+      "loss": 0.1671,
+      "step": 19427
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.021890655817455,
+      "learning_rate": 7.616969337591595e-09,
+      "loss": 0.1625,
+      "step": 19428
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.200794353149773,
+      "learning_rate": 7.552834798931763e-09,
+      "loss": 0.162,
+      "step": 19429
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9148040032386322,
+      "learning_rate": 7.488971305965598e-09,
+      "loss": 0.1706,
+      "step": 19430
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9170760257674808,
+      "learning_rate": 7.425378860425048e-09,
+      "loss": 0.1772,
+      "step": 19431
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1806206086463553,
+      "learning_rate": 7.36205746403651e-09,
+      "loss": 0.1531,
+      "step": 19432
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8990746509040047,
+      "learning_rate": 7.299007118516388e-09,
+      "loss": 0.1441,
+      "step": 19433
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1953447975827665,
+      "learning_rate": 7.236227825574427e-09,
+      "loss": 0.1672,
+      "step": 19434
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1967523379896308,
+      "learning_rate": 7.173719586914818e-09,
+      "loss": 0.1612,
+      "step": 19435
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.091873566076658,
+      "learning_rate": 7.111482404231762e-09,
+      "loss": 0.1483,
+      "step": 19436
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3397730582631586,
+      "learning_rate": 7.049516279215018e-09,
+      "loss": 0.1588,
+      "step": 19437
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2216177742464538,
+      "learning_rate": 6.987821213544355e-09,
+      "loss": 0.1521,
+      "step": 19438
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2590376447335887,
+      "learning_rate": 6.926397208892877e-09,
+      "loss": 0.1614,
+      "step": 19439
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5650243589807098,
+      "learning_rate": 6.8652442669281394e-09,
+      "loss": 0.1776,
+      "step": 19440
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9726278730987528,
+      "learning_rate": 6.804362389306596e-09,
+      "loss": 0.185,
+      "step": 19441
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9403291135586567,
+      "learning_rate": 6.743751577682478e-09,
+      "loss": 0.1619,
+      "step": 19442
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9548342447709992,
+      "learning_rate": 6.683411833697806e-09,
+      "loss": 0.1671,
+      "step": 19443
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2582047847405757,
+      "learning_rate": 6.623343158990159e-09,
+      "loss": 0.1707,
+      "step": 19444
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8723806454457383,
+      "learning_rate": 6.563545555189343e-09,
+      "loss": 0.1594,
+      "step": 19445
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4617744785679774,
+      "learning_rate": 6.504019023916286e-09,
+      "loss": 0.1528,
+      "step": 19446
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.967536921726817,
+      "learning_rate": 6.444763566786361e-09,
+      "loss": 0.1491,
+      "step": 19447
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3426213804774425,
+      "learning_rate": 6.385779185407171e-09,
+      "loss": 0.1664,
+      "step": 19448
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0871880246818988,
+      "learning_rate": 6.327065881377437e-09,
+      "loss": 0.1806,
+      "step": 19449
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8191155304043163,
+      "learning_rate": 6.2686236562903294e-09,
+      "loss": 0.1644,
+      "step": 19450
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8209864529546275,
+      "learning_rate": 6.210452511731246e-09,
+      "loss": 0.1499,
+      "step": 19451
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9778755032833552,
+      "learning_rate": 6.152552449278925e-09,
+      "loss": 0.1752,
+      "step": 19452
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9975335549067313,
+      "learning_rate": 6.094923470502112e-09,
+      "loss": 0.185,
+      "step": 19453
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8685575465714487,
+      "learning_rate": 6.03756557696511e-09,
+      "loss": 0.1568,
+      "step": 19454
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9752549270513626,
+      "learning_rate": 5.980478770224452e-09,
+      "loss": 0.1645,
+      "step": 19455
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.002505500250835,
+      "learning_rate": 5.923663051826678e-09,
+      "loss": 0.154,
+      "step": 19456
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6090929823221471,
+      "learning_rate": 5.867118423314999e-09,
+      "loss": 0.1711,
+      "step": 19457
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1652972258786725,
+      "learning_rate": 5.810844886221523e-09,
+      "loss": 0.1514,
+      "step": 19458
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9219302327292703,
+      "learning_rate": 5.754842442073916e-09,
+      "loss": 0.1825,
+      "step": 19459
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.075261151804333,
+      "learning_rate": 5.699111092389853e-09,
+      "loss": 0.146,
+      "step": 19460
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8768162885658606,
+      "learning_rate": 5.643650838682568e-09,
+      "loss": 0.1537,
+      "step": 19461
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0508940154630682,
+      "learning_rate": 5.588461682455304e-09,
+      "loss": 0.1637,
+      "step": 19462
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.346320532701835,
+      "learning_rate": 5.53354362520575e-09,
+      "loss": 0.1735,
+      "step": 19463
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0066814892312403,
+      "learning_rate": 5.478896668423828e-09,
+      "loss": 0.1506,
+      "step": 19464
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0829562803149226,
+      "learning_rate": 5.4245208135905725e-09,
+      "loss": 0.1518,
+      "step": 19465
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2651296228774314,
+      "learning_rate": 5.370416062181472e-09,
+      "loss": 0.1628,
+      "step": 19466
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9646946338958992,
+      "learning_rate": 5.31658241566535e-09,
+      "loss": 0.1564,
+      "step": 19467
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.692309308973504,
+      "learning_rate": 5.2630198754999304e-09,
+      "loss": 0.1626,
+      "step": 19468
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0675900182191573,
+      "learning_rate": 5.209728443140716e-09,
+      "loss": 0.1559,
+      "step": 19469
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0059903839169024,
+      "learning_rate": 5.156708120032106e-09,
+      "loss": 0.1625,
+      "step": 19470
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4617637087129334,
+      "learning_rate": 5.103958907611839e-09,
+      "loss": 0.1519,
+      "step": 19471
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1075830315746373,
+      "learning_rate": 5.051480807312103e-09,
+      "loss": 0.1766,
+      "step": 19472
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1481469399042683,
+      "learning_rate": 4.999273820553985e-09,
+      "loss": 0.1509,
+      "step": 19473
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1386676238961613,
+      "learning_rate": 4.947337948756348e-09,
+      "loss": 0.1763,
+      "step": 19474
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.072465063966951,
+      "learning_rate": 4.895673193325845e-09,
+      "loss": 0.1605,
+      "step": 19475
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8936326244245394,
+      "learning_rate": 4.8442795556657984e-09,
+      "loss": 0.1624,
+      "step": 19476
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9424084453406708,
+      "learning_rate": 4.793157037168428e-09,
+      "loss": 0.1414,
+      "step": 19477
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0914969007378266,
+      "learning_rate": 4.742305639221512e-09,
+      "loss": 0.1686,
+      "step": 19478
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2472389434505498,
+      "learning_rate": 4.6917253632039475e-09,
+      "loss": 0.1683,
+      "step": 19479
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0606679946176014,
+      "learning_rate": 4.64141621048797e-09,
+      "loss": 0.1839,
+      "step": 19480
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1715536705735645,
+      "learning_rate": 4.591378182438044e-09,
+      "loss": 0.1557,
+      "step": 19481
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2390844663128597,
+      "learning_rate": 4.541611280410862e-09,
+      "loss": 0.1746,
+      "step": 19482
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8912082389176219,
+      "learning_rate": 4.492115505757566e-09,
+      "loss": 0.1536,
+      "step": 19483
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.905352948158156,
+      "learning_rate": 4.442890859820414e-09,
+      "loss": 0.1621,
+      "step": 19484
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0956200005472319,
+      "learning_rate": 4.393937343933896e-09,
+      "loss": 0.1582,
+      "step": 19485
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8536426185940816,
+      "learning_rate": 4.345254959426948e-09,
+      "loss": 0.1571,
+      "step": 19486
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.229837492542071,
+      "learning_rate": 4.296843707619625e-09,
+      "loss": 0.1574,
+      "step": 19487
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.95821743748653,
+      "learning_rate": 4.2487035898242106e-09,
+      "loss": 0.161,
+      "step": 19488
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3127235685010321,
+      "learning_rate": 4.200834607348547e-09,
+      "loss": 0.1621,
+      "step": 19489
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2313703140225252,
+      "learning_rate": 4.153236761488266e-09,
+      "loss": 0.173,
+      "step": 19490
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4217470558916299,
+      "learning_rate": 4.105910053536777e-09,
+      "loss": 0.1551,
+      "step": 19491
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0070122397761707,
+      "learning_rate": 4.058854484777497e-09,
+      "loss": 0.1475,
+      "step": 19492
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1355502059955587,
+      "learning_rate": 4.012070056484963e-09,
+      "loss": 0.1322,
+      "step": 19493
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0643535971160303,
+      "learning_rate": 3.965556769930379e-09,
+      "loss": 0.1466,
+      "step": 19494
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9244812085559133,
+      "learning_rate": 3.9193146263749595e-09,
+      "loss": 0.1604,
+      "step": 19495
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3720180064592766,
+      "learning_rate": 3.873343627073256e-09,
+      "loss": 0.1453,
+      "step": 19496
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1020529971092021,
+      "learning_rate": 3.827643773270939e-09,
+      "loss": 0.1655,
+      "step": 19497
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8882727079169191,
+      "learning_rate": 3.782215066208128e-09,
+      "loss": 0.1532,
+      "step": 19498
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8608176999058977,
+      "learning_rate": 3.73705750711717e-09,
+      "loss": 0.138,
+      "step": 19499
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6619832340491163,
+      "learning_rate": 3.692171097223751e-09,
+      "loss": 0.1516,
+      "step": 19500
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9864753391422965,
+      "learning_rate": 3.647555837744676e-09,
+      "loss": 0.1703,
+      "step": 19501
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8821308274352536,
+      "learning_rate": 3.603211729890088e-09,
+      "loss": 0.1549,
+      "step": 19502
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0563855911607007,
+      "learning_rate": 3.5591387748634687e-09,
+      "loss": 0.1603,
+      "step": 19503
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.144958580490145,
+      "learning_rate": 3.5153369738583078e-09,
+      "loss": 0.156,
+      "step": 19504
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4642207318698972,
+      "learning_rate": 3.471806328065874e-09,
+      "loss": 0.1537,
+      "step": 19505
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2838464816610706,
+      "learning_rate": 3.428546838664115e-09,
+      "loss": 0.1542,
+      "step": 19506
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1749911154214572,
+      "learning_rate": 3.3855585068287564e-09,
+      "loss": 0.1595,
+      "step": 19507
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9491328852325978,
+      "learning_rate": 3.3428413337244224e-09,
+      "loss": 0.145,
+      "step": 19508
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0850935098083156,
+      "learning_rate": 3.3003953205101857e-09,
+      "loss": 0.1725,
+      "step": 19509
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.13078152777645,
+      "learning_rate": 3.2582204683362372e-09,
+      "loss": 0.1457,
+      "step": 19510
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3915491422700175,
+      "learning_rate": 3.216316778348327e-09,
+      "loss": 0.1465,
+      "step": 19511
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3624287232659398,
+      "learning_rate": 3.1746842516833243e-09,
+      "loss": 0.1536,
+      "step": 19512
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9363607116738734,
+      "learning_rate": 3.1333228894692147e-09,
+      "loss": 0.1725,
+      "step": 19513
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1789056989836593,
+      "learning_rate": 3.092232692827324e-09,
+      "loss": 0.1567,
+      "step": 19514
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1716631678255858,
+      "learning_rate": 3.0514136628745363e-09,
+      "loss": 0.1741,
+      "step": 19515
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9295582498624071,
+      "learning_rate": 3.0108658007155235e-09,
+      "loss": 0.1647,
+      "step": 19516
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4339393414590325,
+      "learning_rate": 2.970589107452737e-09,
+      "loss": 0.1936,
+      "step": 19517
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.183292046360521,
+      "learning_rate": 2.930583584176416e-09,
+      "loss": 0.1558,
+      "step": 19518
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7651834428055864,
+      "learning_rate": 2.890849231973469e-09,
+      "loss": 0.1538,
+      "step": 19519
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5131433513827075,
+      "learning_rate": 2.851386051919702e-09,
+      "loss": 0.1729,
+      "step": 19520
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6187138358686646,
+      "learning_rate": 2.8121940450875907e-09,
+      "loss": 0.1875,
+      "step": 19521
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.12632087757575,
+      "learning_rate": 2.7732732125396177e-09,
+      "loss": 0.168,
+      "step": 19522
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.022487523217451,
+      "learning_rate": 2.7346235553304955e-09,
+      "loss": 0.1642,
+      "step": 19523
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.059501541651189,
+      "learning_rate": 2.696245074509385e-09,
+      "loss": 0.1501,
+      "step": 19524
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9766296923988161,
+      "learning_rate": 2.6581377711176747e-09,
+      "loss": 0.158,
+      "step": 19525
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0403942539814275,
+      "learning_rate": 2.620301646188983e-09,
+      "loss": 0.194,
+      "step": 19526
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1544990838019817,
+      "learning_rate": 2.5827367007491555e-09,
+      "loss": 0.1649,
+      "step": 19527
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7138085897003958,
+      "learning_rate": 2.545442935816267e-09,
+      "loss": 0.1536,
+      "step": 19528
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8899480974681093,
+      "learning_rate": 2.5084203524039507e-09,
+      "loss": 0.157,
+      "step": 19529
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9575638072030332,
+      "learning_rate": 2.4716689515147386e-09,
+      "loss": 0.1568,
+      "step": 19530
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5434583311996226,
+      "learning_rate": 2.4351887341467206e-09,
+      "loss": 0.148,
+      "step": 19531
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.157022762265355,
+      "learning_rate": 2.3989797012879957e-09,
+      "loss": 0.1785,
+      "step": 19532
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2032012954869988,
+      "learning_rate": 2.363041853922221e-09,
+      "loss": 0.1566,
+      "step": 19533
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.016008852523729,
+      "learning_rate": 2.327375193024173e-09,
+      "loss": 0.1645,
+      "step": 19534
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8736000709689397,
+      "learning_rate": 2.291979719559745e-09,
+      "loss": 0.1517,
+      "step": 19535
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8238877899880166,
+      "learning_rate": 2.25685543449039e-09,
+      "loss": 0.1519,
+      "step": 19536
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0281866407460118,
+      "learning_rate": 2.2220023387686805e-09,
+      "loss": 0.1596,
+      "step": 19537
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8196043357201952,
+      "learning_rate": 2.1874204333394157e-09,
+      "loss": 0.1506,
+      "step": 19538
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5190087216009494,
+      "learning_rate": 2.1531097191418438e-09,
+      "loss": 0.1812,
+      "step": 19539
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.078921323523072,
+      "learning_rate": 2.1190701971052218e-09,
+      "loss": 0.1842,
+      "step": 19540
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.16461521444874,
+      "learning_rate": 2.0853018681532557e-09,
+      "loss": 0.1488,
+      "step": 19541
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0488166753749382,
+      "learning_rate": 2.051804733202989e-09,
+      "loss": 0.1616,
+      "step": 19542
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9667513288660236,
+      "learning_rate": 2.018578793161474e-09,
+      "loss": 0.1601,
+      "step": 19543
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9142594167680658,
+      "learning_rate": 1.985624048931323e-09,
+      "loss": 0.1453,
+      "step": 19544
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9533579569326212,
+      "learning_rate": 1.952940501405154e-09,
+      "loss": 0.176,
+      "step": 19545
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8812100501930893,
+      "learning_rate": 1.9205281514700356e-09,
+      "loss": 0.1419,
+      "step": 19546
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.502962961960015,
+      "learning_rate": 1.8883870000063753e-09,
+      "loss": 0.154,
+      "step": 19547
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0899422449191924,
+      "learning_rate": 1.856517047883477e-09,
+      "loss": 0.1636,
+      "step": 19548
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9187958817022984,
+      "learning_rate": 1.8249182959684253e-09,
+      "loss": 0.1584,
+      "step": 19549
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8635351997259035,
+      "learning_rate": 1.7935907451172019e-09,
+      "loss": 0.1413,
+      "step": 19550
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.0210498380084243,
+      "learning_rate": 1.7625343961791275e-09,
+      "loss": 0.1537,
+      "step": 19551
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.689591606566535,
+      "learning_rate": 1.7317492499968614e-09,
+      "loss": 0.165,
+      "step": 19552
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3730827056860522,
+      "learning_rate": 1.7012353074052912e-09,
+      "loss": 0.1652,
+      "step": 19553
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9159153423477189,
+      "learning_rate": 1.6709925692326435e-09,
+      "loss": 0.1768,
+      "step": 19554
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.328847538504899,
+      "learning_rate": 1.6410210362993729e-09,
+      "loss": 0.1754,
+      "step": 19555
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9906761771767876,
+      "learning_rate": 1.6113207094181626e-09,
+      "loss": 0.1257,
+      "step": 19556
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8607195647789683,
+      "learning_rate": 1.5818915893939246e-09,
+      "loss": 0.152,
+      "step": 19557
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.877998645264861,
+      "learning_rate": 1.5527336770260193e-09,
+      "loss": 0.1637,
+      "step": 19558
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.148237424551445,
+      "learning_rate": 1.5238469731049254e-09,
+      "loss": 0.1579,
+      "step": 19559
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3199269131975926,
+      "learning_rate": 1.4952314784144606e-09,
+      "loss": 0.1764,
+      "step": 19560
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5449672703207997,
+      "learning_rate": 1.4668871937306706e-09,
+      "loss": 0.1767,
+      "step": 19561
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9551654351745247,
+      "learning_rate": 1.4388141198218297e-09,
+      "loss": 0.1352,
+      "step": 19562
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9266927361479481,
+      "learning_rate": 1.4110122574506612e-09,
+      "loss": 0.1586,
+      "step": 19563
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8555127548715358,
+      "learning_rate": 1.3834816073687862e-09,
+      "loss": 0.1689,
+      "step": 19564
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.947450765781288,
+      "learning_rate": 1.3562221703267153e-09,
+      "loss": 0.1448,
+      "step": 19565
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5330077115508238,
+      "learning_rate": 1.3292339470605264e-09,
+      "loss": 0.168,
+      "step": 19566
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.185404398834737,
+      "learning_rate": 1.302516938304077e-09,
+      "loss": 0.1527,
+      "step": 19567
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3458661533631802,
+      "learning_rate": 1.2760711447812324e-09,
+      "loss": 0.1749,
+      "step": 19568
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8806027444964389,
+      "learning_rate": 1.249896567210307e-09,
+      "loss": 0.1538,
+      "step": 19569
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3762344261508295,
+      "learning_rate": 1.2239932062996229e-09,
+      "loss": 0.1651,
+      "step": 19570
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3219725268163043,
+      "learning_rate": 1.198361062754172e-09,
+      "loss": 0.1637,
+      "step": 19571
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.108314030942384,
+      "learning_rate": 1.1730001372667332e-09,
+      "loss": 0.1611,
+      "step": 19572
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1122036596028475,
+      "learning_rate": 1.1479104305267553e-09,
+      "loss": 0.1613,
+      "step": 19573
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.960709597396094,
+      "learning_rate": 1.1230919432148046e-09,
+      "loss": 0.1858,
+      "step": 19574
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.322649053633847,
+      "learning_rate": 1.0985446760036766e-09,
+      "loss": 0.1546,
+      "step": 19575
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0212454361300904,
+      "learning_rate": 1.074268629559505e-09,
+      "loss": 0.1615,
+      "step": 19576
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9147610824352448,
+      "learning_rate": 1.050263804539542e-09,
+      "loss": 0.1607,
+      "step": 19577
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.155984862637736,
+      "learning_rate": 1.0265302015965984e-09,
+      "loss": 0.1655,
+      "step": 19578
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.158209229203261,
+      "learning_rate": 1.0030678213746037e-09,
+      "loss": 0.1675,
+      "step": 19579
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0781213576452073,
+      "learning_rate": 9.798766645074953e-10,
+      "loss": 0.1765,
+      "step": 19580
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9225119895905177,
+      "learning_rate": 9.569567316269901e-10,
+      "loss": 0.162,
+      "step": 19581
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3798220622535513,
+      "learning_rate": 9.343080233537028e-10,
+      "loss": 0.1559,
+      "step": 19582
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5883719794381368,
+      "learning_rate": 9.119305403015865e-10,
+      "loss": 0.1681,
+      "step": 19583
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9920192181846645,
+      "learning_rate": 8.898242830779336e-10,
+      "loss": 0.1702,
+      "step": 19584
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.011582731494466,
+      "learning_rate": 8.679892522833744e-10,
+      "loss": 0.152,
+      "step": 19585
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8377799966259223,
+      "learning_rate": 8.464254485096579e-10,
+      "loss": 0.1716,
+      "step": 19586
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0654245746696522,
+      "learning_rate": 8.251328723407615e-10,
+      "loss": 0.2018,
+      "step": 19587
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0281990884378034,
+      "learning_rate": 8.04111524354001e-10,
+      "loss": 0.1515,
+      "step": 19588
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0625756720875446,
+      "learning_rate": 7.833614051222515e-10,
+      "loss": 0.1688,
+      "step": 19589
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0201727001752259,
+      "learning_rate": 7.628825152050656e-10,
+      "loss": 0.1583,
+      "step": 19590
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9110095625867899,
+      "learning_rate": 7.426748551597751e-10,
+      "loss": 0.1504,
+      "step": 19591
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1998072018181751,
+      "learning_rate": 7.227384255348302e-10,
+      "loss": 0.1589,
+      "step": 19592
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8852849233106013,
+      "learning_rate": 7.030732268697993e-10,
+      "loss": 0.2041,
+      "step": 19593
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0241911357987625,
+      "learning_rate": 6.836792596986996e-10,
+      "loss": 0.1654,
+      "step": 19594
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.08609554904953,
+      "learning_rate": 6.64556524547777e-10,
+      "loss": 0.1553,
+      "step": 19595
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.56946727399011,
+      "learning_rate": 6.457050219355054e-10,
+      "loss": 0.1606,
+      "step": 19596
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5602719586550857,
+      "learning_rate": 6.271247523736978e-10,
+      "loss": 0.1621,
+      "step": 19597
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.318659581015296,
+      "learning_rate": 6.088157163652853e-10,
+      "loss": 0.1669,
+      "step": 19598
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9424409032077522,
+      "learning_rate": 5.907779144076475e-10,
+      "loss": 0.1562,
+      "step": 19599
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9982593233791778,
+      "learning_rate": 5.730113469903931e-10,
+      "loss": 0.1598,
+      "step": 19600
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5222263838126218,
+      "learning_rate": 5.555160145942485e-10,
+      "loss": 0.1492,
+      "step": 19601
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0665216789776741,
+      "learning_rate": 5.38291917694389e-10,
+      "loss": 0.1641,
+      "step": 19602
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2836376729407852,
+      "learning_rate": 5.21339056759329e-10,
+      "loss": 0.1614,
+      "step": 19603
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8861752562480091,
+      "learning_rate": 5.046574322464803e-10,
+      "loss": 0.1605,
+      "step": 19604
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.573271589681279,
+      "learning_rate": 4.882470446099241e-10,
+      "loss": 0.1373,
+      "step": 19605
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3650632783304832,
+      "learning_rate": 4.721078942948598e-10,
+      "loss": 0.1643,
+      "step": 19606
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9852357360213777,
+      "learning_rate": 4.562399817376051e-10,
+      "loss": 0.1739,
+      "step": 19607
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9335982121563716,
+      "learning_rate": 4.406433073711469e-10,
+      "loss": 0.1748,
+      "step": 19608
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2664477625014325,
+      "learning_rate": 4.253178716162598e-10,
+      "loss": 0.1724,
+      "step": 19609
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9898241612198758,
+      "learning_rate": 4.1026367488927745e-10,
+      "loss": 0.1871,
+      "step": 19610
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8668942232628565,
+      "learning_rate": 3.9548071759876185e-10,
+      "loss": 0.1578,
+      "step": 19611
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.016572741939258,
+      "learning_rate": 3.809690001455035e-10,
+      "loss": 0.152,
+      "step": 19612
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0372104824857318,
+      "learning_rate": 3.667285229236317e-10,
+      "loss": 0.1711,
+      "step": 19613
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2856759711969599,
+      "learning_rate": 3.5275928631839375e-10,
+      "loss": 0.1505,
+      "step": 19614
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2258181466335094,
+      "learning_rate": 3.390612907094859e-10,
+      "loss": 0.1553,
+      "step": 19615
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9124728749818874,
+      "learning_rate": 3.256345364688329e-10,
+      "loss": 0.1467,
+      "step": 19616
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9079859042074966,
+      "learning_rate": 3.124790239594777e-10,
+      "loss": 0.1459,
+      "step": 19617
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8911178077678168,
+      "learning_rate": 2.995947535389121e-10,
+      "loss": 0.1631,
+      "step": 19618
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.200090105682176,
+      "learning_rate": 2.8698172555685634e-10,
+      "loss": 0.1464,
+      "step": 19619
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.945105130525938,
+      "learning_rate": 2.746399403552591e-10,
+      "loss": 0.1725,
+      "step": 19620
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1593818063618362,
+      "learning_rate": 2.6256939826940774e-10,
+      "loss": 0.1595,
+      "step": 19621
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.080925000211112,
+      "learning_rate": 2.5077009962570784e-10,
+      "loss": 0.1645,
+      "step": 19622
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9285369238641832,
+      "learning_rate": 2.392420447450139e-10,
+      "loss": 0.1516,
+      "step": 19623
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3598322147140447,
+      "learning_rate": 2.279852339392985e-10,
+      "loss": 0.1464,
+      "step": 19624
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0434638990818221,
+      "learning_rate": 2.1699966751387303e-10,
+      "loss": 0.1612,
+      "step": 19625
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9800285397183612,
+      "learning_rate": 2.0628534576738746e-10,
+      "loss": 0.1652,
+      "step": 19626
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5006060025917467,
+      "learning_rate": 1.9584226898961e-10,
+      "loss": 0.1796,
+      "step": 19627
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.118456926792813,
+      "learning_rate": 1.85670437465868e-10,
+      "loss": 0.1532,
+      "step": 19628
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3154930824584454,
+      "learning_rate": 1.757698514692763e-10,
+      "loss": 0.1737,
+      "step": 19629
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2594644655238652,
+      "learning_rate": 1.6614051127072929e-10,
+      "loss": 0.1776,
+      "step": 19630
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2309343752309614,
+      "learning_rate": 1.567824171300192e-10,
+      "loss": 0.1664,
+      "step": 19631
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4860321980415303,
+      "learning_rate": 1.4769556930138707e-10,
+      "loss": 0.1441,
+      "step": 19632
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9653230918868438,
+      "learning_rate": 1.3887996803130242e-10,
+      "loss": 0.1638,
+      "step": 19633
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5851984809016977,
+      "learning_rate": 1.3033561355846324e-10,
+      "loss": 0.1601,
+      "step": 19634
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1925293052215677,
+      "learning_rate": 1.2206250611490612e-10,
+      "loss": 0.1721,
+      "step": 19635
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0617452077781544,
+      "learning_rate": 1.1406064592600629e-10,
+      "loss": 0.1635,
+      "step": 19636
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0323618468439235,
+      "learning_rate": 1.063300332082573e-10,
+      "loss": 0.1347,
+      "step": 19637
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2798116231535444,
+      "learning_rate": 9.887066817038105e-11,
+      "loss": 0.1562,
+      "step": 19638
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9078167713684868,
+      "learning_rate": 9.168255101554835e-11,
+      "loss": 0.148,
+      "step": 19639
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9109382670862405,
+      "learning_rate": 8.476568193804824e-11,
+      "loss": 0.1563,
+      "step": 19640
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9894577234866273,
+      "learning_rate": 7.812006112661863e-11,
+      "loss": 0.1612,
+      "step": 19641
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0206860112046388,
+      "learning_rate": 7.174568876111565e-11,
+      "loss": 0.1626,
+      "step": 19642
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3743462768995467,
+      "learning_rate": 6.56425650147341e-11,
+      "loss": 0.1498,
+      "step": 19643
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8318224179362256,
+      "learning_rate": 5.981069005178697e-11,
+      "loss": 0.1659,
+      "step": 19644
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.247774085231404,
+      "learning_rate": 5.425006403214639e-11,
+      "loss": 0.1671,
+      "step": 19645
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2683171725518405,
+      "learning_rate": 4.8960687104582235e-11,
+      "loss": 0.17,
+      "step": 19646
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.001840363433565,
+      "learning_rate": 4.3942559414533734e-11,
+      "loss": 0.1918,
+      "step": 19647
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2938793167991183,
+      "learning_rate": 3.919568109744809e-11,
+      "loss": 0.181,
+      "step": 19648
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0800015855464375,
+      "learning_rate": 3.472005228211117e-11,
+      "loss": 0.1809,
+      "step": 19649
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0588591780403915,
+      "learning_rate": 3.051567308953729e-11,
+      "loss": 0.1805,
+      "step": 19650
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2033718217226848,
+      "learning_rate": 2.6582543634079416e-11,
+      "loss": 0.1423,
+      "step": 19651
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4721066909443041,
+      "learning_rate": 2.292066402120874e-11,
+      "loss": 0.1634,
+      "step": 19652
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.723042537451307,
+      "learning_rate": 1.9530034353065775e-11,
+      "loss": 0.1664,
+      "step": 19653
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9442132049513063,
+      "learning_rate": 1.6410654719578588e-11,
+      "loss": 0.1605,
+      "step": 19654
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9540795193476649,
+      "learning_rate": 1.3562525205124134e-11,
+      "loss": 0.1505,
+      "step": 19655
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1669133271587695,
+      "learning_rate": 1.0985645887418017e-11,
+      "loss": 0.1659,
+      "step": 19656
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0352443925181587,
+      "learning_rate": 8.680016837514516e-12,
+      "loss": 0.1578,
+      "step": 19657
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2268663110505775,
+      "learning_rate": 6.645638116475894e-12,
+      "loss": 0.1621,
+      "step": 19658
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1273871387987642,
+      "learning_rate": 4.882509779813305e-12,
+      "loss": 0.1569,
+      "step": 19659
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3269521982201904,
+      "learning_rate": 3.39063187637656e-12,
+      "loss": 0.171,
+      "step": 19660
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2765834298934498,
+      "learning_rate": 2.1700044450234657e-12,
+      "loss": 0.1679,
+      "step": 19661
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9155996421638892,
+      "learning_rate": 1.2206275190607132e-12,
+      "loss": 0.1621,
+      "step": 19662
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8992063234029176,
+      "learning_rate": 5.425011262438773e-13,
+      "loss": 0.137,
+      "step": 19663
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.401467580475726,
+      "learning_rate": 1.3562528211608085e-13,
+      "loss": 0.1682,
+      "step": 19664
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0157853651703213,
+      "learning_rate": 0.0,
+      "loss": 0.1681,
+      "step": 19665
+    },
+    {
+      "epoch": 1.0,
+      "step": 19665,
+      "total_flos": 4376372611039232.0,
+      "train_loss": 0.19586870987702348,
+      "train_runtime": 255341.6294,
+      "train_samples_per_second": 19.716,
+      "train_steps_per_second": 0.077
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 19665,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 10000,
+  "total_flos": 4376372611039232.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}